bloodem wrote on 2022-03-25, 20:25:

Necrodude wrote on 2022-03-25, 19:37:

What do you guys get with Vanilla Quake3 stock settings?
Demo four.

This is what I got with the most recent SS7 platform I tested (Asus P5A rev 1.04 + AMD K6-3+ 570 @ 633 MHz): https://youtu.be/jrV6vPGR3Yc?t=1570

I didnt hear any sounds. Was it without a sound card? I guess that you would get around 45 to 55 fps with a sound card.

Necrodude wrote on 2022-03-26, 06:40:

I didnt hear any sounds. Was it without a sound card? I guess that you would get around 45 to 55 fps with a sound card.

Yes, this is explained at the beginning of the video (+ in the description).
The K6-3+ test was a follow-up to a previous K6-2+ test that I had done on the same platform with same settings and at the same 633 frequency. In that particular video (which is 46 minutes long), I had tested b0th with and without sound.

To keep the K6-3+ video a bit shorter I only benchmarked without sound, because it's already pretty clear from all the data in the two videos that, in most games, the K6-3+ will be somewhere around 5-10% faster than the K6-2+ is (with or without sound).

So, since the K6-2+ @ 633 MHz managed to score 44.2 FPS in Quake 3 (with sound), the K6-3+ (at the same frequency) will be around ~ 47 FPS.

You can find more details about these videos in this thread here: VIDEO BENCHMARKS (Original title: Settling the VIA C3 performance debate once and for all)

bloodem wrote on 2022-03-26, 07:16:

Yes, this is explained at the beginning of the video (+ in the description). The K6-3+ test was a follow-up to a previous K6-2+ […]
Show full quote

Necrodude wrote on 2022-03-26, 06:40:

I didnt hear any sounds. Was it without a sound card? I guess that you would get around 45 to 55 fps with a sound card.

Yes, this is explained at the beginning of the video (+ in the description).
The K6-3+ test was a follow-up to a previous K6-2+ test that I had done on the same platform with same settings and at the same 633 frequency. In that particular video (which is 46 minutes long), I had tested b0th with and without sound.

To keep the K6-3+ video a bit shorter I only benchmarked without sound, because it's already pretty clear from all the data in the two videos that, in most games, the K6-3+ will be somewhere around 5-10% faster than the K6-2+ is (with or without sound).

So, since the K6-2+ @ 633 MHz managed to score 44.2 FPS in Quake 3 (with sound), the K6-3+ (at the same frequency) will be around ~ 47 FPS.

You can find more details about these videos in this thread here: VIDEO BENCHMARKS (Original title: Settling the VIA C3 performance debate once and for all)

Thanks! 😀 I will go out in my garage and fire upp my k6-3+ 400@616 on a Asus p5a rev 1.04 , 128mb PC133 cl2 sdram, geforce 256 SDR.
It shall be interesting to se what I get with vanilla q3, stock settings with sound.

Edit: Im back! I got 43 frames per second with Vanilla Quake3 with stock settings.

Necrodude wrote on 2022-03-26, 08:19:

Thanks! :) I will go out in my garage and fire upp my k6-3+ 400@616 on a Asus p5a rev 1.04 , 128mb PC133 cl2 sdram, geforce 256 […]
Show full quote

bloodem wrote on 2022-03-26, 07:16:

Yes, this is explained at the beginning of the video (+ in the description). The K6-3+ test was a follow-up to a previous K6-2+ […]
Show full quote

Necrodude wrote on 2022-03-26, 06:40:

I didnt hear any sounds. Was it without a sound card? I guess that you would get around 45 to 55 fps with a sound card.

Yes, this is explained at the beginning of the video (+ in the description).
The K6-3+ test was a follow-up to a previous K6-2+ test that I had done on the same platform with same settings and at the same 633 frequency. In that particular video (which is 46 minutes long), I had tested b0th with and without sound.

To keep the K6-3+ video a bit shorter I only benchmarked without sound, because it's already pretty clear from all the data in the two videos that, in most games, the K6-3+ will be somewhere around 5-10% faster than the K6-2+ is (with or without sound).

So, since the K6-2+ @ 633 MHz managed to score 44.2 FPS in Quake 3 (with sound), the K6-3+ (at the same frequency) will be around ~ 47 FPS.

You can find more details about these videos in this thread here: VIDEO BENCHMARKS (Original title: Settling the VIA C3 performance debate once and for all)

Thanks! 😀 I will go out in my garage and fire upp my k6-3+ 400@616 on a Asus p5a rev 1.04 , 128mb PC133 cl2 sdram, geforce 256 SDR.
It shall be interesting to se what I get with vanilla q3, stock settings with sound.

Edit: Im back! I got 43 frames per second with Vanilla Quake3 with stock settings.

I guess I need to run 3Dmark2000 as well. So that I dont get to off topic 😀

So what WPCRED settings did you guys use?

Tyan s1564
k6-3+ 550 at 400mhz
Audigy 2 zs
512mb 60ns
Radeon 9200 pro 128bit
Windows xp sp2

Andddd radeon 7500 64 bit

Has anyone tried to change the PCI latency timer to a higher value? I changed it from 32 to 64 in my ASUS P5A BIOS. I Gained 5 extra frames per second in Quake3. I will run 3dmark2000 tomorow. Lets hope i will get some more points

Necrodude wrote on 2022-03-30, 21:55:

Has anyone tried to change the PCI latency timer to a higher value? I changed it from 32 to 64 in my ASUS P5A BIOS. I Gained 5 extra frames per second in Quake3. I will run 3dmark2000 tomorow. Lets hope i will get some more points

That is weird, I would have thought that a larger timer would decrease performance

What other options do you have? Can you map them all out

I found this when googling

https://archive.techarp.com/showFreeBOG41e9.h … ang=0&bogno=138

It seems that a higher latency gives the graphic card more time to send data. A too high latency could cause issues, because other cards gets to wait.

I also tried to raise the latency to 128. I didnt have time to look at it in detail. But it seemed that I gained another 1.2 extra fps i in Quake3. But I need to do some further testing.

Edit:
128 was a bit to much. I felt the system became a bit stuttery. So I settled on 96. I gained 1 extra frame per second in Quake3.
I will try to run some benches of 3dmark2000 during the weekend.

Necrodude wrote on 2022-03-31, 06:13:

I found this when googling […]
Show full quote

I found this when googling

https://archive.techarp.com/showFreeBOG41e9.h … ang=0&bogno=138

It seems that a higher latency gives the graphic card more time to send data. A too high latency could cause issues, because other cards gets to wait.

I also tried to raise the latency to 128. I didnt have time to look at it in detail. But it seemed that I gained another 1.2 extra fps i in Quake3. But I need to do some further testing.

Dang! Pretty awesome!

I don’t think my freeway mobo has this option, I wonder if there is a way to add it.🤔

Top score of a non Super 7? My TXP4 with an FX 5500 128 megabyte/64 bit card using the 43.45 driver force installed as an FX 5200 Ultra. Interestingly, I tried a 256 megabyte/128 bit FX 5200, but I could not top this score with it - a 128 bit card might do better in some systems, but in this case I believe it's so limited by the CPU that the extra memory and better bandwidth made no difference. My NVS 280 64/64 card could also match this score on this system.

🤣, 7.55 FPS.

Repo Man, try with a Geforce 2 and driver 7.76?
I have yet to try Aquamark 🤣. Was one of my favorites at that time, used to bench the crap out of my AMD 1700+ 9800Pro 256

Skalabala wrote on 2022-04-07, 07:05:

Repo Man, try with a Geforce 2 and driver 7.76?
I have yet to try Aquamark 🤣. Was one of my favorites at that time, used to bench the crap out of my AMD 1700+ 9800Pro 256

A Geforce 2 in PCI is a very rare card, I've never seen one listed on Ebay.

Repo Man11 wrote on 2022-04-07, 08:30:

A Geforce 2 in PCI is a very rare card, I've never seen one listed on Ebay.

There always are quite a few on ebay (like this one), but their price is usually exaggerated.

Slowly getting to 5k. This is on a Freeway VIA motherboard.

supercordo wrote on 2022-04-11, 15:24:

Slowly getting to 5k. This is on a Freeway VIA motherboard.

Very good. Two megabytes of onboard cache that can handle 124 MHz FSB makes me jealous as neither of my MVP3 boards has cache that will run at that speed.

If you haven't already, try adjusting your AGP apeture size; I've found that reducing it can increase your score.

Repo Man11 wrote on 2022-04-11, 15:33:

supercordo wrote on 2022-04-11, 15:24:

Slowly getting to 5k. This is on a Freeway VIA motherboard.

Very good. Two megabytes of onboard cache that can handle 124 MHz FSB makes me jealous as neither of my MVP3 boards has cache that will run at that speed.

If you haven't already, try adjusting your AGP apeture size; I've found that reducing it can increase your score.

I will try that. I believe it is currently at 64.

2048KB L3 running at 124MHz? How?! Now that is an achievement by its own!
Jealous on the Ti500!

Skalabala wrote on 2022-04-12, 08:18:

2048KB L3 running at 124MHz? How?! Now that is an achievement by its own!
Jealous on the Ti500!

With This beauty. Freeway fw-5vgf+/ultra. I need a faster CPU. This is the fastest i can get this one to go. Going to try to mode a K6-2+ 570. https://www.ultimateretro.net/en/motherboards/10151