Internet Explorer Performance Lab: reliably measuring browser performance


A big part of this blog is going behind the scenes to show you all the work that goes into the engineering of Windows 8.  In this post we take a look at something we all care very deeply about–as engineers and as end-users–real world web performance.  We do a huge amount of work to get beyond the basics of anecdotes and feel as we work to build high performance web browsing.  This post is authored by Matt Kotsenas, Jatinder Mann, and Jason Weber on the IE team, though performance is something that every single member of the team works on.  –Steven

Web performance matters to everyone, and one engineering objective for Internet Explorer is to be the world’s fastest browser. To achieve this goal we need to reliably measure browser performance against the real world scenarios that matter to our customers. Over the last five years we designed and built the Internet Explorer Performance Lab, one of the world’s most sophisticated web performance measurement systems.

The IE Performance Lab collects reliable, accurate, and actionable data to inform decisions throughout the development cycle. We measure the performance of Internet Explorer 200 times daily, collecting over 5.7 million measurements and 480GB of runtime data each day. We understand the impact of every change to the product and ensure that Internet Explorer only gets faster. This blog post takes a deep look at how the IE Performance Lab is designed and how we use the lab to ensure we’re continually making the web faster.

In this post, we present:

  • Overview of the IE Performance Lab
  • Lab infrastructure
  • What (and how) we measure
  • Testing a scenario
  • Results investigation
  • Testing third-party software
  • Building a fast browser for users

Overview of the IE Performance Lab

In order to reliably measure web performance over time, a system needs to be able to reproducibly simulate real world user scenarios. In essence, our system needs to create a “mini version of the Internet.”

The IE Performance Lab is a private network completely sealed from both the public Internet and the Microsoft intranet network, and contains over 140 machines. The lab contains the key pieces of the real Internet, including web servers, DNS servers, routers, and network emulators, which simulate different customer connectivity scenarios.

Although this may appear complex at first glance, this approach allows all sources of variance to be removed. By controlling every aspect of the network, down to individual packet hops and latencies, our tests become deterministic and repeatable, which is critical to making the results actionable. In the IE Performance Lab, activity is measured with 100 nanosecond resolution.

Diagram shows content servers connected to Network emulators, connected to DNS servers, connected to Test clients, connected to Raw data storage, connected to Data analysis, connected to SQL database.

This type of network configuration also provides a great amount of flexibility. Because we’re simulating a real world setup, our lab can accommodate nearly any type of test machine or website content. The IE Performance Lab supports desktops, laptops, netbooks, and tablets with x86, x64, and ARM processors, all simultaneously. Similarly, because the lab uses the Windows Performance Tools (WPT), we can run the same tests using different web browsers, toolbars, anti-virus products, or other third-party software and directly compare the results.

WPT provides deep insight into the underlying hardware. Using WPT, we can capture everything from high-level CPU and GPU activity, to low-level information such as cache efficiency, networking statistics, memory usage patterns, and more. WPT allows us to measure and optimize performance across the stack to ensure that the hardware, device drivers, Windows operating system, and Internet Explorer are all efficiently optimized together.

A single test run takes 6 hours to complete and generates over 22GB of data during that time. This highly automated system is staffed by a small team that monitors operations, analyzes results, and develops new infrastructure features.

Lab infrastructure

The Performance Lab infrastructure can be broken into three main categories: Network and Server, Test Clients, and Analysis and Reporting. Each category is designed to minimize interaction across components, both to improve scalability of testing and to reduce the possibility of introducing noise into the lab environment.

A large room full of computers

Here’s a view of the IE Performance Lab, including a number of test and analysis machines on our private network.

Network and server infrastructure

Let’s start by discussing the DNS servers, network emulators, and content servers; all the components that together create the mini Internet. Over the next three sections we’ll work our way from right to left in the architectural diagram.

Content servers

Content servers are web servers that stand in for the millions of web hosts on the Internet. Each content server hosts real world web pages that have been captured locally. The captured pages go through a process we refer to as sanitization, where we tweak portions of the web content to ensure reproducible determinism. For example, JavaScript Date functions or Math.Random() calls will be replaced with a static value. Additionally, the dynamic URLs created by ad frameworks are locked to the URL that was first used by the framework.

After sanitization, content is served similarly to static content through an ISAPI filter that maps a hash of the URL to the content, allowing instantaneous lookup. Each web server is a 16-core machine with 16GB of RAM to minimize variability and ensure that content is in memory (no disk access required).

Content servers can also host dynamic web apps like Outlook Web Access or Office Web Apps. In these cases, the application server and any multi-tier dependencies are hosted on dedicated servers in the lab, just like real world environments.

Network emulators

Since many sources of variability have been removed, network speeds no longer reflect the experiences of many users with slower connections. To simulate real world customer environments, a test can take advantage of network emulation to understand the performance across the wide range of networks in use today. The lab supports emulating several DSL configurations, cable modems, 56k modems, as well as high-bandwidth, high-latency environments like WAN and 4G environments. As HTTP requests are passed to the emulator, it simulates network characteristics like packet delay and reordering, then forwards the request on to the web hosts. Upon receiving a response, emulation is again applied and then passed back to the test client.

Using dedicated hardware for network emulation provides the most realistic testing environment possible, and significantly reduces the observer effect. Although dedicated hardware adds cost and complexity compared to proxy or software-based solutions, it’s the only way to accurately measure performance. Browsers limit the number of simultaneous proxy connections to prevent proxy saturation, so using a proxy for network emulation has the unintended effect of sidestepping domain sharding and other optimizations made by the webpage. Additionally, local network emulation will compete with the browser for local machine resources, especially on low power machines.

DNS servers

Like real world DNS servers, the lab’s DNS servers link the content servers to the test clients. The lab also uses a different DNS server for each network emulator, meaning that changing from one network speed to another is as simple as changing the DNS server. In these cases, instead of resolving domain names to the web hosts, the DNS server resolves all domain names to the associated network emulator.

Test client configurations

We want to ensure that Internet Explorer consistently runs fast across all classes of computer hardware. The lab contains over 120 computers used to measure Internet Explorer performance. We refer to these as test clients; they range from high-end x64 desktops, to low-powered netbooks, to touch-first tablet devices, and everything in between. Because repeatability of measurements is paramount, all test clients are physical machines.

A long desk and two shelves, each containing 12 or more computers

Internet Explorer Performance Lab change comparison machine pool

Different machine classes contain both discrete and integrated graphics platforms to ensure that Internet Explorer continues to take full advantage of hardware acceleration across the ecosystem of devices. Above is our main machine pool. These PCs approximate the average consumer experience over the lifetime of a Windows 7 or Windows 8 PC. Machines are ordered from the OEM to be identical; they all come from the same manufacturing lot and their performance characteristics are verified prior to use.

Since the lab runs 24/7, hardware failures are inevitable. Replacing failed components with identical parts from a different manufacturing lot almost always results in the repaired computer running faster than the other machines in the pool. While this difference would be unnoticeable in the real world, when you’re measuring down to 100 nanoseconds, even a few cycles can impact the results! If after a repair a machine no longer runs identically to the rest of the pool, it is removed from the lab and the pool’s size permanently shrinks. In response, the lab’s purchases include extra “buffer” machines, so that when a failed machine is removed from the pool, the excess capacity provides a cushion, and the lab’s operations are not affected.

Pool Name

# Machines

Form Factor

Processor

Arch

Clock Speed

RAM

Graphics

Main Pool

32

Desktop

Core i5 750 (Lynnfield)

64-bit

2.66GHz

4096MB DDR3

NVIDIA GeForce 310

To add hardware breadth, we have additional machine pools that run the spectrum of consumer scenarios. Good performance on these machines ensures that IE uses the underlying hardware effectively across the PC ecosystem.

Pool Name

# Machines

Form Factor

Processor

Arch

Clock Speed

RAM

Graphics

High‑end 1

20

Desktop

Core i7 870

64‑bit

2.93GHz

4096MB DDR3

ATI Radeon HD 4550

High‑end 2

4

Desktop

Xeon 5150 (Woodcrest)

64‑bit

2.66GHz

8192MB DDR2

ATI Radeon X1950 Pro

Mid‑range 1

6

Desktop

Core 2 Duo (Wolfdale)

64‑bit

3.0GHz

4096MB DDR2

Intel GMA 4500

Mid‑range 2

15

Desktop

Core 2 Duo E6750

64‑bit

2.66GHz

4096MB DDR2

ATI Radeon HD 2400 XT

Mid‑range 3

25

Desktop

Core i5 2500

64‑bit

3.30GHz

4096MB DDR3

Intel HD Graphics 2000

Mid‑range 4

6

Desktop

Core 2 Duo (Conroe)

64‑bit

2.66GHz

4096MB DDR2

ATI Radeon HD 2400XT

Mid‑range 5

4

Desktop

Core 2 Duo (Conroe)

64‑bit

2.4GHz

4096MB DDR2

ATI Radeon X1950 Pro

Low‑power 1

2

Laptop

Atom Z530

32‑bit

1.6GHz

2038MB DDR2

Intel GMA 500

Low‑power 2

4

Netbook

Atom N270

32‑bit

1.6GHz

1024MB DDR2

NVIDIA ION

Low‑power 3

2

Netbook

Atom N450

64‑bit

1.66GHz

1024MB DDR2

Intel GMA 3150

Low‑power 4

4

Netbook

Atom N270

32‑bit

1.6GHz

1024MB DDR2

Intel GMA950

Low‑power 5

4

Slate

ARM

32‑bit

Prototype hardware

Assortment of laptop and desktop PCs on two shelves

Low-powered test machines. Each one is in a different state of testing.

If even more diversity is needed, the IE Performance Lab can also make use of the Windows Graphics Lab. The Windows Graphics Lab stocks nearly every graphics chipset manufactured. PCs can be configured into nearly any permutation imaginable and then used for performance testing. The Windows Graphics Lab is invaluable for diagnosing graphics problems across chipsets and driver revisions.

Analysis and reporting servers

Collection and analysis of test results are divided into two separate steps. By offloading analysis to dedicated machines, the test clients can begin another performance run earlier, and more powerful server class machines can be used to perform the analysis more rapidly. The sooner the analysis completes, the more efficiently we can identify performance changes.

For analysis, we use 11 server class machines, each of which has 16 cores and 16GB of RAM. During analysis, each trace file is inspected and thousands of metrics are extracted and inserted into a SQL server. Over the course of 24 hours these analysis machines will inspect over 15,000 traces that will be used for trend analysis.

Two server racks

Pictured are two of several server racks which contain file servers, a SQL server, and several analysis and content servers.

The SQL Server used to store the nearly 6 million measurements we collect each day is a 24 logical core machine with 64GB of RAM. Reports can be generated directly from SQL, or results can be inspected using either an HTML-based comparison application or WCF service that provides results in XML or JSON formats.

What (and how) we measure

With the infrastructure in place, let’s review the different types of scenarios measured in the Performance Lab, and the tools we use to gather metrics.

Scenarios measured daily

The Performance Lab focuses on real world scenarios that matter to users. As a result, we run over 20,000 different tests daily. These tests fall into four, sometimes overlapping, categories:

4 overlapping circles: Loading Content, Interactive Web Apps, IE "The Application", Synthetic Platform Benchmarks

Loading content – Navigating from one page to another is still the most common activity inside a web browser. Loading web content is also the only category that touches most of the browser’s eleven subsystems. Loading web content is a prerequisite for the other categories of scenarios.

Interactive web apps – This category covers what is sometimes referred to as content creation, AJAX applications, or Web 2.0 sites. It includes interacting with popular news and social networking sites as well as interacting with mail and document applications like Outlook Web Access and Office Web Apps.

IE “the application” – Important but often forgotten are scenarios that interact with the browser itself. Common interactions include opening or closing the browser, switching tabs, using browser features like History and Favorites, and panning and zooming with both keyboard and mouse, and touch inputs.

Synthetic benchmarks – Rarely forgotten but often overstated are synthetic benchmarks like WebKit SunSpider. Benchmarks can be a useful engineering tool as they are designed to stress individual browser subsystems and accentuate differences between browsers. However, in order to maximize those differences, benchmarks often resort to atypical usage patterns or edge cases.

Real world patterns

When measuring performance it is important to ensure that the tests reflect real world usage patterns. Most Software Engineering textbooks refer to this process as workload modeling, or application usage modeling. To ensure that the Performance Lab measures real world patterns, the Performance Lab uses real Web pages that represent real world patterns and exercise different browser subsystems.

In order to determine which sites to use for testing, we regularly crawl millions of sites and compile a list of site attributes and coding patterns. We use 68 different data points to determine commonalities across sites – things like the depth and width of the resulting DOM, CSS layout patterns, common frameworks used, international features, and more. From the results we chose sites that best represent the common patterns and diversity of the broader Web.

Engineering metrics

Performance is a multi-dimensional problem. The only way to get an accurate view of performance is to understand the scenario you’re testing, and how the hardware and OS interact with the browser. Here’s a closer look at five important performance metrics in the context of loading a major sports site for the first time.

Chart comparing Display time, elapsed time, CPU time, resource uitilization, and power consumption

Display Time – Display Time measures the time from when the user performs an action until the user sees the result of that action on the screen.

Elapsed Time – Most sites continue to perform background work after content has been displayed to the screen. Examples might include downloading the next email in a web mail application or sending analytics back to a provider. From the user’s perspective, the site might appear finished; however, significant work is often occurring which can impact overall responsiveness.

CPU Time – Modern web browsers are almost exclusively limited by the speed of the CPU. Offloading work to the GPU and making the CPU more efficient makes a large impact on performance.

Resource Utilization – Building a fast browser means ensuring resources across the entire PC work well together, including network utilization, memory usage patterns, GPU processing, graphics, memory, and hundreds of other dimensions. Since users run several applications at the same time on their PCs, it’s important for browsers to responsibly share these resources with other applications.

Power Consumption – Increasing power efficiency leads to longer the battery life in mobile scenarios, lower electricity costs for the device, and a smaller environmental impact.

Concentrating only on a single metric creates an overly simplistic view of performance. By focusing on a single metric, humans naturally tend to optimize for that metric, often at the expense of other equally important metrics. The only way to combat that tendency is to measure all aspects of performance, and then make the tradeoffs consciously, rather than implicitly.

In total, the Performance Lab measures over 850 different metrics. Each one provides part of the picture of browser performance. To give a feel for what we measure, here’s a (non-exhaustive) list of key metrics: private working set, total working set, HTTP request count, TCP bytes received, number of binaries loaded, number of context switches, DWM video memory usage, percent GPU utilization, number of paints, CPU time in JavaScript garbage collection, CPU time in JavaScript parsing, average DWM update interval, peak total working set, number of heap allocations, size of heap allocations, number of outstanding heap allocations, size of outstanding heap allocations, CPU time in layout subsystem, CPU time in formatting subsystem, CPU time in rendering subsystem, CPU time in HTML parser subsystem, idle CPU time, number of threads.

Windows event tracing infrastructure

Metrics are gathered using Windows Event Tracing Infrastructure (ETW) and VMMap. ETW is the Windows-wide event logging system that is used by many Windows components and third-party applications, including the Windows Event Log. ETW logging APIs are extremely low level and low overhead, which is critical for performance testing.

The view shows 6 graphs stacked vertically. Graphs are named CPU Usage by Process, Generic Events, WinINet End-to-End Downloads, IE CPU Breakdown, WinInet Transfer Setups, and IE Repaint.

The trace viewer included in WPT, xperfview.exe, is a powerful visualizer that allows correlation and overlaying kernel, CPU, GPU, I/O, networking, and other events. WPT also supports stack walking. Stack walking takes a snapshot of the program’s callstack at regular intervals and saves the stack as part of the trace. By correlating ETW events with stacks, WPT will display not only what work was being done, but the callstack associated with that work and the amount of time spent doing that work, with 10 microsecond resolution. Stack walking can be enabled on any process, even one that does not use ETW events. The drawback to stack walking is that it requires debugging symbols to decode the stacks, and is susceptible to aliasing.

Testing a scenario

The final piece of the puzzle is the actual test process. Testing can be broken into 3 phases: setup, testing, and errors and cleanup. Here’s a flowchart of the entire process to follow along.

A complex flow chart, starting with "User requests run" and ending with "Run is marked finished"

Setup

The process starts when a user requests a run through the lab website or automation framework. The run is placed into a priority queue with other pending runs. When a test client becomes available, it checks the queue and starts the highest priority job that it can. First, the test client installs the Test OS specified. The IE Performance Lab supports testing on Vista, Windows 7, and Windows 8. The test client installs a fresh copy of the Test OS for every run so the machine always starts in a known good state.

Once the Test OS is installed, the client configures WPT, VMMap, and the test harness. The run also specifies a number of IE settings such as the homepage, use of Suggested Sites, InPrivate browsing, and others. Any third-party software is also installed and configured at this point.

The final step before testing is ensuring that the test client is idle to minimize test interference. Windows defines a concept of idle tasks. Idle tasks are a way for Windows and other developers to schedule non-critical work to happen at a later time when the user is not competing for resources. OS idle tasks include prefetching or SuperFetching, disk defragmentation, updating search indexes, and others, depending on OS version and configured services. To ensure that no idle work is done during the tests, the idle task queue is flushed. Additionally, Windows Defender is paused and the log location for the test harness is marked as excluded from the Windows Indexing Service to prevent log and trace files from causing the indexer to start during a test run. Testing is done in multiple passes to minimize the number of providers needed, since additional providers increase the observer effect. The first pass is always a warm-up pass. Warm-up ensures that the browser binaries are “warm” and that the maximum amount of cachable page content is available in the WinINET cache. Subsequent passes each focus on a specific type of instrumentation, like stack walking, memory tracing, and I/O and registry tracing.

Errors and cleanup

If at any time during the test the browser crashes, the test pass is considered failed and the run moves on to the next test pass. If at any time during the tests Windows crashes, the computer reboots and the OS is reinstalled, since its state cannot be guaranteed. If the number of retries exceeds the threshold, the whole run is considered failed and the machine moves on to another run to prevent endlessly trying to test an unstable build.

When all the test cases are complete, the test client uploads the logs and traces for analysis. The test client then returns to an idle state and begins polling for a new run.

Results investigation

Each metric is tracked change-over-change. We run each test case a minimum of ten times, and duplicate runs on at least two different machines to create the sample population. Using statistical tools, uncharacteristic results can be automatically flagged for investigation. A variance change is also considered a regression. Users interact with IE under a wide range of circumstances and on a wide range of hardware, and one of our goals is to ensure a smooth and predictable experience every time.

In addition to automated analysis, a triage team investigates the daily results to watch for trends and other interesting behaviors. Manual investigation cannot be eliminated because many statistical tools assume both a normal distribution and that all samples are independent. Neither assumption may be strictly true for our measurements. Some activities in IE are driven by a timer from the OS, meaning results are also dependent on when (along the timer’s cycle) the page load begins. A page load that starts right before or after a timer interrupt may do more or less work because IE must service the interrupt at different points in the page-load process. This interruption can have a rippling affect that leads to a bimodal distribution. Also, because we use repeated trials (and we don’t wipe the machine between iterations) the next trial is influenced by previous trials. Here’s a sample Elapsed Time graph for Bing Maps for change-over-change comparison.

A bar chart with a red line superimposed. A mouse pointer hovers over one point in the chart, and next to this is a tooltip listing max, median, min, and other info.

The red series shows the median value of each test run, and grey bars show the range. Hovering over a test run will show the iterations for the metric (in blue) as well as a tooltip that provides the exact values for minimum, median, max values, as well as the absolute and relative difference with the previous test run. The tooltip shown in this image also provides additional context like the build being tested, and a quick link to our source control system to view the changes in the build.

The combination of automated analysis and manual investigation provides the IE team with reliable and actionable data for performance tuning.

Testing third-party software

Many third-party applications depend on Trident, the network stack, and other IE components. Extensions like BHOs and toolbars load within the IE context. Other applications, like security software, can inject themselves between IE components. These applications become part of the IE stack, and can lead to poor performance. The Performance Lab is capable of measuring the impact of third-party software on browsing real world content in a controlled environment. These studies are important to IE and the ecosystem because users generally cannot quantify the impact of popular software on their browsing experience.

When testing third-party software impact, we compare a run with third-party software installed, with a clean run with only IE installed, to determine the impact of the software. In particular, we are interested in measuring two metrics: startup time and navigation time. Startup time measures the time it takes to launch the browser and navigate to an URL, whereas navigation time measures the time it takes to navigate to an URL when the browser has already been launched. Startup will also include the time that third-party applications take to load their IE extensions.

Using cached content allows repeatability in our measurements. Further, by measuring a cached site, we can definitively know that a performance regression is caused by the third-party software and not by differences in the site. Whenever measuring the impact of third-party software, we also validate our findings by testing startup and navigation on a direct connection to the Internet, to verify that the testing environment is not responsible for any deltas.

Many third-party applications offload work during a page navigation to cloud services. While parallelization of work and use of cloud services are excellent techniques to improve performance, some applications wait synchronously for the results from the network, blocking the navigation in the process. There are many real world scenarios, like strict firewalls, WAN connections, and offline scenarios, where such patterns can lead to poor performance for users. Third-party software should never process synchronously in response to an IE or user action, and should batch UI and DOM updates to minimize disruption.

Building a fast browser for users

Real world browser performance matters. Measuring performance at scale is a significant investment and a full-time job, but the results are well worth the effort. The data gathered by the Internet Explorer Performance Lab is instrumental in our understanding of browser performance and of the underlying PC hardware, and in developing a fast, fluid, and responsive web experience for users.

—Matt Kotsenas, Jatinder Mann, and Jason Weber for the Internet Explorer Performance Team

Comments (143)

  1. Don says:

    And we thought Microsoft isn't doing anything about IE 🙂

  2. Yousaf says:

    This is probably the longest blog post I have ever come across. A world record? We write short posts on our blog http://www.elevatelocal.co.uk/blog

  3. FremyCompany says:

    Great article. I wonder how the Google and Microsoft browser testing labs compare.

  4. Ford says:

    What a long post… it is worth reading though…

  5. FremyCompany says:

    Another question: how do you deal with gradual computers' performance decrease?

  6. John says:

    I think this article is supposed to be on IE blog not here since this is the place to discuss mostly about features of Windows 8 and showing demo. If it's about IE 10 in Windows 8 then we should know some facts about it rather  than just bunch of data. By the way, I don't want to read the post that has no video demonstration on it.

  7. w3force says:

    I don't think this post is anything to do with windows 8.

  8. John says:

    Except the Metro styled IE, features from IE10 in desktop are the same as IE9 (I already checked developer preview). I'm wondering that if Microsoft and the team drop features creation process and start focusing on the performance and standard implementing only?

    I want more features for desktop IE, faster launch, smoother UX, easy to bookmark page, less annoying notification.

  9. pual says:

    @ John it's still a good read about the work is done into the IE and a final release of windows 8.

  10. Not even read half yet but VERY detailed.

    I noticed that IE10 on the Win 8 Dev Preview was extremely fast and fluid, it ran faster and smoother than any other browser i've used (and i often use many browsers each day – i'm a web developer) now i know one of the reasons why.

    And can someone remove Yousaf's post? It's pretty much spam.

  11. Kathy93 says:

    Be aware Windows 8, new Mac OSX is coming in the name of Mountain Lion!

  12. @John …. This should be amazingly obvious, but the Developer Preview was nowhere near complete. Neither is the IE10 version thats on it.

  13. ghost says:

    A lot of testing and in result Chrome is faster and firefox the best for extension. I won't never use IE again because with Adblock plus it is useless

  14. John says:

    @Beany3001

    I agree with you, I just want to urge them go further in order to remain in their throne. I used to use Microsoft-only products and that's things that I'm proud of. However, I start worrying about their development process since Chrome's is taking the share. I want it to avoid this potential threat. As the user's perspective, we need nothing more than just sneak peek of the next generation of IE. IE Blog has been long posting all about boring data, performance, and all. In here, I hope this is the time we should learn something about the upcoming features rather than knowing it's getting faster/smoother since this is inevitable trend and everyone knows it.

  15. Xero says:

    Sometimes if the system shuts unexpectedly, on next boot IE9 forgets all the "user preferences". These include:

    1. The :visited hyperlinks, the popular sites on about:Tabs page.. though the history (Ctrl+H) and that of onebar (Ctrl+E) remains intact

    2. Bing language preference.

    3. Forgets the autocomplete fields' value.

    and so on..

    Please prevent IE10 from this disaster!

  16. As a web developer I have to say stop collecting data and start focusing on W3C Recommendations!!! IE always been the worst browser to optimize!

  17. Real McCoy says:

    Run this most simple DOM manipulation test nontroppo.org/…/Hixie_DOM.html, which covers the areas, such as; append, prepend, index, insert and remove.

    Time taken to accomplish this test by various browsers is as follow:

    Apple Safari 5.0.5:

    172ms

    Google Chrome 12:

    319ms

    Mozilla Firefox 6.0.1:

    886ms

    Internet Explorer 9:

    44430ms

    Internet Explorer 10 Developer Preview (v: 10.0.8102.0):

    41185ms

    As it is affirmed, there is a significant difference in performances and IE is lagging behind vividly as compared to Safari & Chrome. Please address these crucial areas which are required by almost every next-generation interactive/rich web-application.

  18. ABM says:

    Sometimes, we receive links to download files as a plain-text email. Supposedly, the file type is PDF or some other which launches inside the bowser via corresponding add-on, and our intent is to download the file in that scenario, we would have to wait for the file to load completely then it can be saved.

    There must be number of other scenarios as well, which may require an option to manually enter the URL in download manager. Please provide us with a "Create Download" button in download manager, so the user can enlist the download by manually entering the URL to save a file, resource or the entire webpage within the download manager.

    Also, introduce the paste shortcut Ctrl+V in download manager for a single entry and Ctrl+Shift+V for batch entries.

    OR as for Windows 8, a collaborative "Download Hub" would be nice which may host all kinds of internet downloads segregated by the applications, like windows update, browsers, p2p yada yada..

  19. Kathy93 says:

    IE color tab is not really colorful. Because it has only a few boring colors…that somehow turns out to be useless to catch the eyes on. Please add more vivid colors to the tabs in order to make this feature more catchy and useful.

  20. One could argue that IE and Chrome are now about the same speed. But the problem isn’t there anymore. IE simply doesn’t offer the browsing experience other browsers (above all, Chrome) have got us used to, that is, a browsing experience: 1) made much more seamless thanks to an efficient and exhaustive autofill (name, passwords, forms, etc.); 2) made much more useful thanks to the easy installation of extensions relevant to our daily web experience.

  21. Steven88 says:

    "Be aware Windows 8, new Mac OS X is coming in the name of Mountain Lion!"

    By Kathy93

    It's true, incredible news, it will be released just one year after Lion.

    Apple Mountain Lion is Mac OS X 10.8.

    Maybe they wanted to make a 10.8 vs. 8 war,  LOL

    Number 8's war ;)))

    The best desktop OS is (and will be) Mountain Lion.

    Windows 8 will emphasize this thing.

    Sorry Microsoft… you have already lost. No news for the Windows desktop.

    Game over. Apple won.

  22. @Real McCoy: Nice!

    On my machine:

    Apple Safari 5.1.2 = 50ms

    Chrome 15 = 88ms

    Opera 11.61 = 121ms

    Chrome 17 = 143ms

    Chrome 16 = 157ms

    Chrome 12 = 160ms

    Firefox 10.0.1 = 339ms

    Firefox 4 = 343ms

    Firefox 3.6 = 418ms

    Internet Explorer 9 = 15039ms

    Internet Explorer 8 = 27254ms

    Internet Explorer 7 = 50423ms

    Internet Explorer 6 = 52135ms

    I was deleted PevPrev after 2 weeks, so I can't measure IE10.

    I'm going to install Consumer Preview just to see that they remove the start orb also. And to see that they not make start screen optional, and right after than I'm going to format that particion.

    And yup, I think the same about Mountain Lion. Apple smells blood!

  23. butilikethecookie says:

    Were these benchmarks for Metro ie or desktop ie. Metro ie is VERY slow and sludgeie.

  24. ..if win 8 will works just as good as this blog, than it's f*** up!

  25. Great effort, so what is this mean?

  26. it is Mountain Lion.and Apple who are worried about windows 8,reacting by announcing OS X Mountain Lion with a blog to boot.windows 8 is coming with full force, with the help of the oems be very afraid.

  27. Win8 Relevance says:

    Soooo… is this lab used to test Win8 performance, or was this just posted on the wrong blog?  Great detailed post, but doesn't seem to apply here, unless we are all missing something.  The browser and OS are two different things, no?

  28. jader3rd says:

    For me there are two places where IE performance interferes with using IE. First is when I'm starting IE and the disk is paging heavily (like when I just logged in). IE will be trying to load the home page and I'll select the address bar, start typing in the address (or query) I want to go to and BAM, IE will over write what I'm typing with the home page address. Next, a webpage begins to load and for one reason or another it's taking a long time to load. I want to use the search field on the page, and so I start typing in it. If the page hasn't fully loaded my typing is appended to the text that's in the search field. That's is so annoying, it should always clear the text in the field.

    Besides those two issues I'm generally satisfied with IE performance.

  29. @nana688: I don't think they are worried. I think they feel the opportunity to divide attention. They feel that this win7 with some ugly and almost useless tiles is not going to be an opponent. And for them a single update will be enought to do the job.

    They just feel the weakness.

  30. Sam says:

    Wihen will IE10 be avaliable for public beta?

  31. dhiren says:

    Since you're talking about IE performance, can someone very high up on the Windows chain of command find whoever on the IE team marked this connect bug:

    connect.microsoft.com/…/very-poor-gif-animation-performance

    as "Wont Fix", beat them thoroughly with a very heavy stick, and make sure that this is fixed in IE10?

    It's utterly shocking to see that the IE team feels that they can suck up 100% cpu core usage trying to render an animated gif, because someone decided that "animated gif usage is declining".

  32. Uminatsu says:

    @Real McCoy: Thanks for the link! It appears that IE fails miserably in this performance test.

    On my iPhone 4 (the old single-core model, not 4S) I can get 763ms.

    On my Quad-Core i7 desktop IE9 I get 11859ms.

    Pathetic!

  33. @Prettysure keep wishing and dreaming,number of windows 8 sold in the first 6-8 months will tell the story.

  34. To all those that reference benchmarks, please note that they covered benchmarks in their post. I have continually found that IE9 is faster in real world applications than Chrome is. I use Chrome as my main browsers, but every time I use IE9 for browsing pages that Chrome just won't render correctly I'm amazed at how well it performs. In Chrome I will routinely have performance problems when using many tabs. The browser will crash, the add-ins will crash. Changing tabs and scrolling will be painfully slow. Generally Chrome doesn't degrade very nicely under pressure. IE9 doesn't have this problem that I've observed.

    The reason I use Chrome is because it more faithfully renders standards and as a web developer I need that. Firefox has wonderful debugging tools, but it still can't let go of memory properly. Opera is very nice, but many sites use content that won't render correct there too (mostly plugin stuff).

    Note to IE developers: a page isn't a browser window. If I have 10 browsers open, each with 10+ pages, that would be 100 pages. In Chrome that appears as 10 browsers when I mouse over the Chrome icon on my Win7 task bar. Since I open a new browser for each task I'm on, it isn't a problem that I can't see the other 9+ pages in each browser. In IE that same scenario would yield 100+ pages and I can no longer see the thumbnails of the pages. That is another reason I don't use IE; while it performs well the UI doesn't scale well to high-usage scenarios.

  35. I have the same issues as @jader3rd on my netbook. My desktop is fast enough that I don't notice those problems.

  36. fr says:

    Good to see a post on IE performance as whilst it was a focus of IE9 there hasn't been much mention of further work on this in IE10.  

    One of my major performance concerns is after IE has been open for a long period with regular use (eg over a week).  I don't think it's that unusual to leave your browser open permanently and not shut down regularly, but IE9's performance does tend to significantly deteriorate until it becomes unreliable and has to be restarted.  Memory usage for some iexplore.exe processes grows quite high and often does not seem reduce again,  perhaps IE needs to have a similar effort to Firefox's memshrink project.  It also needs to be tested to reflect this style of usage, with very dynamic pages such as Gmail, Facebook, Twitter etc left open and interacted with, and with a large number of other tabs opened and closed to simulate normal and power user usage patterns.  Other browsers I use do seem to handle this kind of usage better for longer.

    Hope to see a new IE preview release on Windows 7 soon!

  37. ghost says:

    I hope you do a better UI for IE10 because the minimalist UI for IE9 is very bad. chrome and firefox have a great UI for simple and power users. Because IE is so buggy I believed you did a minimalist UI to prevent us for having a lot of tabs.

  38. abdo says:

    i really wat use IE10 but some times make me crazzy because still slow then others…i hope can fix this soon.

    http://www.full-windows8.com/

  39. Don't you think that it's so pathetic how Sinofsky sneaking out the way of the question about "will it be possible to use the classic start menu in 8, etc."

    This behavior is not worthy of a windows dev team manager. It's pretty shifty, dishonest and not the straight way.

    Sinofsky, you're not a "straight" man. And I don't mean your sexual orientation, I mean your mental one.

    After you wrote down that BS like 'if someone opens an application they will only use that one'

    and:

    “I think what is worth thinking about is that this provides a unified mechanism for today's system tray notifications or just app popups that just steal focus and take over the screen.” – you are not a serious man to me.

    Sinofsky’s got the cheek..

    I won't buy used car from you. Nor I won't buy your next OS.

  40. @c_barth: You're right – "a page isn't a browser window"

  41. oskingen says:

    I'm glad for your effort to get us a great browser. The road is long and you still have some works to do.

    I hope to see more posts about new features of IE10. Did you also check the following post:

    social.msdn.microsoft.com/…/463dbc6d-8bc5-4aca-9b70-b8084ed268cf

  42. Ryan says:

    I think where IE in particular falls behind in relatively performance is in the IE "The Application" area. Some of the functional things in particular can be time-consuming, quite frustrating, and even a serious drag on productivity:

    – When you click the Back button, why isn't the response from memory and nearly instantaneous? Why does IE seem to insist on re-rendering the page?

    – When you are filling out a form–particularly with large text fields–and you misclick or somehow end up navigating by mistake (e.g., Alt + left arrow), why doesn't IE remember/recover any of your text? Typically, all the text is lost, and you have to fill out everything again. If you lose a lot of text in a text field, this is particularly aggravating, since you have to reconstruct your writing from scratch. Some folks go so far as to compose in Notepad or Word and then cut and paste into IE to make sure they don't lost their work. Word, Excel, etc. had sophisticated recovery mechanisms for a long time now to make sure you never lose any work. Why doesn't IE?

  43. BARACK OBAMA says:

    Hey guys!

    I love to speak about your great job, so i will speak quickly.

    I need ONE Thing that i miss since WINDOWS VISTA it is:sorting headers in list view.

    iF anybody read this please comment.

  44. Steve Oz says:

    Performance is all well and good Microsoft, but if you can't provide an interface to match, it's pointless. Competitors are light years ahead with user defined 'speed dials' for the easy addition of their most frequently used websites etc Here's a suggestion: how about IE having LIVE TILE speed dial and speed dial groups of users favorite websites on the IE default page?

  45. Ben says:

    Internet Explorer: the number one browser used to download a better browser!

    I have to say that I hate IE with a passion. I will never use a Mac or any apple products and love windows 7 and quite a few ms products. However I think you should just drop IE. I spend hours and hours fixing IE bugs and trying to make a webpage look the same like chrome/firefox/safari etc. Why can't you follow proper standards? IE9 is still just as terrible at following standards as the versions before it.

    Honestly I don't even care about performance that much, why don't you get the standards correct before you worry about such things. It is the only concern I have about buying a windows phone that would force me to use IE. Also, you have to make it much less painless to upgrade. Why don't you follow Googles example and let the browser update in the background unless you specifically tell it not to? We still have people using IE6 and honestly it is slowing down web development.

    Death to IE!

    Cheers,

    Ben

  46. Adam says:

    I think, this post is very appropriate here because IE10 will be an important part of the Windows 8. The core of IE10 will run those Metro apps written in HTML5 and JS so reliability is a huge concern since this will decide wether you can use web technologies as well as Silverlight. (The performance could be pretty much the same for small projects, if ran locally.)

    IE9 is also very reliable in rendering nearly every page. (Also because everybody just has to optimize for it because of its marketshare.)

    Today, I opened the Apple.com page with the Mountain Lion announcement with the latest stable Google Chrome (17) and it rendered the font badly. IE9 performed just perfectly well here.

    I am looking forward to the IE10 using the bleeding edge underlying technologies as well as the further adoption of so called web standards. Some features like drag & drop, better support for codecs (WebM?) and many others would be nice.

    Variability with the UI could be better for power users as well. (Btw. There is lots of space, which is reserved just for the frame of the Aero style window, where the minimize, maximize and close buttons are… some healthy 20 px? Why don't you use that space for tabs or something else.)

    Last but not least: Thank you for an amazing post, which is very informative and link-rich so everybody can get the needed information to understand fully what the given paragraph is about. There has been lots of work on this article as well as on the testing. Thank you!

  47. JGodo says:

    Some news about including webGL in explorer in near future?

  48. Shortcut freak says:

    Missing shortcut:

    While holding the Ctrl button, if you press browser's back or forward button it should open the corresponding (previous or next) page in new tab. This behavior is present in Firefox, Chrome and Safari.

    Incomplete implementation:

    While holding the Ctrl button if you press 0 (zero) on qwerty keyboard, the zoomed page is reverted to normal. But if the 0 is pressed on numpad, the shortcut doesn't work. FF doesn't have this restriction with Numpad's 0 neither does IE reserve ctrl+num_0 shortcut for anything else.

    Why Microsoft??? Why should I open the ticket on connect for this stupid mistake by your dev then testing and user-interaction teams' ignorance and then wait for 2-4 months so you close my ticket as Won't Fix with a hope to get it in next release !!!! Com'on address this little hiccup in the name of evolution..

  49. Fordp says:

    Performance is but one aspect. It's not important at all when the user experience is so much worse than other browsers like Chrome/Firefox. IE could be the fastest browser ever and most users would still prefer an easier to use, more feature rich and nicer browser than saving a few ms. The only reason people use IE is because its part of Windows.

  50. @Real McCoy says:

    This is because IE encourage you to mitigate changings on "live DOM". You can create an on-the-fly DOM object in JS, then attach components with it and finally publish the DOM to the corresponding location. Do not commit every change on the DOM. Taking care of this would make your interactive application faster on cross-browser! Also try to learn the DOM fragmentation concepts with less interference with the Live DOM.

  51. "Real world browser performance matters"

    1. Why suddenly care now that the horse has bolted and you have a lot of competition?

    2. Why doesn't IE have an option to *block all ads* instead of forcing users to constantly mess about with hosts files?

  52. Very interesting. Thanks for the post.

  53. i like the speed of IE i currently use IE9 which is good but it annoys me that i cant have the same layout as IE8 because im not a fan of a shrunken browser like chrome and while working with windows 8 and IE10 i realy dont like the browser but thats not to do with the way it looks its because a lot of sites i visit have broken layouts and pages look messy.  i have noticed on some sites when scrolling down a white line appears in the centre of the page and everything is distorted and i know we should use the compatibility button but that has no effect whatsoever.  i dont fully understand why each new version of IE breaks webpages and yet new versions of firefox and chrome continue to display them perfectly and ive had no choice but to use chrome while in windows 8 just so i can browse the web without problems.  faster browsing is good but speed is no use if you cant spin the web

  54. @Ben I'd totally appreciate it if you would use the contact email above and send the URLs to pages you worked on that were troubling for IE9 along with a pointer to the constructs, or even better just paste some code in that gave you difficulty.

  55. Anon says:

    And yet Chrome is better still. And it became as such in only 3 years. Why is so? Why is it that IE with its vast testing imfrastructure is still behind. Why is IE's security model still worse than Chrome's? Why can't IE update popular add-ons and plug-ins automatically like Chrome? Why does every ActiveX and plug-in developer for IE have to create their own little updator service? Why is writing add-ons anyway so hard compared to Chrome even after 15 years? And why is even with a gigantic testing lab, IE slower than Crhome? Chrome even pre-renders a page while you are typing its address in the address bar.

  56. I'm happy to see a focus on performance in IE10, especially since it might be running on some low-end ARM platforms.  However, it won't matter much if the browser lacks the basic capabilities of some of its competitors–namely extensions and plugins.  

    On my Developer Preview tablet, I tried using Metro based IE10 for as long as I could, but eventually the lack of Adblock Plus was too much for me.  At least I didn't have to install Flashblock.  🙂  So, I'm typing this from Firefox on my Win8 tablet, in desktop mode.  

    Oh, and not allowing plugins basically makes Metro IE useless for me at work.  I can't even log my time and attendance (Java app), much less use one of the Silverlight based internal applications we spent millions of dollars developing over the years.

    I'm curious to see if the recently announced Metro version of Firefox will find some way of enabling extensions in the WinRT sandbox; I sure hope so, as that will make Win8 Metro a lot more attractive.  That is assuming, of course, that Microsoft doesn't pull and Apple and ban competing browsers from the app store.

  57. RE66 says:

    about "capture everything from high-level CPU and GPU activity, to low-level information such as cache efficiency, networking statistics, memory usage patterns"

    You forgot about a very import issue and it is the mount of Local Disk I/O usage. From my experience of using both FF and IE the local IO usage of IE is about  10 time higher in IE compared to FF.

    Every new Tab open cause an intensive IO load ( I assume that instead of moving cache data from one process to the new one the new one have to read all again )

    This is one of two  main reasons I use FF as my default browser.

  58. LOL @ trying to get attention back on Windows 8 after the Mountain Lion release, I love this competition ^_^

  59. fengchun says:

    Not related to IE performance, but related to IE color management. IE 9 only support reading c olor profile in the image but ignore the output device (monitor) profile. Will this be corrected in IE 10?

  60. John Mitas says:

    What a joke … IE is dead to me… I'm forever in the Chrome world… and if not chrome ill use FireFox..

    Kill IE … kill it off!!!

  61. boomie says:

    It is not about "browser performance", it is about "browser responsiveness"!

    WinPhone team got that already and here we have super responsive winphone 7 that is joy to use, same is with metro interface on slates.

    I've been having some bad experience with IE9 not because it is slow, but because it was not tested by actual power users.

    Firefox taking 5 seconds to render a page AND remaining responsive to my actions perceived as much better experience than IE rendering page for 4 seconds and being hung during that time.

    My theory is some of the IE problems are happening because it is too polite in memory managing. When FF refuses to let memory go under heavy load, IE frees it and takes hit to response time when it tries to load memory from swap.

    Also, the big thing to me are lack of customization, I have not found a way to disable showing "close" buttons on inactive tabs which made me close the tab instead of switching to it too many times and other stuff like awkward tab saving using "reopen last" menu.

  62. Will Internet Explorer 10 and color management? I'm using a wide-gamut monitor and color management is a must.

    Another question: Will Windows 8 support 10bit color ?

  63. I predict windows 8 sales will be will more than ios/ Mountain Lion /ipad  and iphone  all    combined in 6-8 months

  64. Frank says:

    As a end user, I just care about my experience when I using IE browser. Actually, IE is not the slowest browser but it's also not the fastest browser in the world from my experience.

    I tried a lot of browsers in the past few years. Till now,I think chrome, firefox and Avant browser are the top three when it comes to the speed of a browser.

    IE, I think could be the most stable browser in my opinion. It has less problems when I using it, and most of the problems can be fixed by reseting IE settings.

  65. xpclient says:

    The efforts Microsoft takes to engineer IE for great performance are appreciated, had no idea about the scale of IE's Performance Lab. But the percentage of Windows customers who benefit from performance improvements in Internet Explorer is far less because Microsoft has started cutting off platform support for older OSes still in widespread use just because they are older. The performance improvements only reach the ones who use the latest Windows release. Whereas other browsers do *the right thing* in not abandoning their customers which is far more important. Just keep the web platform separate from the OS while still providing web integration in the OS. MS ties their browser to OS upgrades – the single biggest issue for IE's continually falling market share. How old an OS does not matter **more than how many people are using it**. Mozilla, Google and Apple have blasted MS for being terribly irresponsible in dropping XP support for IE9 and Vista support for IE10 but MS doesn't care. I will always continue to use IE because it offers the best usability and balanced feature set catering to end users and web developers.

    For me, IE9 was a huge improvement in tab loading speed and page loading/rendering speed. But the quantum leaps in performance reach a smaller set of users because of lack of XP support.

    Coming back to actual feedback on IE performance, three examples of the pages that give me the biggest, more frequently, consistently reproducible hangs and crashes in IE9 on Windows 7 x64 in my own usage are:

    1.  History pages of any article on Wikipedia which show 500 history edits. For example on this page: en.wikipedia.org/…/index.php , try to compare any two revisions of the article in IE9 standards mode and notice how sluggish clicking on the radio buttons and scrolling up/down is.

    2.  Microsoft's own forums like Windows Developer Preview forum pages (social.microsoft.com). IE often hangs on very long posts with dozens of replies and the tabs crash, *other browsers don't*.

    3.  Many RSS feeds with long-posts and complex JavaScript running. Like the //build/ session RSS feeds on Channel9.

    Another big issue even for IE9/IE10 is it doesn't release memory from tabs as efficiently. Often when I have many tabs loaded (15 on average), then I sleep or hibernate the machine and resume the next day (even with lots of memory, less HDD paging and no IE addons), I have to close IE and re-open the session just to reduce the memory usage of IE. For IE vNext, MS should focus on freeing per-tab memory more efficiently and of course JavaScript rendering speed.

  66. Sebastiaan Mindreau says:

    Great to hear you're working on improving IE for Windows 8.

    But instead of just aiming for speed, why don't you aim for less memory leaks?  Since you went to a per-tab process, IE is the main cause of memory swapping in my environment due to too high memory usage.  This should be addressed!

  67. JGod says:

    Microsoft should drop IE and build a new browser from scratch based in webkit. Just my opinion. IE has already lost the browser battle. And it deserved it.

  68. Extensibility in F12 Developer Tools just like Firebug says:

    1. While inspecting with the F12 developer tools, the asynchronous page update doesn't immediately update the code in devtools’ HTML tab. We need to close and reopen devtools to seek the updated code.

    2. When we click 'edit' in the HTML tab, it should narrow down the code to that of the selected tag and it's innerHTML (only) in edit mode.

    3. Implement autocomplete for Style while creating or editing rules in both CSS and HTML tabs.

    4. Preview of images onmouseover in tags such as <img src=….>, <input type=image….> or image embedded in stylesheet.

    5. Implement an easy way to add a new Style rule in HTML tab rather than switching to the CSS tab. For example; pressing <enter> for the first time should highlight-to-edit the name of the selected rule, next <enter> should switch the focus to that rule’s value then next rule's name and its value and since so forth, until it reaches the end of "that selector" (only). After that, the next <enter> should cause the creation of new rule's name-value pair under that selector. So this way, if the users have to create new rule promptly, rather than going to CSS tab and find the selector then right click to select AddRule, within the (right-side: Styles of) HTML tab, they select the last value of the desired selector and hit enter to create a new rule.

  69. Please make IE 10 fast and fluid.  IE 9 on my machine is sluggish and hangs frequently.

    Keep the TABS bar in separate line.

    Ctrl + 0 from numpad not resetting zoom to 100 % is irritating

    Add functionality in download manager. If i am download a heavy file and close IE in the middle of download , I should be able to resume it after restarting IE or the computer.

  70. Also, while saving a web page, I can not perform other tasks, like opening a new tab and viewing the content of the web page. Waiting till the download completes is frustrating. Please fix this issue

  71. alvatrus says:

    @JGod

    Which webkit version? There are a lot out there, and they all render subtly differently.

  72. Will says:

    the best test for IE is….

    run facebook home page and go to the bottom, click load more, go to bottom load more, load more, etc

    run twitter and go to the bottom, load and load and load tweets

    *it's not that I click hundred of times 'load more' I just click it kinda twice

    you'll see that IE9 can't handle facebook or twitter …. both using a computer with or without dedicated graphic card, IE9 becomes superslow and graphics start to dissapear or start 'like flashing' in the web page.

    *the same happens in several websites, also flash is super slow on ie9

    *try the same in chrome and you'll see that it always run fast …. aka chrome is better than ie9

    *and I am right, tHat's why I always see Microsoft employees using chrome in Microsoft events

  73. Will says:

    oh ur moderating, so my recent comment won't appear, well at least have it saved for you to make better IE…. I would love to use IE as my default browser, I would love that it can handle tabs as fast as chrome does, I would love it to render web pages correctly, I would love it to work fast 😉

  74. Terry A. Davis says:

    Wow!  Lots of work.  God is just.  Do you think Linux is existentially cursed until they offer a warranty?  It seems fair to reward taking responsibility and risk by offering a warranty.

  75. Real McCoy says:

    @@Real McCoy, check out the replies from IE team for the connect feedback connect.microsoft.com/…/a-dom-manipulation-test-ie-performance

    They simply closed the issue as if it doesn't matter. The question is; if other browsers can swiftly update the DOM objects, why can't IE's Trident do so? Don't tell me about any security issues or "fragmentation being a better approach"…  Issues reported on connect are simply ignored by some sealed heads.. they don't even bother to discuss it with the real developers of IE who apparently don't even know if anything like "connect" exist! Hence connect is just an eye-wash.. Microsoft needs some interactive platform like conceptshare.com to gather the feedback in visual/prototyping format from their clientele rather wasting our time on orphaned website like connect.microsoft.com !

  76. erix says:

    @Real McCoy:

    on my machine IE9 runs in 150 ms.

  77. erix says:

    A second call to the test page gave 122 ms ==>

    Total elapsed time: 122ms

    Breakdown (fraction shows time relative to append time):

     Append:  1.00; 30ms

     Prepend: 1.07; 32ms

     Index:   0.53; 16ms

     Insert:  1.10; 33ms

     Remove:  0.37; 11ms

  78. Per says:

    @Real McCoy:

    CHROME:

    Total elapsed time: 156ms

    Breakdown (fraction shows time relative to append time):

     Append:  1.00; 47ms

     Prepend: 0.66; 31ms

     Index:   0.66; 31ms

     Insert:  0.66; 31ms

     Remove:  0.34; 16ms

    IE 10.0.8102.0:

    Total elapsed time: 20642ms

    Breakdown (fraction shows time relative to append time):

     Append:  1.00; 148ms

     Prepend: 1.01; 150ms

     Index:   102.51; 15172ms

     Insert:  11.11; 1644ms

     Remove:  23.84; 3528ms

  79. Per says:

    How about a  Media Center or a multimedia blogpost?

  80. erix says:

    @real: in your earlier posts you mentioned values for IE 9 like 44430ms.

    I just wanted to show what real IE9 values are.

    I did not test it yet on IE10, but for sure Win8 / IE9 is a preview now, nothing like tuned GA release.

  81. erix says:

    @Real McCoy: my values for IE10 / Win8 are similar to what you see (around 32000 ms). But I would certainly not judge a product on a very early preview release. The IE9 value seems to better represent how IE can perform on such a test.

  82. Stras says:

    Steven88 and By Kathy93

    The prophet is dead! you are a phone and a tablet today.

    The world is a better place without the insults and exaggerations of Steve Jobs.

  83. Alex Acosta says:

    You need to dump the Activex plugins and support Mozilla plugins (like Chrome/Safari do)

    See the feedback from the author of my favorite plugin (Adblock Plus)

    adblockplus.org/…/adblock-for-internet-explorer-is-not-an-option

  84. AgentSmith says:

    Nice work! Let me know when you release a version for Linux. I'll be happy to use it.

  85. Eagleizer says:

    You cannot judge a  browser just  by counting milliseconds of  computing.

    I use both Crome and IE8. The inteface makes me work much faster on IE8.

    (With Google toolbar)

    For the same reason I do not use IE9.  

    I want a fully customizable interface, like in the good old XP days!!!

    … then you can give computing power…

  86. pmbAustin says:

    I just want the IE10/Desktop team to bring back the separate search textbox/control.  I *loathe* the combined "one bar".  It's the one thing I CONSISTENTLY miss from IE8.  I actually just enabled the favorites bar and stuck both Google and Bing icons on there, so I can do searches the way I want, and not pollute the stupid address bar drop down with search queries, but I still miss the easy ability to do the same search in multiple engines with ONE CLICK.  At least make it an option so I can enable it.  I want tabs on their own row, and the address bar and search box on the same line.

    I also think that allowing more and better customization of the UI (where the home button, favorites button, tools button, etc go) would remove many objections others have.

    But mostly?  We need more frequent updates.  We need 10.0 an then four months later 10.1 and then four months later 10.2, etc.  We need a steady drum-beat of performance improvements, and feature/HTML5 enhancements… done in a way that won't freak businesses out (i.e. new versions don't "break" but only enhance things).  Without this?  IE will continue to lose market share.  It doesn't matter if IE10 is the fastest browser out of the gate, if it becomes second place in four months as the others catch up, then third place in eight months, and a year later, it's dead last.  The competition is not standing still.

    I have no doubts that IE10 will be the best IE yet.  If you can make the UI more customizable, bring back the dedicated search box for those of us that use search INTENSELY and prefer to operate that way, and roll out HTML5 enhancements and support and updates more timely than once every 12-18 months… and you'll have a serious winner here.

  87. Real McCoy says:

    @Erix, are you sure about the score? In the latest release of IE9 (9.0.5) the results on my Window 7 machine (dell laptop; corei3, 4gb ram) are:

    Total elapsed time: 22369ms

    Breakdown (fraction shows time relative to append time):

     Append:  1.00; 375ms

     Prepend: 1.10; 413ms

     Index:   40.52; 15194ms

     Insert:  5.19; 1948ms

     Remove:  11.84; 4439ms

  88. John says:

    @Steven Sinofsky

    Sorry for being off topic but I would like to know if the new redesigned logo of Windows 8 make it way to Consumer Preview. I want to test how it feel like in the O.S since many people has criticized this design as much as the metro UI.

    Thank you

  89. erix says:

    @real: I'll recheck tomorrow.

  90. Stefano says:

    There's a new Windows logo! It's official.

    windowsteamblog.com/…/redesigning-the-windows-logo.aspx

    So the pictures I saw in the past few days were real…

    😀

    It looks great.

  91. tuv says:

    wow all that work. And yet open source browser engines such as webkit and firefox are beating you. Microsoft has many more developers than apple, yet apple is soon going to be twice as big as microsoft.  Isn't that the very definition of poor management!!!!!!!!!!!!!!!!!!!!!!!!!!!

  92. reDim says:

    It's amazing how Windows manages to lose credibility every day.

    Loses more credibility in the little things that one known computer company that sells overpriced hardware pieces (check out the ram, holy god)

    Can't be that a good OS has a bad performance browser that has no eyes for web standards.

    I'm sorry about that.

  93. JGodo says:

    The new windows 8 oficial logo is a very horrible logo from all logo design standars ( I´m graphic designer myself ). As horrible as metro in desktop.

    Nothing more to say about it.

  94. @Steven Sinofsky:

    Off topic: I saw the new Windows logo, and I actually like it, except for one thing. Why is the entire logo colored in simply a flat sky blue? Personally, I think that the words "Windows 8" should be in royal blue and the window pane design should retain the traditional red, green, blue, and yellow design.

  95. what's the xperf command you use to trace issues?

    And have you fixed the terible slow XML file loading from Internet Explorer 9?

  96. Ryan says:

    As a perf engineer I just wanted to say: Cool!

    To me, the SQL reporting setup is by far the most interesting part. Definitely giving me some ideas… 🙂

    Thanks for sharing!

  97. A. Neto says:

    I like the look of Internet Explorer 9, it's the prettiest of all the browsers, but I must admit, it suck a lot. It crashes in a lot of sites, specially YouTube. I read it is related to some plugin (or whatever), but I should not be worried about which plugin I should install in my browser, I just want it to work nice and open all the sites and don't crash, but IE can't do that for me. Firefox have some nice finishing too, a lot of spetacular addons, but is heavier (at least it appears to be), so Chrome is the only one that rest. I don't really like it, it's ugly, but it work, and fast, and don't crash. Please, make IE 10 like that, but looking like the IE9.

  98. Fr@nk$inatr@ says:

    Cool story bro!  In what chapter do you talk about IE10?

  99. Antony Jones says:

    I'm sorry, but who cares how fast a browser is if it can't render content properly?

    Try making a browser which adheres properly to standards, supports a full reference implementation of real CSS3 and runs Javascript in a predictable and conventional way, passes the ACID test suites, doesn't need to lie about itself in comparison charts, and hey – just for fun, runs on multiple Operating Systems, and then I'll care about how /fast/ it is.

    An interesting post nonetheless, but the subject matter is largely irrelevant.

  100. etacarinae says:

    @dhiren:

    "Animated GIF use overall is not particularly high and is declining. Moreover, sites with many animated GIFs are rare."    

    Haha! I guess they've never been exposed to tumblr, forums, livejournal, imgur and innumerable other portals that the youth of today all inhabit, where GIF usage is undergoing a huge resurgence.

    @Steven Sinofsky:

    I'd echo @pmbAustin and other's request for the return of a separate search box, the "tab list" (not to be confused with the useless quick-tabs that was chosen to be retained), and the general ability to move and customise the interface (IE7, IE8) which has now been depreciated. Why? It's fine to hide things, however at least give us the option to bring them back if we so desire. Firefox and Opera do.

    These requests have never been acknowledged by any of your IE team, Steven. I'd like to see if you will acknowledge the "dumbed-down by design" that seems to now be the norm at MSFT.

  101. Bob says:

    So IE is being tested on an imaginary Internet, no wonder IE sucks so. I have a feeling that IE will never catch up to Chrome.

    I just don't understand, all these testing and available resources and you guys still can't come up with the decent browser.  

  102. Nitz Walsh says:

    I was pretty disappointed in the preview release that IE, whether Metro or Desktop, still does not smoothly re-size pages like in iOS and now OSX Lion.  That is by rendering the page into an off-screen buffer when using dynamic resizing to get great response, then when resizing is stopped re-render the page in full res.  

    The responsiveness of this approach is well known to anyone who's used iOS/OSX.  After using that, browsers that actually dynamically resize every element in chunk-chunk-chunk steps just feel archaic.  Scrolling responsiveness as well is fantastic – this really needs to not just come to IE10, but to WIndows 8 as a whole.

    Smooth scrolling and smooth resizing – everywhere.  That will have the biggest user-impacting effect, far more than winning a Javascript benchmark.

    And please, give up the "tabs on address bar" as the default layout.  It just doesn't work, people need to see beyond the first 10 characters of the URL and it limits tabs as well.  It seemed like a desperate ploy to be "different" and say "See, we take up less vertical space than Chrome!", whilst chucking usability out the window.

  103. kisscul says:

    I heard that the HTML5 WebGL technology will run ultra fast and smooth on IE10. That's really cool!

    I'm sure you will kick the ass out of Firefox and Chrome.

    Keep up the good work!

  104. Web-crawler says:

    @Bob

    You obviously don't know much about testing and coding for IE !!

    I agree that IE has to behave better with web pages and with users than it does now, but it doesn't mean the testing procedure is the issue !

  105. Satish says:

    Two IE Browser causing lot of problems. When we open a link it open in Other IE10, and we normally surf in other IE. we have to change screen everytime we want to see other IE page. There is no closing button for default IE which open links.

    Also Games/application which get opened once, could find anything to close them, they always keep pending in background and may affect system performance.

  106. DLT says:

    a bit off topic, but this metro comment made me laugh:

    http://i.imgur.com/vd2WA.jpg

  107. Paul Coddington says:

    As an interesting aside on the issue of full-screen presentation in Metro IE, I've been playing around while reading this article.  This web site is comfortably readable in a window taking up about 1/2 to 2/3 of a 1280×1024 display.  If you maximise the window, the site becomes extremely uncomfortable and difficult to read, as the text is now wrapped to the full screen.  On a wide-screen display, full screen mode would be unusable.  This means that many web sites, including this one, require major reworking to operate with a full-screen only mode browsing experience.  I guess zooming would address the issue, but then that is not full screen.  Perhaps 2 tabs side-by-side?  It will be interesting to see if the user can remain in control of the text-width in the new scheme of things.  Should have checked this out before dropping my copy of the developer preview, but at least the consumer preview is nearly here (yay!).

  108. mojtaba alemi says:

    1.Why there is no AMD System in your test list?

    2.Why IE9 even have problems with big sites like gsmarena.com (you can not open posts in new tab) and many other small websites ?

  109. @Mojtaba alemi, it's not called the Wintel monopoly for no reason.

  110. ThridEyeBlind fanboy says:

    @IE team, please push some real development in Trident DOM parser. I'm surprised to see over 100 comments for IE.. which means people still expecting from you and there is still a slight hope for IE to revive its position!!!!

    Mozilla's Kraken: krakenbenchmark.mozilla.org/index.html

    Google V8: http://code.google.com/p/v8/

    DOM manipulation (posted by RealMccoy): nontroppo.org/…/Hixie_DOM.html

    If you able to make IE gain outstanding result in these three tests in comparison with Safari, FF and Chrome… You will make everybody excited! otherwise you are just dogging the "real" challenge… and wasting time in performance testing… IE9 won Apple's sunspider contest, after that there is no non-MS vendor test which qualify MSIE as a super hero 🙂

    I dare you … get distinction in those three conformance tests FTW and I will never use conspiring-Chrome or any non-IE browser…  else as always I'll b riding on the FF-hound

  111. ReMark says:

    Cmon guys, IE10 on Win8 should be a thunderbolt, take your times and listen to users.

    If you wanna ship a IE10, ship it in the best way.

  112. Andre says:

    @ IETeam.

    A lot of effort for little result.

    Kraken Mozilla: krakenbenchmark.mozilla.org / index.html

    Google V8: http://code.google.com/p/v8/

    We are critical because we still have some faith in IE, but it might end up.

  113. James Preston says:

    Id like to bring up the point about no AMD systems being in the farm as well?

    Surely it would be intresting to see how IE works on the new fusion APUs.

  114. @James and others worrying about AMD APU's says:

    Search on Bing or Google … you will find something like..

    windows7themes.net/lots-of-windows-8-devices-and-tablets-with-amd-apus.html

    so yes! Windows 8 will consume CPU, GPU, APU, ARM all kinds of systems!!!

  115. John says:

    I have not figured out how to use IE 9 for anything other than the simplest of tasks.. If I want to do anything serious it is IE8, or the current versions of Firefox or Chrome for me. Sorry Microsoft but you have lost the plot.    

  116. Stefano says:

    I suggest the use of a password even if a user hasn't created one for access, for the RESET and REFRESH utilities. It's too easy to delete a user's PC contents in Windows 8. If I can access a user's desktop I can delete all of its files etc… Too easily.

  117. @John says:

    Can you specifically mention what kind of "task" you are unable to carry out in IE9 that you do in crippled chrome?

  118. Andy says:

    Some people should start to live a real life and stop trolling on internet. Trust me… you're not "cool" as you could think.

  119. Todd Garrison says:

    From the WOA post:

    "During Windows 8, we worked to define a new standard to describe the configuration of the system with tables, so software can simply read the table and configure the system."

    I wonder whether this work ties in with the rumors of the Win8 kernel and surrounding stack coming to Windows Phone?  Some of the Windows Phone rumors are that additional resolutions, processors etc … will be supported.  Right now Windows Phone runs on a number of Snapdragon processors that are essentially ARM-like.

    Putting the two pieces together it seems like a very reasonable technical operation to rip the kernel and related stacks out from under Windows Phone, and replace them with the "MinWin" piece of WOA with appropriate configuration tables for the Windows Phone chassis specification(s).  And going forward, based on everything described in this post, the new WOA core would provide tremendous flexibility for hardware manufacturers.  Windows Phones could be based on any of the WOA compatible SoC chips from the 3 ARM vendors, and likely even from Intel, and phones could be configurable to the extent allowed by the configuration tables, and the new chassis configurations that will no doubt surface.

    It certainly seems like the future is a stack of base services that runs everywhere, and supports WinRT: on Windows Phone, on WOA, on traditional x86/64 PCs, on XBOX, on …

    On top of those WinRT runtime environments are a couple of tailored software environments:

    Windows Phone for pocket sized devices.

    WOA for thin, light, long life tablets and laptop-lets.

    Regular "Windows 8" for desktop PCs and laptops.

    XBOX for big screen media and game playing.

    The WinRT platform will mean that application logic can be written once and deployed to any of these platforms.  The challenge for the developer is to choose which screens to target, and to write appropriate interfaces for those screens.  This represents the "app by app" version of the problem that Microsoft has tackled themselves by realizing that phones, PCs/Tablets, and large screens like XBOX each need a tailored UI that is "screen specific".  And each will need apps with Metro UIs that play to their individual strengths.

    Windows everywhere.  An O/S UI layer tailored for the device you are using, and a foundational platform (WinRT) that allows application developers to fully reuse code and focus on user experience through tailored UIs.

    I'm excited about the future.

  120. Fun thing says:

    The evolution of speed (Microsoft advertising on their site):

    i.microsoft.com/…/Hotmail_0213_v2_530x320_lt.jpg

    Looks like a…. Mountain Lion (OS X 10.8 by Apple).

    🙂

  121. @Fun thing

    That's a cheetah. There was a verison of OS X called Cheetah, but it is long gone.

  122. Fun thing says:

    @WindowsVista567

    They're brothers, same family 😉

    I think that Macs will win the desktop "battle". Windows 8 will compete against the iPad. If prices will be right, Windows 8 could win in the tablet market, a fully portable operating system is better than just a system like iOS. Apple's error in my opinion is using iOS in iPads, (Mac) OS X would be perfect but maybe they wouldn't sell any more Macs (mini, iMac, Macbook pro, etc…). It's all about marketing strategies….

    I have two suggestions for Microsoft:

    1) Lower the price of Windows (8) Home Premium Edition (no more than $49 for the upgrade version)

    2) Don't remove Windows 7 from stores as soon as Windows 8 is released, don't do it, please, leave choice to your desktop users who very likely won't like the desktop (the always switching from Metro back and forth thing) of Windows 8.

  123. Todd Garrison says:

    I was just fooling around on the Developer Preview, swiped in the charms bar and had a "momentary thought" after accidentally touching the devices charm.  Microsoft already HAS something called devices.  It shows up in Windows Live online when you connect computers to your mesh, or get a Windows Phone.

    The only thing that shows up in the devices hub right now is the "connect to projector" widget.  I wonder whether the devices hub will see a most amazing improvement when the Consumer Preview is released.  If the devices charm provides access to the devices that show in my windows live profile it will be wicked.

    Imagine misplacing your phone, picking up your tablet (which is logged on using your Windows Live ID), swiping out the charms bar, tapping the devices icon, and having direct access to the "ring my phone" option that is currently buried under 6 feet of snow in Windows Live.

    This kind of feature would bring about a totally amazing connected reality.  Need to get to some files on another device in your universe? Just Swipe Charms > Devices > Other PC > Shared Folders > Done.

  124. Irfanfare says:

    I love IE10, almost subconsciously I pick up IE from 2 other browsers I have downloaded, without apparently knowing the reason for my choice. Now I know why I do.

  125. winstar says:

    in some times windows 7 SP1 icons become like to these :

    http://www.uploadkon.ir/…/dc49d320b39f992c1f804c1ecb8eaa88.png

  126. @winstar says:

    I know it occurs every time the desktop starts up…There are many icons for it the load perhaps? I hope Windows Team will look onto it. At least the Mac doesn't have this issue!!

  127. Jp says:

    @winstar

    http://www.uploadkon.ir/…/dc49d320b39f992c1f804c1ecb8eaa88.png

    I know it occurs every time the desktop starts up…There are many icons for it the load perhaps? I hope Windows Team will look onto it. At least the Mac doesn't have this issue!!

  128. Stefano says:

    @winstar

    In Windows 7 sometimes a program that has an icon in the taskbar, when launched creates another icon (duplicated icon) of itself…

    I've seen it doing it with Firefox and iTunes. For Firefox the solution is deleting and recreating its icon on the taskbar. I still have to try it with the iTunes icon.

  129. Stefano says:

    Solution for iTunes:

    1) Open iTunes

    2) Right click on its icon in the taskbar and select "add to taskbar"

    Only in this way it doesn't create a duplicate 😉

  130. John Kowalsky says:

    Hey guys! I hate IE because is SLOW like a pig, but i LOVE OFFICE 2007 AND WINXP WHY?

    BECAUSE IS NOT FAST/ NOT STABLE BUT  

    CUSTOMIZABLE!!!!!!!!!!!!!SO: as we see there is no progress to CHROME or Firefox!!!!!!!!!

    I need ONE Thing that i miss since WINDOWS VISTA it is:sorting headers in list view.

    iF anybody read this please comment.

  131. Lower prices says:

    You MUST lower the cost of Windows. Absolutely. Don't continue to be greedy as in the past. Times are changed. Or else Windows 8 will be a total failure. The complete commercial game Alan Wake on Steam is sold for just $29.99! In just two days the game makers have gotten back what they spent to make the game… It's time to REIMAGINING the PRICES!

  132. The moment MS dropped Silverlight, they killed IE along with it…. way to go.

  133. Huki says:

    There should be a way to circumvent IE's compatibility mode. Any page without a proper DOCTYPE or X-UA-Compatible tag is forced to render in a legacy document mode. Not to mention the blacklist database which can only be turned off from the hidden menu bar. It wouldn't hurt to add one option in the advanced setting to force the use of IE-Edge document mode (considering there is even a setting to turn off automatic crash recovery).

  134. Mayur Prayag says:

    Current IE Team is not brilliant ,talented.They can't think from user's view.Steven ,Please Remove IE development team ,appoint new team like Windows 7 team or give project to Google ,they will develop IE like Chrome and later you tell world that we(Microsoft) developed "the IE"…..but finally we will get better IE ….Please reply me or guaranty me that future IE will be a benchmark …..the best…..better than Chrome……

    Below are IE's failure Reasons:

    1) IE's UI feels very heavy when first opened.

    2) IE is not simple means back button and forward button is not light.

    3) IE is slow

    4) IE takes more time when loading Facebook and Twitter…..!!!!!!

    5) Give IE's development to  "Google" and they will tell you how to develop a web browser.

    6) Please,copy the chrome UI ,responsiveness , simpleness…..then and then only IE will be great         otherwise whatever mini Internet,millions test you do ,it is of no use……

    7) chrome feels very light and IE feels like ton of load on our head..

    8) when we open any site IE ask for default browser in a very sluggish manner…

    9)  IE smart screen filter very slow takes much time

    10) Remove status bar from IE

    11) Take some development lessons from Google Team so that you can develop OS like Android and browser like Chrome….

    12) Within 3 years Google showed your position in market(Browser)…

    13) Microsoft is not very serious about IE…

    14) Current IE 9 is like Windows Vista ,make some strict steps and give IE 10 like Windows 7,8 performance …..

    15) Make proper Acid 3 tests,because when user sees Acid 3 tests after release on various sites,they give very less marks to Internet Explorer….

    16) Make UI clean,again I'm giving warning otherwise within 5 years you will lost completely..

    17) When we Right Click on empty space,a lot of items appear…..Remove many items(Translation etc) from right click

    18) Keep less items for Right click …..we want that

    19)Current IE Team is not brilliant ,talented.

    20)They can't think from user's view

                                                              Most Importantly, All will agree that,        "  Last thing ,we all people wanted that IE must beat chrome…….when we double click chrome icon on desktop ,it opens smoothly and feel softs,smoother …..whereas IE feels like scratches on our Screen,skin etc……!!!!!!!!!!!!!!!!.."

    I hope you will take a lesson from above .I'm a end user.

  135. Mayur Prayag ( India ) says:

    Current IE Team is not brilliant ,talented.They can't think from user's view.Steven ,Please Remove IE development team ,appoint new team like Windows 7 team or give project to Google ,they will develop IE like Chrome and later you tell world that we(Microsoft) developed "the IE"…..but finally we will get better IE ….Please reply me or guaranty me that future IE will be a benchmark …..the best…..better than Chrome……

    Below are IE's failure Reasons:

    1) IE's UI feels very heavy when first opened.

    2) IE is not simple means back button and forward button is not light.

    3) IE is slow

    4) IE takes more time when loading Facebook and Twitter…..!!!!!!

    5) Give IE's development to  "Google" and they will tell you how to develop a web browser.

    6) Please,copy the chrome UI ,responsiveness , simpleness…..then and then only IE will be great         otherwise whatever mini Internet,millions test you do ,it is of no use……

    7) chrome feels very light and IE feels like ton of load on our head..

    8) when we open any site IE ask for default browser in a very sluggish manner…

    9)  IE smart screen filter very slow takes much time

    10) Remove status bar from IE

    11) Take some development lessons from Google Team so that you can develop OS like Android and browser like Chrome….

    12) Within 3 years Google showed your position in market(Browser)…

    13) Microsoft is not very serious about IE…

    14) Current IE 9 is like Windows Vista ,make some strict steps and give IE 10 like Windows 7,8 performance …..

    15) Make proper Acid 3 tests,because when user sees Acid 3 tests after release on various sites,they give very less marks to Internet Explorer….

    16) Make UI clean,again I'm giving warning otherwise within 5 years you will lost completely..

    17) When we Right Click on empty space,a lot of items appear…..Remove many items(Translation etc) from right click

    18) Keep less items for Right click …..we want that

    19)Current IE Team is not brilliant ,talented.

    20)They can't think from user's view

                                                              Most Importantly, All will agree that,        "  Last thing ,we all people wanted that IE must beat chrome…….when we double click chrome icon on desktop ,it opens smoothly and feel softs,smoother …..whereas IE feels like scratches on our Screen,skin etc……!!!!!!!!!!!!!!!!.."

    I hope you will take a lesson from above .I'm an end user.

  136. BRAZIL says:

    Best Suport to all CSS and HTML5 Atributes!!! to all!! 100% msdn.microsoft.com/…/cc351024(VS.85).aspx

  137. HELLO_WORLD says:

    Most of what you folks are complaining about with regard to IE performance is caused by Add-Ons and security software hooks into IE.  Put IE9 and MSSE on a clean Win7 install and it screams.  Try IE9 on Windows Phone 7, it kills mobile Safari and Chrome.

    MS is fixing the problem, it's called IE10 on ARM based devices.  No more security software, no more add-ons…

    Also, in regards to standards, the problem these days is often not IE9.  Many times IE9 is implementing the final version of a standard while other browsers had their own implementation before the final version was settled and never modified their behavior to the final spec…

  138. temp says:

    It is clear that most of us are complaining about IE and this is going to continue unless Microsoft decide to take a big step forward innovation for IE. And notice that only power users come on this blog. I also have the same feeling, IE is not designed anymore for power users. It still have a good market share only because it is included with the OS. But this won't last forever. I hope MS will wake up before it is too late!

  139. temp says:

    Give us what we want in a browser but not what you think is good for us (like useless features pinned websites in IE9…)

  140. hamakaze(from japan) says:

    IE10 is please for 7 also support it. We ask, especially emphasizing the safety and balance the lack of speed and memory usage. The official version is very fun. Win8 and simultaneous release is a good us official version.

    One other, please let me block the ads here and there. I want to thank you for thinking in your company or other companies make the add-on or something for that. Even though Firefox or Chrome, you can with IE can't hate. So don't shove wrong click, and to avoid the serious….

  141. checho says:

    better provide a video with those long posts

  142. Not much to argue with in this post. Thanks for all this info.

    "To add hardware breadth, we have additional machine pools that run the spectrum of consumer scenarios. Good performance on these machines ensures that IE uses the underlying hardware effectively across the PC ecosystem."

    No AMD or Via? Core 2 Duo seems over-represented. Shouldn't the testing be biased to current generation chips, which are more representative of the CPUs that IE10 will be used on? Bias should also be towards low-end chips, especially mobile CPUs where battery life is critical. High-end CPU browser performance is relatively a non-issue.

    "Loading web content is also the only category that touches most of the browser’s eleven subsystems."

    The IE Blog article linked to is fascinating. So much time dedicated to Javascript! So browsing might be much faster with scripting completely disabled, but this would kill the functionality of too many sites. Perhaps a really quick way of enabling Javascript for a site – both temporarily and/or permanently – might be valuable?

    "Synthetic benchmarks – Rarely forgotten but often overstated are synthetic benchmarks like WebKit SunSpider. Benchmarks can be a useful engineering tool as they are designed to stress individual browser subsystems and accentuate differences between browsers. However, in order to maximize those differences, benchmarks often resort to atypical usage patterns or edge cases."

    This made me wonder what the *least* synthetic benchmark might be. Perhaps comparative ecommerce sales per time-on-site or time-on-product-page? Can't find a good link but i believe web design researchers like Jakob Neilsen have found a strong inverse correlation between page loading times and web sales. If that is true, faster browsers might show improved ecommerce results. If not true, then browser speeds are not the bottleneck, and at least in regard to online sales outcomes, would mean that all the major browsers are essentially identical in speed.

    "Idle tasks are a way for Windows and other developers to schedule non-critical work to happen at a later time when the user is not competing for resources. OS idle tasks include prefetching or SuperFetching, disk defragmentation, updating search indexes, and others, depending on OS version and configured services. To ensure that no idle work is done during the tests, the idle task queue is flushed."

    Presumably this is done via the command '%windir%system32rundll32.exe advapi32.dll,ProcessIdleTasks'. Anyway, this section got me thinking. Firstly, would running this command regularly myself stop the near-constant harddrive chatter that some non-SSD machines seem to experience? If yes, then maybe giving users an easily accessible method of running this command (Computer right-click menu?) would be a nice touch? (The idea being to run the command when stepping away from the PC for several minutes.)

    Secondly, the IE cache does not as far as i know, have any predictive type functionality like that of Prefetch and Superfetch. The only stuff in the cache is from sites the user has already visited. Could this be improved on by speculative behavior that was determined by the user's browsing history, combined with the site ranks of related sites? More specifically, the files within the Temporary Internet Files folder might not only relate to already browsed sites, but would also include content from the worlds most popular sites, and sites that were both popular and had strong browsing association to the the user's Favorites and History, but had *not* necessarily been visited by the user. In this scenario, the IE cache might talk to Bing, and download web content as a background task. In a sense, Bing would then become a quasi-proxy server for all IE users.

    I hope MSFT UI developers have a good understanding of the subjective aspects of UI/UX performance. This excellent post covers UX time scales. http://www.useit.com/…/timeframes.html It would be fair to note that the move from the Start Menu to Start screen involves moving from the 0.1 to 1.0 second bracket – a subjectively large retrograde step from a UI responsiveness PoV. This will surely require attention prior to the release candidate stage. More generally, i wonder if these subjective power of 10 timeframes are used as benchmarks within Windows development? Apparently, moves across these boundaries are much move significant for a user than moves with the boundaries.

    "Many third-party applications depend on Trident, the network stack, and other IE components. Extensions like BHOs and toolbars load within the IE context. Other applications, like security software, can inject themselves between IE components. These applications become part of the IE stack, and can lead to poor performance."

    Yuk. I support the move away from extensions and add-ons.

    Re the remarks @Paul Coddington, the readability issues of maximized and fullscreen browser windows should be given serious consideration. Half-screen browser windows are the way to go. Put it this way; What scenario would provide the most immersive experience – a webpage in a fullscreen browser session with excessive line length (from a readability PoV), or a half-screen session on a black background, displaying the same page with much closer to optimal line length? The second scenario might also improve battery life.

    Any other post on Windows development and testing would be welcome. Having said that, i'm curious to hear more about the Metro Contracts technology.