1. #1
    Firefly
    Joined
    Nov 2002
    Location
    Texas
    Posts
    855

    Anyone here ran 3dmark 2001SE on a Sandy Bridge-E system?

    Anyone here ran 3dmark 2001SE on a Sandy Bridge-E system (i7 3820, 3930k, or 3960x)? I recently compared the scores in 3Dmark 2001SE of my i7 2600k and my i7 3930k, and my i7 2600k I get like 61000 and on my 3930k I get like 52000 which really doesn't make sense. It's the Dragothic tests and the Nature test that are much slower on my 3930k than my 2600k and this brings down the overall score in 3dmark 2001SE for me on my 3930k. Can anyone here with a SB-E CPU and Windows 7 test out 3dmark 2001SE and run just the Dragothic and Nature tests and tell me what the fps were on each of those tests? Also tell me what video card you are using during those tests? I would appreciate it. I'm using Windows 7 Professional x64 and a GTX 580.

  2. #2
    Junior Proxycon Emperor
    • Computer Specs
      • Motherboard or Laptop:
      • X79-UP4
      • Chipset:
      • X79
      • CPU:
      • i7 3930k
      • Memory:
      • 32GB Corsair Vengence
      • Memory Speed:
      • 1866
      • Video Card(s):
      • EVGA GTX 780 SC
      • Operating System:
      • Windows 7 Professional

    Joined
    Jan 2001
    Location
    England
    Posts
    10,505

    Re: Anyone here ran 3dmark 2001SE on a Sandy Bridge-E system?

    Unless it was run on same drivers at the same clocks on cpu and gpu you're going to get results like these.

    Newer drivers will have less dx7/8 optimisations no one is benching them, and is a waste of resources. Also the 3960k isn't always faster than a 2600k clock for clock as the larger L3 cache takes a latency hit. But you buy lga2011 more for the platform than cpu.

  3. #3
    Firefly
    Joined
    Nov 2002
    Location
    Texas
    Posts
    855

    Re: Anyone here ran 3dmark 2001SE on a Sandy Bridge-E system?

    Quote Originally posted by DhinShin View Post
    Unless it was run on same drivers at the same clocks on cpu and gpu you're going to get results like these.

    Newer drivers will have less dx7/8 optimisations no one is benching them, and is a waste of resources. Also the 3960k isn't always faster than a 2600k clock for clock as the larger L3 cache takes a latency hit. But you buy lga2011 more for the platform than cpu.
    The X79 is a superior platform than the P67 though. The reason is because the hardware is better on that platform and not as much limitations on the X79. L3 cache latency should not really make much of a difference because I read that L2 cache has around a 90-95% hit rate but that was what I read 14 years ago and maybe with the requirments of more RAM in software over they years, the L2 cache rate for 256KB maybe got lower and 256KB of L2 cache can't mask the slower L3 cache and RAM speed as well as in the mid-late 90's?

  4. #4
    Senior Onion
    • Computer Specs
      • Motherboard or Laptop:
      • P7P55D-E PRO
      • Motherboard BIOS:
      • BIOS 1703
      • Chipset:
      • P55
      • CPU:
      • i7 875k
      • CPU Overclock Speed:
      • 4Ghz
      • CPU Cooler:
      • Noctua NH-D14
      • Memory:
      • 4x4GB DDR3 Ram
      • Memory Speed:
      • 1600Mhz
      • Video Card(s):
      • Gigabyte GTX 670
      • Video Card Speed:
      • 1319/7000
      • Sound Card:
      • Asus Xonar Essence STX
      • PSU:
      • Antec 650 EarthWatt
      • Hard Drive(s):
      • 2x60GB SSD, 1x 1TB HDD WD
      • Optical Drives:
      • LiteOnIt DVD burner
      • Operating System:
      • Windows 7 64Bit SP1

    Joined
    Mar 2002
    Location
    New York City
    Posts
    3,211

    Re: Anyone here ran 3dmark 2001SE on a Sandy Bridge-E system?

    I do not have any kind of SB. I do have i7-875k Lynnifield:

    http://mnetcs.com/thumb/storage/b5182189.JPG

    54080 3DMark01SE
    Ram speed is on 800(1600Mhz on dual channel.)
    Last edited by OrionCheung; April 21, 2012 at 23:54.

  5. #5
    Futuremark Staff
    Joined
    Jun 2000
    Location
    Finland
    Posts
    9,031

    Re: Anyone here ran 3dmark 2001SE on a Sandy Bridge-E system?

    I should also note that 3DMark 2001SE is not a suitable benchmark for modern systems. Videocard is completely irrelevant for it (it is 100% CPU on today's systems) and even there the bottleneck is such that I can completely see how 2600K would trump over the Sandy-E CPUs.

    Use 3DMark 11 or 3DMark Vantage for those systems. They give you accurate comparisons.

  6. #6
    Fresh Member
    • Computer Specs
      • Motherboard or Laptop:
      • Asus Rampage IV Extreme (MIPS R4E Ni/BLK Block Kit)
      • Chipset:
      • X79
      • CPU:
      • Intel i7 3930K
      • CPU Overclock Speed:
      • 5178Mhz
      • CPU Cooler:
      • Swiftech Apogee HD + XSPC EX420 + Alphacool UT60 240 + UT60 560
      • Memory:
      • 16GB G.Skill Ripjaws Z 2133CL9-GBZH
      • Memory Speed:
      • DDR3-2584 10-12-10-26 1T
      • Video Card(s):
      • EVGA GTX670 FTW 2GB SLI w HK Hole Edition 680 Blocks/Backplates + GTX560Ti 448cores PhysX (1098c)
      • Video Card Speed:
      • 1394(core) / 7448(mem) - Bgears Blasters 140mm + Koolance FAN-12025HBK/12038HBK P-P Fans
      • Sound Card:
      • Creative Titanium HD
      • PSU:
      • NZXT HALE90 850W
      • Hard Drive(s):
      • Samsung 830 256GB, Plextor M5P-Xtreme 512GB, 2x WD10EZEX 1TB RAID0, 3x WD VR 1TB RAID0, 2x WD RE 4TB
      • Optical Drives:
      • LG Slim Slot-Load Blu-Ray Burner
      • Operating System:
      • Windows 7 Pro x64 SP1

    Joined
    Jul 2013
    Location
    Ohio
    Posts
    1
    I don't get scores "all over the map" like some people, in fact they are quite consistent...

    Benching my 3930K at speeds from stock up to 5.38Ghz, there was an almost linear correlation between CPU speed and performance.

    If you want more points, I suppose you could disable 4 of the 6 cores (run bench on Core 1; leave Core 0 for the OS)?

Posting Permissions

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