My bitchin S939 takes a shot at it with a GTX-680.
-Opty 185 @ 3133 MHz
-4GB DDR-418 CL2-3-2-5
-GTX 680
-Win7 64 SP1

Edit:
Hmm, I thought this was a good score, but now I'm looking at agent_x007's PD-4521/Win7 score. My CPU score is nice and high, but for some reason the game tests are lower...that's strange. I wonder if I'm running into some sort of platform bottleneck here. PCIe 1.1 or DDR1 RAM, maybe?

From an Optiplex 780
This is the best of Core 2 era integrated graphics from Intel. E8600 on Q45 with GMA X4500HD running Win7.
This actually outperforms the Radeon 9600 Pro that was in my first real gaming rig. The 9600 Pro only scored 3250 with a P4-1800.

The CPU score is a little low, but I've noticed that systems based on Intel IGPs tend to score lower in the CPU tests than systems with NVIDIA GPUs. Even my S939 Opteron (above) is doing better!

Standard Def Steve wrote:

Edit:
Hmm, I thought this was a good score, but now I'm looking at agent_x007's PD-4521/Win7 score. My CPU score is nice and high, but for some reason the game tests are lower...that's strange. I wonder if I'm running into some sort of platform bottleneck here. PCIe 1.1 or DDR1 RAM, maybe?

Well, ASUS P5B I used in my test can't do pcie 2.0, so that can't be it.
My CPU is slower (?)...but I think my GPU is "at fault" here (I was using a bit more powerfull than your gtx 680).

The ASRock 4CoreDual-SATA2 strikes again, this time with an E7600, the fastest dual core CPU it supports.

With AGP Radeon X1950Pro:

With GeForce 8800GTS 640MB:

@up 4CoreDual-SATA2 and E7600 is the fastest Dual Core you can get ?
Challenge Accepted 😀
PDC E5800 @ 4GHz/1066MHz

^Worth pointing out is that Your GPU is clocked higher than mine.

agent_x007 wrote:

@up 4CoreDual-SATA2 and E7600 is the fastest Dual Core you can get ?
Challenge Accepted 😀
PDC E5800 @ 4GHz/1066MHz

Yet the E7600 is beating the Pentium in the CPU tests and in GT1, which is the most CPU intensive game test. 😀

Though it's probably all down to cache size. The memory performance on these boards is abysmal, so cache makes a bigger difference than it normally would. I tested one of the old 1MB Pentium E2220 @ 3.2GHz/1066 on this board and the performance just sucked.

Interesting to see how close the GTS 450 and 8800GTS are in these tests. I wonder if the 8800 was holding back my Doom 3 score a bit.

It's a shame this thread hasn't had more love. Thought I'd wade in with my offering.

CPU: Phenom II X4 955 Black Edition @ 3.2GHz (Validation: hc0cbe)
GPU: 2x Sapphire Radeon HD4850 X2 (Validation: ev8d9)
Storage: Some OEM Samsung 850 EVO or similar (SATA III)

I have no idea how this result compares really - the cards are pretty old, but the CPU shouldn't be bottlenecking at all. I guess I'll know if I overclock and see a change or not. For reference, it scored nearly 17000 with a single X2 card. Click for validation.

Phenom is serious bottleneck for this kind of GPU power. For example - check my score with 2x HD 4830 and Kaby Lake i7. http://hwbot.org/submission/3495400_havli_3dm … 830_84017_marks or 2x HD 4870 + i5 2500k http://hwbot.org/submission/2210501_havli_3dm … 70_114477_marks

Here's the same XP gaming rig that was also benchmarked on the 3DMark 2001 Mega Thread.

Specs: ASUS A8V, AMD Athlon 64 3200+ @ 2.0 GHz, (2x1GB) 2GB PC-3200 DDR SDRAM, BFG nVidia Geforce 7800GS OC 256MB AGP 8x, Windows XP Pro SP3
Score: 11741

Standard Def Steve wrote:

From an Optiplex 780 This is the best of Core 2 era integrated graphics from Intel. E8600 on Q45 with GMA X4500HD running Win7. […]
Show full quote

From an Optiplex 780
This is the best of Core 2 era integrated graphics from Intel. E8600 on Q45 with GMA X4500HD running Win7.
This actually outperforms the Radeon 9600 Pro that was in my first real gaming rig. The 9600 Pro only scored 3250 with a P4-1800.

The CPU score is a little low, but I've noticed that systems based on Intel IGPs tend to score lower in the CPU tests than systems with NVIDIA GPUs. Even my S939 Opteron (above) is doing better!

http://i.imgur.com/YI9LXo5.png

with my 9500Pro I noticed something odd, depending on the driver version the score changed a lot with driver version, there must be some cheating or very good optimization going on with the fastest version, I think the 9600 can go a bit higher than that? sadly I can't test right now due to an accident with my 9500PRO (I had some confusion with the AGP lock and one small capacitor 😒 very embarrassing, but should be relatively easy to fix )

I'm not sure the 3dmark score for GMA 4500 translates into gaming, even with my HD 2000 I found games that felt poor even compared to the 9500, and the score for it is higher than the 9500

regarding CPU tests, I noticed that my PCI 8400GS is super slow in terms of CPU score, so I think the test is somehow bottlenecked by that and other aspects of the GPU/interface, so I can see Intel IGPs lowering the CPU score.

SPBHM wrote:

with my 9500Pro I noticed something odd, depending on the driver version the score changed a lot with driver version, there must be some cheating or very good optimization going on with the fastest version, I think the 9600 can go a bit higher than that? sadly I can't test right now due to an accident with my 9500PRO (I had some confusion with the AGP lock and one small capacitor 😒 very embarrassing, but should be relatively easy to fix )

I'm not sure the 3dmark score for GMA 4500 translates into gaming, even with my HD 2000 I found games that felt poor even compared to the 9500, and the score for it is higher than the 9500

regarding CPU tests, I noticed that my PCI 8400GS is super slow in terms of CPU score, so I think the test is somehow bottlenecked by that and other aspects of the GPU/interface, so I can see Intel IGPs lowering the CPU score.

I've noticed that about PCI video cards as well. With my Radeon 9250 PCI, a P4-2400 only scores 149 in the CPU test. A 3GHz Core 2 Duo only scores 157. Some serious bottlenecking going on there.

havli wrote:

Phenom is serious bottleneck for this kind of GPU power. For example - check my score with 2x HD 4830 and Kaby Lake i7. http://hwbot.org/submission/3495400_havli_3dm … 830_84017_marks or 2x HD 4870 + i5 2500k http://hwbot.org/submission/2210501_havli_3dm … 70_114477_marks

I didn't realise how true this is - I overclocked the CPU by 600MHz and I got 91900, then boosted the GPUs to get 95550. Might be able to push it above 100000!

brassicGamer wrote:

havli wrote:

Phenom is serious bottleneck for this kind of GPU power. For example - check my score with 2x HD 4830 and Kaby Lake i7. http://hwbot.org/submission/3495400_havli_3dm … 830_84017_marks or 2x HD 4870 + i5 2500k http://hwbot.org/submission/2210501_havli_3dm … 70_114477_marks

I didn't realise how true this is - I overclocked the CPU by 600MHz and I got 91900, then boosted the GPUs to get 95550. Might be able to push it above 100000!

100K is definitely possible. it only takes a GTX 970 and slightly over 4GHz to do so.

Still, there's a huge difference between the Phenom II and a Core i7 running the same video card (see the first picture on the first page).

Standard Def Steve wrote:

100K is definitely possible. it only takes a GTX 970 and slightly over 4GHz to do so.

HayiwHT.png

Still, there's a huge difference between the Phenom II and a Core i7 running the same video card (see the first picture on the first page).

Yep, I found that astounding to be honest. But hey, I've built a 'Shit Hot For '09' system, not a modern system so, for that purpose (DirectX 10) it's probably respectable.

Any K6 owners having any luck getting the CPU tests to run? Running '03 on my K6-2/500 with a GF4 Ti4400, it completes the first three game tests, then crashes to the desktop without any errors when it gets to the CPU tests. It's not the lack of SSE; I've seen it run on a Tbird Athlon before.

Standard Def Steve wrote:

Any K6 owners having any luck getting the CPU tests to run? Running '03 on my K6-2/500 with a GF4 Ti4400, it completes the first three game tests, then crashes to the desktop without any errors when it gets to the CPU tests. It's not the lack of SSE; I've seen it run on a Tbird Athlon before.

You don't need CPU tests, to get a 3DMark 03 overall score.

agent_x007 wrote:

Standard Def Steve wrote:

Any K6 owners having any luck getting the CPU tests to run? Running '03 on my K6-2/500 with a GF4 Ti4400, it completes the first three game tests, then crashes to the desktop without any errors when it gets to the CPU tests. It's not the lack of SSE; I've seen it run on a Tbird Athlon before.

You don't need CPU tests, to get a 3DMark 03 overall score.

I'd just like to see how my K6-2/500 compares to the PIII Katmai-500, which scores 105 in the CPU test.

"Madness" test 2 (3DMark 03 v3.4.0) :

Core i5-3470 @ 3391MHz
Intel HD 2500.

I thought this was fairly impressive, considering that HD 2500 is the slower Ivy Bridge IGP. Some Ivy processors have HD 4000, which is almost twice as fast.

The CPU score is really high too. Clocked at a conservative 3391 MHz, it's much faster than Phenom II at 4070 MHz. And this is all running on the IGP, which typically produces lower CPU scores than systems with Nvidia graphics.

"Project Madness" 3DMark 03 (after Win98 native reinstall) :