qqevzses
Sign in with
Sign up | Sign in

Benchmark Results: 3DMark 11 And PCMark 7

Windows 8: Double-Checking Performance On Core i7-3770K
By , Achim Roos

The benchmarks that Thomas ran a couple of days ago for his story Windows 8: Does AMD's Bulldozer Architecture Benefit? did not include Futuremark's synthetic PCMark and 3DMark tests. After all, although they're based on real-world Windows applications, there are times where we simply cannot reconcile the numbers we generate with these tests to our application testing.

What PCMark does do, however, is give us a more holistic look at system performance, whereas our application-based metrics generally isolate the CPU. With that in mind, let's have a look at how one machine handles two operating systems.

3DMark 11

For the most part, we see similar performance in both operating systems. The only exception is the Physics sub-test, where Windows 7 surprises us with a win. This benchmark is decidedly threaded, and there really isn't any good reason we can come up with for this difference to exist.

PCMark 7

Again, it's not necessarily clear why the results pan out the way they do, with Windows 8 claiming big leads in the overall benchmark and particularly the Computation sub-test. However, that result (based on a very processor-heavy workload) seems to contradict the 3DMark Physics sub-test (also very CPU-dependent). Could the difference be in PCMark's handling of Quick Sync technology, which we've seen in the past affects this benchmarks results heavily?

Ask a Category Expert

Create a new thread in the Reviews comments forum about this subject

Example: Notebook, Android, SSD hard drive

Display all 44 comments.
This thread is closed for comments
Top Comments
  • 25 Hide
    mayankleoboy1 , November 6, 2012 3:35 AM
    You are using winrar 3.98 , which isnt big on multithreading. Winrar4.2 is a huge change from the older versions, with quite aggressive multithreading. Please drop the old version and use the newer. :) 
  • 18 Hide
    mayankleoboy1 , November 6, 2012 3:38 AM
    most of these benchmarks are basically CPU bound. Can you do a mix test that does CPU + IO activity , like an Antivirus software ? or uncompressing 10GB+ rar/7z archives ?
    That would test if Microsoft has made changes in the scheduler to balance throughput VS response time.
Other Comments
  • 25 Hide
    mayankleoboy1 , November 6, 2012 3:35 AM
    You are using winrar 3.98 , which isnt big on multithreading. Winrar4.2 is a huge change from the older versions, with quite aggressive multithreading. Please drop the old version and use the newer. :) 
  • 18 Hide
    mayankleoboy1 , November 6, 2012 3:38 AM
    most of these benchmarks are basically CPU bound. Can you do a mix test that does CPU + IO activity , like an Antivirus software ? or uncompressing 10GB+ rar/7z archives ?
    That would test if Microsoft has made changes in the scheduler to balance throughput VS response time.
  • 7 Hide
    Darkerson , November 6, 2012 3:53 AM
    Well, at least things are on an overall even scale, unlike when vista came out and was decidely less snappy than XP. But with all the other modifications MS has made, such as the nice boost to start up and shut down speeds, its still "feels" faster than 7, even if it really isnt, overall. At any rate, I still dont feel this is "Vista 2" as a lot of people were making it out to be before launch.
  • 0 Hide
    SteelCity1981 , November 6, 2012 4:26 AM
    Overall Windows 8 doesn't improve over Windows 7 as both OSes are evenly matched and trade off performance over each other in diff benchmarks and you would think a newer OS like Windows 8 that is suppose to use less resources then Windows 7 should have outperformed Windows 7 across the board and yet it doesn't.
  • 9 Hide
    agnickolov , November 6, 2012 5:40 AM
    Alas, you skipped the Visual Studio benchmark. It showed slight aberration in Win7's favor on the AMD platform, so I was curious if that remained true for Intel as well.
  • 4 Hide
    manwell999 , November 6, 2012 7:40 AM
    Why bother doing performance comparisons if whenever the performance drops for the new version out comes the excuse "it will be fixed in the driver". What evidence is their to say this?

    A double-check means check-check, not check-excuse. Conclusions are drawn from the data at hand, not from speculation on what might be in the future.

    Otherwise cite the manufacturers acknowledgment that performance issues are going to be fixed and hold them to it.
  • -2 Hide
    darkstar845 , November 6, 2012 9:24 AM
    Is there a way to customize windows 8 to make it look like the desktop picture in the first page?
  • 5 Hide
    warmon6 , November 6, 2012 9:51 AM
    darkstar845Is there a way to customize windows 8 to make it look like the desktop picture in the first page?


    no. those are 2 different images (windows 7 and 8) that's been fused together.

    now, if you want that start button on windows 8, you could look at this article and try one of the 2 solutions they give.

    http://www.pcworld.com/article/2012810/another-way-to-restore-the-windows-8-start-button-and-menu.html
  • 3 Hide
    roflmaonow , November 6, 2012 10:22 AM
    warmon6no. those are 2 different images (windows 7 and 8) that's been fused together.now, if you want that start button on windows 8, you could look at this article and try one of the 2 solutions they give.http://www.pcworld.com/article/201 [...] -menu.html


    I just tried out pokki, it's really good for free program. Thanks for the heads up. I do like Start8 more though.
  • 7 Hide
    zyky , November 6, 2012 11:37 AM
    When did everyone forget how to make proper graphs. Can we get everything in the same sort order instead of sticking the winner on top each time? It's much harder to compare at a glance how it is now.
  • 4 Hide
    RazorBurn , November 6, 2012 11:41 AM
    The Two features of windows 8 that persuaded me to upgrade are

    1. Storage Spaces
    2. Data Deduplication

    jeffwouters.nl/index.php/2012/01/disk-deduplication-in-windows-8-explained-from-a-to-z/
    http://blogs.technet.com/b/filecab/archive/2012/05/21/introduction-to-data-deduplication-in-windows-server-2012.aspx

    Hope tom's will have a in depth look at Data Deduplication..
  • 2 Hide
    warmon6 , November 6, 2012 11:59 AM
    roflmaonowI just tried out pokki, it's really good for free program. Thanks for the heads up. I do like Start8 more though.


    Np, while i haven't cared if there a start button or not (been trying it out since the dev preview which did have it), I know other people have cared.

    I'm just showing options to those that want it back.
  • -9 Hide
    SteelCity1981 , November 6, 2012 12:23 PM
    TheCapuletLol, what?


    i'm sorry if that was over your head.
  • 8 Hide
    nekromobo , November 6, 2012 3:16 PM
    What about double checking battery life on laptops for both win7/8 and AMD/Intel camps?

    Kinda would think that win8 be way better in reserving battery?
  • 3 Hide
    maxinexus , November 6, 2012 3:16 PM
    Where is the comparison of PC boot, resume, or shut down times?
  • 0 Hide
    neiroatopelcc , November 6, 2012 3:20 PM
    Please compare windows server 2012 as well. You'll need to install the Wireless feature and enable the windows audio service, then it's like windows 8, but with desktop as default ui :) 
    maybe consider desktop experience feature too if those things are needed
  • 0 Hide
    clonazepam , November 6, 2012 5:35 PM
    I hadn't seen this Fritz benchmark before. I just tried version 4.2. I'm using an i7-2600k at 4.6ghz, with 16GB RAM at 1866. Let's just say my results are less than half of what you got on the i7-3770k so I think I'm missing something. Maybe I have too much overhead in processes, etc, or something...

    Edit: I closed about 40 processes, funny Chrome had 12 processes running taking over 2GB memory in total. Still, I managed to only get a score of 6006 lol...
  • 3 Hide
    PreferLinux , November 6, 2012 8:20 PM
    zykyWhen did everyone forget how to make proper graphs. Can we get everything in the same sort order instead of sticking the winner on top each time? It's much harder to compare at a glance how it is now.

    Worse, some graphs are in a particular (or is it random?) order, and some graphs have the winner on top. It is totally disgusting.
  • 0 Hide
    clonazepam , November 6, 2012 8:28 PM
    I believe it draws more attention to the graph's results. Read through an entire newspaper. Recall what you remember the next day, and most likely, it'll be discrepancies, typos, and misspellings. Another example is when you see a commercial, and each time its tweaked just slightly different so it sticks out like a sore thumb. ok done bs'ing...
Display more comments