http://www.win-raid.com/t15f37-NVIDIA-Optimiz … -Vista-Win.html
If someone is willing to give them a try. I'll be free to play with my hardware around Christmas 🙁
But ... ha, I do "steal" some time at work to read the forum 😀
Hey it just occurred to me that in my quest to figure out why VIA seems sluggish I did some 2D GUI speed tests years ago. I even posted about it here and the thread is still buried here.
Essentially VIA AGP causes performance degradation for the 2D GUI acceleration of graphics cards. I used a benchmark called Tom2D. On NVIDIA and Intel AGP performance is considerably better. I ran the test on G400, Radeon 8500 and GeForce 3. The chipsets tested were K8T800, 865G and nForce2 U400.
wrote:http://www.win-raid.com/t15f37-NVIDIA-Optimiz … -Vista-Win.html
If someone is willing to give them a try. I'll be free to play with my hardware around Christmas 🙁
But ... ha, I do "steal" some time at work to read the forum 😀
I have used these on nForce4 and its siblings like 6150. With Win7 I only recommend getting the Ethernet driver for nf4. Use the Windows drivers for everything else. Nforce3 I don't know anything about but I suggest trying any Windows stock drivers first.
With XP I am not entirely sure if there are benefits to using these remixes over NVIDIA official downloads. The remixes do seem to work fine though.
wrote:Hey it just occurred to me that in my quest to figure out why VIA seems sluggish I did some 2D GUI speed tests years ago. I even posted about it here and the thread is still buried here.
Essentially VIA AGP causes performance degradation for the 2D GUI acceleration of graphics cards. I used a benchmark called Tom2D. On NVIDIA and Intel AGP performance is considerably better. I ran the test on G400, Radeon 8500 and GeForce 3. The chipsets tested were K8T800, 865G and nForce2 U400.
I have to chime in since Athlon XP is basically the last IBM compatible in my house besides the gaming rigs. I too noted terrible performance in Windows XP (which is what finally drove me to use Linux on my PCs). I still have this rig with an ASUS A7V333 motherboard and a Athlon XP-M 2500+ in it. The curious thing is that using Windows does feel like an exercise in futility whereas Linux runs fantastic, even today. I remember back in 2005 when Ubuntu was fresh out of the oven I tried it and to my horror discovered that I could up the settings in DooM3 and it still performed better.
In Windows XP I could run DooM3 at medium quality and a resolution of 800x600 and it would still occasionally stutter.
In Ubuntu I could run DooM3 at medium quality and a resolution of 1024x768 and maintain a higher average framerate than Windows, without the odd stutters.
My theory is that Windows simply did not have proper support for the Athlon XP era hardware. Maybe the VIA Windows drivers were crap or something. Today my rig runs at 1826 MHz (166 MHz FSB) and sure enough YouTube will choke on it, but if I download the clip that Flash can't handle and play it back on MPlayer, it will run beautifully.
So yeah, I'll cast my vote for the 'obsolete but not quite retro lot'.
[Install Win95 like you were born in 1985!] on systems like this or this.
It's the VIA AGP that is apparently the problem. nForce2 performs better. Though I only tested GDI performance. I don't know of 3D is indeed slow too. That is something that reviewers would have exposed I think. Few reviews looked at 2D speed. In fact the Tom2D benchmark was born from the question of why Radeon HD 2000-5000 series cards seemed so slow at 2D. Toms Hardware wanted to take a closer look. It was a rare moment of GUI performance investigation.
Here is the old thread of results.
VIA AGP impact on GUI performance
wrote:http://www.win-raid.com/t15f37-NVIDIA-Optimiz … -Vista-Win.html
If someone is willing to give them a try. I'll be free to play with my hardware around Christmas 🙁
But ... ha, I do "steal" some time at work to read the forum 😀
INTERESTING! I have an old 939 system and a 775 system. the 775 system has ACHI but, my SSD caps out at 5.9 on the WEI, perhaps this driver(s) could help?
They are just a mix of driver versions newer than the official NVIDIA downloads.
If you are running a SSD on an nForce chipset it is probably running at SATA 150 rate. The chipsets don't work properly with SSDs for some reason. This isn't really a practical issue though because the main benefit of SSDs is the latency, not so much the bandwidth.
wrote:They are just a mix of driver versions newer than the official NVIDIA downloads.
If you are running a SSD on an nForce chipset it is probably running at SATA 150 rate. The chipsets don't work properly with SSDs for some reason. This isn't really a practical issue though because the main benefit of SSDs is the latency, not so much the bandwidth.
So i was right to not worry about reliability/garbage collection, the board just doesn't have the through put. are there any performance benefits from these newer drivers?
I think you are probably best off with the stock Windows IDE driver. But I do like the Ethernet drivers in his packs. I saw CPU usage reduction in Win7 on nF4 when transferring at full gigabit speed, compared to the driver 7 includes.
i don't think it's an IDE driver cause all the AHCI function are turned on.
wrote:i don't think it's an IDE driver cause all the AHCI function are turned on.
Specifically - which NVIDIA chipset do you have?
wrote:... Murder caps? ... The number of the beast is 3300uf […]
... Murder caps? ... The number of the beast is 3300uf
Far from it pal, as a matter of fact they're top-notch MBZs, a great Rubycon series of caps.
Exceptionally high ripple and low ESR (the MBZ 6.3V 3300uF 10mm are 2800 ripple & just 0.012 ESR). Superb caps for mobo CPU VRM input/output, and, provided the case they've been living on was decently ventilated (heat is an assasin of electrolytic capacitors), probably still in good health after a decade.
You can identify the Rubycon brand caps by the vent pattern at the top, like a 'K' letter.
Let the air flow!
wrote:You can identify the Rubycon brand caps by the vent pattern at the top, like a 'K' letter.
That is good to remember 😀
Im starting to think that the Zalman 7700 cu was to blame when it comes to the smoked CPU.
I diddnt notice that it was mounted badly when I removed it but its a heavy heat sink.
I guess if the weight wasnt 100% evenly distributed over the core that could be enough to smoke the CPU.
I had moved the motherboard with the CPU and heat sink mounted earlier the same day. But its strange that it didn't happen at once when I powered the system on if the heat sink is to blame.
New PC: i9 12900K @5GHz all cores @1.2v. MSI PRO Z690-A. 32GB DDR4 3600 CL14. 3070Ti.
Old PC: Dual Xeon X5690@4.6GHz, EVGA SR-2, 48GB DDR3R@2000MHz, Intel X25-M. GTX 980ti.
Older PC: K6-3+ 400@600MHz, PC-Chips M577, 256MB SDRAM, AWE64, Voodoo Banshee.
wrote:That is good to remember 😀.
The Electrolytic Cap Spotter's Field Guide
The Solid Cap Spotter's Field Guide
🤣
THESE are A7N8X murderers: http://oi42.tinypic.com/fd9jjc.jpg. Notice all eight bloated KZGs are 'taking off' from the mobo due to bottom bulging 😵 .
Let the air flow!
wrote:I have to chime in since Athlon XP is basically the last IBM compatible in my house besides the gaming rigs. I too noted terribl […]
wrote:Hey it just occurred to me that in my quest to figure out why VIA seems sluggish I did some 2D GUI speed tests years ago. I even posted about it here and the thread is still buried here.
Essentially VIA AGP causes performance degradation for the 2D GUI acceleration of graphics cards. I used a benchmark called Tom2D. On NVIDIA and Intel AGP performance is considerably better. I ran the test on G400, Radeon 8500 and GeForce 3. The chipsets tested were K8T800, 865G and nForce2 U400.
I have to chime in since Athlon XP is basically the last IBM compatible in my house besides the gaming rigs. I too noted terrible performance in Windows XP (which is what finally drove me to use Linux on my PCs). I still have this rig with an ASUS A7V333 motherboard and a Athlon XP-M 2500+ in it. The curious thing is that using Windows does feel like an exercise in futility whereas Linux runs fantastic, even today. I remember back in 2005 when Ubuntu was fresh out of the oven I tried it and to my horror discovered that I could up the settings in DooM3 and it still performed better.
In Windows XP I could run DooM3 at medium quality and a resolution of 800x600 and it would still occasionally stutter.
In Ubuntu I could run DooM3 at medium quality and a resolution of 1024x768 and maintain a higher average framerate than Windows, without the odd stutters.My theory is that Windows simply did not have proper support for the Athlon XP era hardware. Maybe the VIA Windows drivers were crap or something. Today my rig runs at 1826 MHz (166 MHz FSB) and sure enough YouTube will choke on it, but if I download the clip that Flash can't handle and play it back on MPlayer, it will run beautifully.
So yeah, I'll cast my vote for the 'obsolete but not quite retro lot'.
Hey my god parents have windows XP Athlon XP machine that is a tad slugish, does this problem carry over to windows 7? or possibly even 8.1? i know you need mods to get 8.1 to run on older hardware though.
ORrrrr, does it just have 512kb cache versus 1mb?
