Windows 2000 doesnt use HT Correctly?


Recommended Posts

Well a friend of mine is running a new p4 setup and has his windows 2000 cd which he is going to use. He has a chip with HT and 800mhz system bus, a few people told me HT doesnt work correctly in 2k, that XP is by far better, he doesnt want to buy XP unless he has to, he likes 2k and wants to stick with it. Did SP4 fix this problem? Thanks for you input

Link to comment
https://www.neowin.net/forum/topic/226593-windows-2000-doesnt-use-ht-correctly/
Share on other sites

  OptiPlex said:
It's not recommended you use Windows 2000 and HT together so turn off HT in the BIOS before installing it.

Win2k SHOULD run smoothly with HT on but I hear there's some negative impact on performance.

UNLESS YOU MULTITASK

cant forget that.

so unless he multitasks alot, turn it off, but I would leave it on, good to know its there

  Quote

Windows 2000 Professional is limited to two processors and doesn?t recognize the difference between a dual processor hyperthreaded system and a quad processor non-hyperthreaded systemThe 2nd logical processor on each physical processor would never be used.b>

  Quote
For Windows NT and Windows 2000, the answer i"It doesn't even know." These operating systems are not hyperthreading-aware because they were written before hyperthreading was invented.b> If you enable hyperthreading, then each of your CPUs looks like two separate CPUs to these operating systems. (And will get charged as two separate CPUs for licensing purposes.Since the scheduler doesn't realize the connection between the virtual CPUs, it can end up doing a worse job than if you had never enabled hyperthreading to begin with.b>
  Quote

A1 = CPU #1

A2 = Logical CPU #1

B1 = CPU #2

B2 = Logical CPU #2

Suppose you have two CPU-intensive tasks. As far as the Windows NT and Windows 2000 schedulers are concernedall four processors are equivalent, so it figure it doesn't matter which two it uses. And if you're unlucky, it'll pick A1 and A2, forcing one physical processor to shoulder two heavy loads (each of which will probably run at something between half-speed and three-quarter speed), leaving physical processor B idle; completely unaware that it could have done a better job by putting one on A1 and the other on B1.i>

Multitasking wouldn't even work properly in Windows 2000, those two statements explain it decently.

And HT has been known to make Win2k unstable or crash (not on all machines but some)

And in some cases it's been known to hurt performance.

It looks like HT causes more problems in Win2000 so my recommendation ... turn it off or hope it won't b0rk your install or anything...

Hmm

I am a bit confused here.

if the reasons why not to use it with w2k is true then why is it that Dual Xeon CPus used in servers running W2k server show up as 4 CPus and are used accordingly?

Here of all places i would notice a degradation in performance and response in the domain but this is not true.

Xeon 2.4 (533fsb) use HT technology without any problems in W2k server (same core as W2k pro).

My 3ghz prescott (w2k) runs all the same apps at the same level and power as my XP pro equivilent machine including 3dmark 2003,2005 & pcmark2004.

I dont have any problems with HT in W2k and performance in games and opengl Rendering is actually better in W2k with my ht cpu.

This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.
  • Posts

    • Win11 being 2.3x faster than Win10, so that means without changing any hardware, Doom Eternal will run at over 2x the framerate? Or that any computational stuff will be done 2x faster just because of OS? Press X for doubt. I wonder where they pulled these metrics out of coz they make no sense. Not to mention my laptop with Ryzen 2500U that used to run Windows 11 simply CAN'T anymore because of BS arbitrary requirements Microsoft has imposed in Win11 that weren't there intiially. So, that laptop is now running Ubuntu or Fedora and runs at 1000000000000000000000x faster speeds because Windows 11 just doesn't work at all so it gets DNF.
    • “Ask Photos” is coming to more Google Photos users in the US by Paul Hill Google has announced several improvements to a Gemini-powered Google Photos feature called Ask Photos. The feature launched last year in early access and it was great for longer search requests, however, when users typed “Dog” or “Beach” looking for those types of pictures, Ask Photos underperformed. Now, Google has addressed this and is bringing the feature to more eligible users in the United States. The search giant said that users enjoyed asking queries such as “suggest photos that’d make great phone backgrounds” or “what did I eat on my trip to Barcelona?” and getting Gemini-powered responses to these complex questions. Unfortunately, simple searches like “beach” or “dogs” were generating less than optimal results and people complained. To remedy the situation, Google has brought the best of Google Photos’ classic search feature to Ask Photos to improve latency. Easy, short requests will be dealt with by the old search mechanism and when you ask complex queries, it will switch over to the Gemini-powered search results. Now that Google has addressed this main issue of simple searches by integrated functionality from the previous search model, the company feels more confident to open up beyond early access and deliver it to more users in the United States. To be eligible, you must be 18+, be in the US, have your Google Account language set to English (United States), and have Face Groups turned on. This feature is only available on Android and iOS. To start using it, open Google Photos and tap on Ask at the bottom, then press Try now and agree to the terms. If you do not see the Ask button, it means you’re not yet eligible. The feature also can’t be used on computers yet. Hopefully, Google will quickly expand this feature outside the US so that international users can try it out too.
    • When are MS going to learn. Traditional home users who email, watch cat videos and perform basic tasks don't care if W11 is slightly faster...assuming it's true. Few home users care about faster speed and more secure claims. In any event, it's irrelevant. MS stopped caring about home users years ago as evident by their hostile, anti choice home user policies.
    • They have. https://www.xda-developers.com...-25h2-update-kind-of-small/
    • Microsoft shares detailed performance benchmarks for Windows 11 vs 10 to show the faster PC by Sayan Sen In the past, Microsoft has always made big claims regarding Windows 11 performance and how it is faster and better than Windows 10. For example, back in 2021, the company stated how the former was designed to get the best out of your system hardware. Later in 2023, Microsoft shared detailed Windows 11 performance improvements and upgrades it achieved on the new OS. While the claims were not substantiated with data, they were certainly quite interesting. Then, in 2024, Microsoft highlighted the differences in performance, citing a paid study, to show Windows 11 was way faster. As with most commissioned studies, the data was not represented fairly. This week, Microsoft has shared new numbers based on its December 2024 testing in a new support document. Here are the performance claims Microsoft made: Top Windows 11 PCs have up to 2.7 more hours of battery life than Windows 10 PCs Windows 11 PCs are up to 2.3x faster than Windows 10 PCs Windows 11 PCs offer up to 3.2x faster web browsing than Windows 10 PCs Windows 11 PCs offer up to 2x faster Microsoft Office productivity than Windows 10 PCs Thus, if you ask Microsoft, it will say that Windows 11 is better than Windows 10 in everything there is, be it web browsing, Office use, battery life, and overall performance in general. The company notes that the results are "based on a 95% confidence interval" for each OS across multiple tests. Microsoft also accounted for outliers. To reach such numbers, Microsoft used the following test metrics: For battery life, a local 1080p 24 fps MP4 video file playback was tested till 90% battery discharge on the Windows Media Player app. During the test, all settings were default except that screen brightness was set to 150 nits and Auto-brightness was disabled. Wi-Fi was connected to a network. For the "2.3x faster" claim, Geekbench 6 results have been cited. For web browsing, Speedometer was used, though Microsoft does not mention the version. And finally, for Office, Procyon Office productivity was used. Interestingly, Microsoft only used Intel CPUs (the company also recently recommended Intel chips for Windows 11 Pro PCs). The company has played it a smarter this time as it has "tested performance, battery life and application capabilities of a selection of Windows 11 PCs in comparison to a selection of Windows 10 PCs", where the Windows 10 PCs feature Intel Core 6th, 8th, and 10th gen chips and Windows 11 PCs pack Intel Core 12th and 13th gen. Thus, this is Microsoft essentially acknowledging that the underlying hardware itself actually plays a big role in the claims it has made. But to be fair to the company, there is also no way to run Windows 11, at least officially, on unsupported PCs, such that a direct comparison can be made. You can find the support document with full battery results in this article here on Microsoft's website.
  • Recent Achievements

    • Week One Done
      Marites earned a badge
      Week One Done
    • One Year In
      runge100 earned a badge
      One Year In
    • One Month Later
      runge100 earned a badge
      One Month Later
    • One Month Later
      jfam earned a badge
      One Month Later
    • First Post
      TheRingmaster earned a badge
      First Post
  • Popular Contributors

    1. 1
      +primortal
      567
    2. 2
      +FloatingFatMan
      177
    3. 3
      ATLien_0
      169
    4. 4
      Michael Scrip
      127
    5. 5
      Xenon
      119
  • Tell a friend

    Love Neowin? Tell a friend!