Standard Def Steve wrote:

That's quite a system! Though I'm a little surprised it scored so low--the BX is supposed to be the fastest P3 chipset at 133MHz. I was able to hit 11460 with a P3-1575 and 9800Pro on an 815 board. I wonder if Win98 is holding your machine back.

I am a bit surprised myself. Expected some more tbh. I guess I can try and get XP installed and test that out.

· Tekram P6BX-A (440BX) mobo
· Tualatin 1400-S @ 1575MHz/150FSB
· 768MB PC-133 SDRAM
· Geforce 6200 64-bit (MSI NX6200AX-TD128LF)
· WinXP Pro SP3 32-bit

Seems a crappy score, many Coppermines in this thread score better, and the other Tuallys cream my arsch with much higher scores 😵 .

Is this just due to the Geforce 6200 being a mediocre GPU, or there is something terribly wrong in this system 😦 ?

And by the way ...

· ASUS M4A89GTD Pro
· Phenom II 960T
· 16GB DDR3 1600
· Radeon 6870
· Win 7 SP1 32-bit

Yeah, 6200 is a mediocre GPU. You should get 9-10k with like 6800.

Wiped my system and installed Windows XP Pro.

Was also playing around with a pair of 1Ghz/256kb Xeons. I didn't expect them to boot on the XG-DLS, but they did, so I did some benchmarks at 750Mhz and 840Mhz (7.5 x 112). I also redid my benchmarks for the 700/1Mb Xeons.

Looks like there's some extra overhead on XP Pro that dropped my scores on these a bit, but not by much. Looks like I lost 1-2 frames in most sections, not really noticeable.

I'm thinking about finding a pair of 933/256kb Xeons to do a more fair comparison since they'll run at the same clock speed as the 700/1Mb Xeons. Also looking for a good i840 based workstation board like the Iwill DCA200 or the Supermicro S2DM3, that would probably level the playing field a bit.

P3 Xeon 1Ghz/256kb SL4HE @ 750Mhz

P3 Xeon 1Ghz/256kb SL4HE @ 840Mhz

P3 Xeon 700/1Mb SL49P @ 700Mhz

P3 Xeon 700/1Mb SL49P @ 784Mhz

On my Athlon-XP 2200, 1.5GB RAM, Geforce 6800, I get 10,000 3DMarks. It fluctuates by about 10 3DMarks on either side.

Here are some more benchmarks from me.

All from the same system just with different CPUs:

Motherboard M6TBA
Memory 2x256MB
OS Windows XP Pro SP2
HDD 4.3GB HDD
Video Geforce 256 SDR 32MB AGP
Sound ESS PCI

Score - CPU
589 - Pentium 2 233
799 - Pentium 2 333 @ 300 (4.5x66)
821 - Celeron 300A (4.5@66)
1001 - Pentium 2 350 (3.5x100)
1500 - Pentium 3 450 (4.5x100)
1620 - Pentium 3 600 (6x100)

After overclocking the GF6200 to just below the apparition of artifacts there is only a slight boost in score. Flogging a dead horse 😵 .

Same specs as in sig. Changed drivers to nForce 45.23. Was previously running 56.72.

Best run on 56.72

Best run on 45.23

m1919 wrote:

Same specs as in sig. Changed drivers to nForce 45.23. Was previously running 56.72.

Best run on 56.72

Best run on 45.23

That's very impressive! Your 784/112 Xeon scored exactly the same as my 1000/133 with a 9800 Pro (page 1).

3DMark01 must like the extra cache. Makes me wonder what a 2MB Xeon would score. 😀

Thought I'd post a single 3DMark 2000 shot just because of the insanely high CPU score.

This is 3DMark2k on a PIII-S at 1575MHz with a Radeon 9800Pro.
732 is BY FAR the best 3D'00 CPU score I've ever seen from a P3. In fact, my Athlon 64 3500+ only scores 817 on this test.

3DMark00 was one of the early SSE optimized tests, which makes me wonder if it's simply ignoring the A64's SSE capabilities. After all, the A64 demolishes the PIII in 3DMark 03's CPU test.

Voodoo3 3000D AGP on Dell 440BX board, Tualatin-1200 @ 900, 512MB, Win98SE, Amigamerlin 2.9 driver

Swapped out the A350 for an A380 (FX5950 Ultra). Running Omega Drivers based on Forceware 53.03.

This is probably as good as it'll get until I get my hands on a pair of 900Mhz 2MBs.

How about something more exotic

cyrix 6x86-133, geforce2mx200

when i ran 3dmark01 it gave a "you need a cpu with mmx support" message, but still let me in(3dmark2000 just quits at this).
it gave no score after testing, because the speed of the "car chase: high detail" test is below 0.1fps, which is considered "test was not run". according to the formula of 3dmark01 score calculation, the actual score is manually calculated to be around 175 points.

then i downclocked the 6x86 to 66mhz(66*1), and switched to a rage128vr videocard because nvidia crashes at any opengl programs with this cpu. both car chase tests reported no fps(below 0.1) this time, and the manually calculated result is a whopping 35 points!

Reference 5950 Ultra, Overclocked to 530Mhz core, 965Mhz memory. Omega Forceware 53.03 Drivers.

Didn't expect to see this thread again!
8K is a pretty amazing score for a 785MHz processor...WOW. Looks like cache size really does affect 3D'01, because my old Coppermine 1000/133 could only score 7620 with a 9800 Pro.

I guess I may as well post the updated score for my tweaked-to-perfection Tualatin beast:
1400-S at 1585MHz - 151MHz bus
Stock-clocked Radeon 9800 Pro 128MB - running Catalyst 7.3
Asus TUV4X mobo - VIA 694 chipset
1.5GB PC133 at 151MHz/CL2
XP SP3 - all updates

I'm really tempted to overclock the 9800Pro to see if I can break 12K, but the cooler is a puny little thing!

Lol, I was surprised I got such an even number.

I'm sure the extra gain was mostly from overclocking the 5950 Ultra. I actually tried to push the my processors to 805, but I got a BSOD soon after starting 3DMark, so I guess I'd need a bit more voltage to get stability. Unfortunately, I can't actually adjust voltages on this board, so I'm topped at 785Mhz until I can find a pair of 900/2MB Xeons. I'll bet I could get over 10k with a pair overclocked to just past 1Ghz.

I've been using 3DMark for many years, and I made a quite comprehensive list. Unfortunately, I did not noted all the time the CPU. Where I did, I will write it down.

Celeron 600 oc 900, TNT2 M64 - 984 3DMarks '01
Celeron 1200, Radeon 9250 - 4270
, GeForce FX5500 - 4748
Duron 1300, Radeon 9250 - 4392
, GeForce FX5500 - 4858
FX5200 - 6146, CPU was some Athlon XP, AFAIR.
FX5700 - 12553
Radeon 7000 - 835
Radeon 9550 - 6043
Athlon XP 1800+, GeForce 2MX - 2751
Intel 845G - 1353
Pentium 4 2800, Intel 915G - 3951
Sempron 2800+ socket 754, GeForce 6200TC PCI Express (cousin's) - 4000
Sempron 2600+ socket 754, GeForce 6200 128bit AGP (mine) - 10174; so, a 128bit 6200 is not that crappy after all!
GeForce 6800 Golden Sample - 17019
Core 2 Duo E8200, Radeon X1650 - 16011
Radeon HD2600 Pro - 24296

And now a series of tests on my previous main rig, an Athlon X2 4200+
Radeon X550 - 8672
GeForce 7300GS - 13621
7300GT - 18658
onboard GeForce 7050 - ~6100
Radeon HD3650 - can't remember, and I can't find the bits&scraps of paper I used to write on.
Radeon HD6790 - 28422; kinda low in '01, but it blew my mind in '03.
Just for the sake of comparision: 42214 '03, 12221 '05, 8653 '06
Same HD6790 on my current Athlon II x3 455: 28658 '01, 49771 '03, 19605 '05, 14210 '06, 9905 Vantage and 2454 '11.

Last but not least, several cards on my number 2 PC, a Sempron 3000+ socket A.
Radeon 7500 - 2342 '01
9250 - 5448
9550 - 7812
GeForce FX5500 - 5090 (kinda low, but then the card has some bad capacitors which need replacement)
GeForce 6600 - 11037 (again, weird, just a bit better than socket 754 Sempron 2600+ and GF 6200 I had until 2007); then again, sk. 754 Sempron is a different beast than sk. A; a cut down Athlon 64 instead of a rebadged Athlon XP.

And a fresh test done on my work computer:
CPU: Intel Pentium Dual-Core E5200 2.5 GHz, 3 Gb DDR2, onboard Intel G31/33

And a second test; can't figure it out why the score dropped a bit; overall, quite decent performance for onboard video, especially Intel.