VOGONS


First post, by kva

User metadata
Rank Newbie
Rank
Newbie

Hello again 😉
I wrote an article about Pentiums Pro and Overdrives. They are all tested in more or less modern software. It is in polish but it doesn't matter that much, charts are universal. Enjoy 😀

http://kva.pl/index.php?link=artykul&id_a=9

Celeron Tualatin vs Celeron Conroe at equal clocks
Pentium Pro 256k vs Pentium Pro 1M vs Pentium II Overdrive!
VIA C3 vs VIA C7

My website all about old hardware

Reply 2 of 43, by kva

User metadata
Rank Newbie
Rank
Newbie

It is, benchmarking took about a month (every test was done at least three times, some of them taking forever...) 😁 . There are more tests working on POD but not working on PPro (because lack of MMX). They will be presented in the next article - POD vs P II.

I should have Dhrystone results, will check later.

Celeron Tualatin vs Celeron Conroe at equal clocks
Pentium Pro 256k vs Pentium Pro 1M vs Pentium II Overdrive!
VIA C3 vs VIA C7

My website all about old hardware

Reply 3 of 43, by lazibayer

User metadata
Rank Oldbie
Rank
Oldbie
kva wrote:

It is, benchmarking took about a month (every test was done at least three times, some of them taking forever...) 😁 . There are more tests working on POD but not working on PPro (because lack of MMX). They will be presented in the next article - POD vs P II.

I should have Dhrystone results, will check later.

That's exciting as well! Are you testing PII-333? Will they both be tested on the same motherboard?

Reply 4 of 43, by havli

User metadata
Rank Oldbie
Rank
Oldbie

Great article!
I did similar tests (also using Pentium Pro... and many other CPUs) and it really takes a extreme amount of time. 😀
Your methodology is quite similar to mine - do you also plan to compare PPro to Pentium MMX, K6, Cyrix M2 or other CPUs from this era? P6 is really strong when running true 32-bit SW.

HW museum.cz - my collection of PC hardware

Reply 7 of 43, by luckybob

User metadata
Rank l33t
Rank
l33t

The issue with a quake benchmark is the video card choice. I did some testing and Quake loves uncacheable speculative write combining (USWC) and not all motherboards/bioses/video cards support this. The Matrox G200 did and it's a 100% increase in FPS with it enabled.

I did some crude testing here: The Ultimate 686 Benchmark Comparison

BTW, still reading through!

It is a mistake to think you can solve any major problems just with potatoes.

Reply 8 of 43, by kva

User metadata
Rank Newbie
Rank
Newbie

I was testing CPUs in a way which allows me to compare anything with anything. The goal is to build a database of scores CPUs from 486 to moderns CPUs. Quake is not suitable for that.

Next one is about VIA C3 vs VIA C7 (both 1Ghz). A little spoiler - C7 isn't always better 😉

Dhrystone:
POD: 149443247
PPRO 256K: 85418837
PPRO 1MB: 83557957

There is no mistake, 256K is slightly faster (every test taken 3 times, AVG presented). I have no idea why.

Celeron Tualatin vs Celeron Conroe at equal clocks
Pentium Pro 256k vs Pentium Pro 1M vs Pentium II Overdrive!
VIA C3 vs VIA C7

My website all about old hardware

Reply 9 of 43, by RJDog

User metadata
Rank Member
Rank
Member

Wow, POD retailed for $599? That seems like a great deal, for the time. Given that, I'm surprised they aren't easier to come by these days.

Reply 10 of 43, by kva

User metadata
Rank Newbie
Rank
Newbie

It wasn't that good. When it was released there was already Pentium II 400 on the market. Better of course than POD. And for PII you could buy mobos with AGP and SD-Ram. That also made a big difference.

Celeron Tualatin vs Celeron Conroe at equal clocks
Pentium Pro 256k vs Pentium Pro 1M vs Pentium II Overdrive!
VIA C3 vs VIA C7

My website all about old hardware

Reply 11 of 43, by kva

User metadata
Rank Newbie
Rank
Newbie

Is anyone interested in comparison dual Pentium II Overdrive vs dual Celeron 366 (Abit BP6 ftw 😉 )?

Celeron Tualatin vs Celeron Conroe at equal clocks
Pentium Pro 256k vs Pentium Pro 1M vs Pentium II Overdrive!
VIA C3 vs VIA C7

My website all about old hardware

Reply 15 of 43, by kva

User metadata
Rank Newbie
Rank
Newbie

Same as above, windows XP SP3, the same set of benchmarks/programs. But I am open for every suggestion and can test for you whatever you want 😉

Celeron Tualatin vs Celeron Conroe at equal clocks
Pentium Pro 256k vs Pentium Pro 1M vs Pentium II Overdrive!
VIA C3 vs VIA C7

My website all about old hardware

Reply 16 of 43, by mrau

User metadata
Rank Oldbie
Rank
Oldbie
kva wrote:

Same as above, windows XP SP3, the same set of benchmarks/programs. But I am open for every suggestion and can test for you whatever you want 😉

quake 3 in smp? 😁 also i always wondered if the memory controller on those multi-ppro boards was up to the task (ie does memory transfer double when 2 cpus are used on tasks where memory speed is the only real factor)

Reply 18 of 43, by kva

User metadata
Rank Newbie
Rank
Newbie
mrau wrote:

also i always wondered if the memory controller on those multi-ppro boards was up to the task (ie does memory transfer double when 2 cpus are used on tasks where memory speed is the only real factor)

No, memory controller is integrated into chipset so it doesn't matter if there is one or two CPUs. So in memory intensive jobs scalling is rather bad. You can se it here:
23_pentium_pro_overdrive_winrar_test_eclipse.jpg

Two PODs are much faster than PPro 1MB cache but the difference in this test is small, thats because of much too slow memory transfers.

I will do my best with Quake.

Celeron Tualatin vs Celeron Conroe at equal clocks
Pentium Pro 256k vs Pentium Pro 1M vs Pentium II Overdrive!
VIA C3 vs VIA C7

My website all about old hardware

Reply 19 of 43, by luckybob

User metadata
Rank l33t
Rank
l33t

instead of winrar- can you give 7zip a shot? I did on my dual p-pro. and posted my results here 7-zip benchmarks for over 60 CPUs

It is a mistake to think you can solve any major problems just with potatoes.