Page 3 of 3 FirstFirst 123
Results 31 to 43 of 43
  1. #31
    Joined
    Oct 2001
    Location
    Jax,Fla,USA
    Age
    48
    Posts
    1,561

    Re: AMD HT Link: Never above 1000mhz?

    The only way to prove it was not with in the margin of error (1-3%) is to repeat the test many times and at several different HTT speeds. That my friend is scientific method, and how you prove your point.

    My point is you can run back to back bechmarks in Prime and get different results with the same settings, usually 1-3% difference.
    Last edited by vick1000; 01-22-2007 at 12:32 PM.
    Quote Originally Posted by JC Denton
    Quote Originally Posted by Bob Page
    I will burn like the brightest star...
    You're gonna burn all right.

  2. #32
    Joined
    Jan 2004
    Location
    Croatia
    Age
    39
    Posts
    453

    Re: AMD HT Link: Never above 1000mhz?

    When I re-run those benchmarks, numerous times, I will post here.

  3. #33
    Joined
    Aug 2000
    Location
    Vvardenfell
    Age
    58
    Posts
    10,887

    Re: AMD HT Link: Never above 1000mhz?

    SuperPI is very unusual as benches go because it involves a massive (and pretty artificial) amount of memory swapping. As such it's about the only app or bench (apart from specifically memory-testing benches) which does on occasion show a difference. Tighter memory settings and faster RAM speeds will also help SuperPI.

    Now try the same trick with any version of 3DMark, even 2001.


    M

  4. #34
    Joined
    Aug 2004
    Posts
    10,610

    Re: AMD HT Link: Never above 1000mhz?

    HTT speeds shouldn't effect it as the traffic it causes is on the memory bus. Except for reading the program (I) and printing the results(O), nothing happens on the HTT bus.


    "The significant problems we face cannot be solved at the same level of thinking we were at when we created them."
    - Albert Einstein (1879-1955)

  5. #35
    Joined
    Nov 2002
    Posts
    9,762

    Re: AMD HT Link: Never above 1000mhz?

    I know. Sometimes I think the Hypertransport is as high as it is for marketing reasons. There's no real need for it to be as fast as it is.

    Intel's FSB...that's a whole other story!
    Q6600
    8GB corsair XMS DDR2
    MSI K9N Platinum 780i
    2*Evga 8800GT at 650Mhz (factory)
    x-fi Platinum
    150GB Raptor, 2TB Others
    Corsair 650W PSU
    Lian Li PC65 case

  6. #36
    Joined
    May 2005
    Location
    Hattiesburg, MS
    Age
    36
    Posts
    259

    Re: AMD HT Link: Never above 1000mhz?

    Bump... any more results???
    AMD 9950 BE @ 3.0 GHz
    AMD 4870
    4 GB Corsair RAM
    Foxcon Mobo
    1 x WD Velociraptor 300GB
    1 x WD Raptor 74GB
    1 x WD 120GB
    Antec Case

    "Pink hearts, yellow moons, green clovers, orange stars... That leprechaun's on aciiiid!"
    -Dave Chappelle/Black Gallagher

  7. #37
    Joined
    Jun 2004
    Posts
    903

    Thumbs down Re: AMD HT Link: Never above 1000mhz?

    I dont think the HT Link makes much difference either way , and neither does Nvidia LinkBoost
    Intel Core i7 920 Nehalem @4Ghz
    CoolerMaster V8
    EVGA X58 3X SLI
    Antec Signature 850W
    Antec Twelve-Hundred
    6GB Corsair Dominator DDR3 1600mhz(Triple Channel)
    EVGA GTS 250 (500mb) x 2 (SLI)
    WD Caviar Black SATA 1TB
    Seagate 500G External
    Samsung SyncMaster T260 (HDMI)
    HK SoundSticks II
    Vista64 SP2

  8. #38
    Joined
    Jul 2003
    Location
    Kannapolis, NC
    Age
    32
    Posts
    1,425

    Re: AMD HT Link: Never above 1000mhz?

    my HTT is reported to be running at 1250 and it's just fine

    I don't think this has anything to do with it, but bumping my FSB up to 250 gave my aquamark 9k from 6k in CPU lol.

    http://valid.x86-secret.com/show_oc?id=179744

    If anyone sees this btw, am I overclocking right? I can only remember it's make the FSB high as possible
    Last edited by Havox; 03-21-2007 at 04:55 PM.

  9. #39
    Joined
    Jan 2004
    Posts
    3,463

    Re: AMD HT Link: Never above 1000mhz?

    If it's stable it's stable. People lower the HT speed to achive stability at virtually no cost to performance. However, if you already have perfect stability, there is little point in changing it.

    Hypertransport speed is relivant in a few cases:

    - Multi-socket systems.

    - If you have a rediculous number of PCI-E lanes and hardware that will utilize it. For example, going over 1000MHz might help a system with dual 8800GTXs each in a full 16x slot, while playing oblivion at 2560x1600 with everything maxed out, while copying a terabyte file from your 12 drive RAID 0 to your other 12 drive RAID 0 (each one on a separate PCI-E 8x raid controler), While blasting porn into your external firewire and USB storage from your maxed out dual-gigabit ethernet....all at the same time.

    Now, when video cards come out that can saturate the 16x PCI-E slots and everyone (meaning mostly no one) starts using 3 or 4 of them, HT will start to really matter.

    Quote Originally Posted by Mr_Speedy View Post
    As you can see, the first set of results are faster in every single test.
    Margin of error.

    Prime95 doesn't even touch the hypertransport bus, so I'm betting somthing else is responsible for the differences.

  10. #40
    Joined
    Dec 2009
    Location
    central new york
    Posts
    26

    Re: AMD HT Link: Never above 1000mhz?

    Quote Originally Posted by oralpain View Post
    If it's stable it's stable. People lower the HT speed to achive stability at virtually no cost to performance. However, if you already have perfect stability, there is little point in changing it.

    Hypertransport speed is relivant in a few cases:

    - Multi-socket systems.

    - If you have a rediculous number of PCI-E lanes and hardware that will utilize it. For example, going over 1000MHz might help a system with dual 8800GTXs each in a full 16x slot, while playing oblivion at 2560x1600 with everything maxed out, while copying a terabyte file from your 12 drive RAID 0 to your other 12 drive RAID 0 (each one on a separate PCI-E 8x raid controler), While blasting porn into your external firewire and USB storage from your maxed out dual-gigabit ethernet....all at the same time.

    Now, when video cards come out that can saturate the 16x PCI-E slots and everyone (meaning mostly no one) starts using 3 or 4 of them, HT will start to really matter.



    Margin of error.

    Prime95 doesn't even touch the hypertransport bus, so I'm betting somthing else is responsible for the differences.
    that is funny even in 2010

  11. #41
    Joined
    Nov 2002
    Location
    In bed with one of my avatar AMD girls :D
    Age
    39
    Posts
    8,876

    Re: AMD HT Link: Never above 1000mhz?

    Quote Originally Posted by oralpain View Post

    Prime95 doesn't even touch the hypertransport bus, so I'm betting somthing else is responsible for the differences.
    So whats a good stability test for HT Link then?

  12. #42
    Joined
    Jul 2001
    Location
    Reality: Playground for the unimaginative.
    Age
    44
    Posts
    3,896

    Re: AMD HT Link: Never above 1000mhz?

    Damn zombie threads coming back to life!

  13. #43
    Joined
    Mar 2009
    Location
    Belding, Michigan
    Age
    70
    Posts
    139

    Talking Re: AMD HT Link: Never above 1000mhz?

    Does that make this The Night of the Living Thread ? (sorry I couldn't resist)
    Last edited by Sidd; 01-30-2010 at 09:07 AM. Reason: sp

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •