Pentium 1@150mhz at 3 different fsb and Multiplier settings benchmarks \ VOGONS

Pentium 1@150mhz at 3 different fsb and Multiplier settings benchmarks

Topic actions

First post, by Imperious

Posted on 2016-11-19, 13:02

Imperious Offline

Rank Oldbie

Rank: Oldbie
Posts: 732
Joined: 2012-06-18, 14:02
Location: Brisbane, Australia

I got this p133 cC0 SY028 off Ebay a couple of weeks ago. I was doing some messing around with overclocking on an ALI Alladin mobo I got, 5ALI61 Rev F. I did get it to
boot at 200mhz but nearly everything I tried to run crashed, no surprise there I guess. I found I could run the same speed of 150mhz at 3 separate fsb and multiplier
settings, so thought I would run some benchmarks and document the results. Obviously the sdram is being severely underclocked at 60mhz and it's hard to find any information
out there about the PCI speed at 75mhz fsb, but I would assume 100fsb = 33mhz PCI and 60mhz fsb = 30mhz PCI. The video card is a PCI Voodoo Banshee.
I used Phils Dos Benchmark version 2 suite for most of these results.

Last edited by Imperious on 2016-11-19, 13:32. Edited 1 time in total.

Atari 2600, TI994a, Vic20, c64, ZX Spectrum 128, Amstrad CPC464, Atari 65XE, Commodore Plus/4, Amiga 500
PC's from XT 8088, 486, Pentium MMX, K6, Athlon, P3, P4, 775, to current Ryzen 5600x.

Reply 1 of 8, by James-F

Posted on 2016-11-19, 13:25

James-F Offline

Rank Oldbie

Rank: Oldbie
Posts: 1505
Joined: 2015-11-30, 04:10

Nice!
So can we say that the boost in performance primarily comes from the memories of the system (L1, L2, RAM, Video)?

my important / useful posts are here

Reply 2 of 8, by Imperious

Posted on 2016-11-19, 13:39

Imperious Offline

Rank Oldbie

Rank: Oldbie
Posts: 732
Joined: 2012-06-18, 14:02
Location: Brisbane, Australia

Looking a bit closer at the Speedsys results, it appears to be the L2 (Motherboard cache) that shows the most improvement from 60fsb to 100fsb, also the SDRAM. The L1 cache doesn't show massive
improvements which makes sense as it's running at 150mhz irrespective of fsb and multiplier.

Atari 2600, TI994a, Vic20, c64, ZX Spectrum 128, Amstrad CPC464, Atari 65XE, Commodore Plus/4, Amiga 500
PC's from XT 8088, 486, Pentium MMX, K6, Athlon, P3, P4, 775, to current Ryzen 5600x.

Reply 3 of 8, by James-F

Posted on 2016-11-19, 13:47

James-F Offline

Rank Oldbie

Rank: Oldbie
Posts: 1505
Joined: 2015-11-30, 04:10

This shows that the Memory works really closely with the CPU.
If I remember correctly upgrading to DDR was a big performance boost back then.

my important / useful posts are here

Reply 4 of 8, by clueless1

Posted on 2016-11-19, 14:13

clueless1 Offline

Rank l33t

Rank: l33t
Posts: 4714
Joined: 2015-12-22, 17:43
Location: Midwest US

Really interesting results, thank you for taking the time to make up the charts and post them.

Look at how much 3dbench is effected by the increase in RAM speed compared to the other benchmarks. It is also the least-demanding benchmark in the bunch. Less pixels to push=greater jump in performance with increase in RAM speed. The idea follows suit with Wolf3D and Chris 320, and to a lesser extent, Doom. The more demanding the benchmark, the less effect faster RAM has.

The more I learn, the more I realize how much I don't know.
OPL3 FM vs. Roland MT-32 vs. General MIDI DOS Game Comparison
Let's benchmark our systems with cache disabled
DOS PCI Graphics Card Benchmarks

Reply 5 of 8, by mrau

Posted on 2016-11-19, 18:01

mrau Offline

Rank Oldbie

Rank: Oldbie
Posts: 1145
Joined: 2015-11-28, 12:43

win quake is software or 3dfx accelerated? i mean this result is either due to banshee power or missing DOS setup imho..

Reply 6 of 8, by Imperious

Posted on 2016-11-20, 01:38

Imperious Offline

Rank Oldbie

Rank: Oldbie
Posts: 732
Joined: 2012-06-18, 14:02
Location: Brisbane, Australia

mrau wrote:
win quake is software or 3dfx accelerated? i mean this result is either due to banshee power or missing DOS setup imho..

3dfx accelerated. One little detail I forgot to add. I don't see a lot of point in Running benchmarks in software mode under windows if the game works in Dos.

Atari 2600, TI994a, Vic20, c64, ZX Spectrum 128, Amstrad CPC464, Atari 65XE, Commodore Plus/4, Amiga 500
PC's from XT 8088, 486, Pentium MMX, K6, Athlon, P3, P4, 775, to current Ryzen 5600x.

Reply 7 of 8, by Imperious

Posted on 2016-11-20, 01:42

Imperious Offline

Rank Oldbie

Rank: Oldbie
Posts: 732
Joined: 2012-06-18, 14:02
Location: Brisbane, Australia

clueless1 wrote:
Really interesting results, thank you for taking the time to make up the charts and post them.

Look at how much 3dbench is effected by the increase in RAM speed compared to the other benchmarks. It is also the least-demanding benchmark in the bunch. Less pixels to push=greater jump in performance with increase in RAM speed. The idea follows suit with Wolf3D and Chris 320, and to a lesser extent, Doom. The more demanding the benchmark, the less effect faster RAM has.

Thanks for Your insights. I might rerun the 3Dbench result later just to be certain I wrote it down right.

Atari 2600, TI994a, Vic20, c64, ZX Spectrum 128, Amstrad CPC464, Atari 65XE, Commodore Plus/4, Amiga 500
PC's from XT 8088, 486, Pentium MMX, K6, Athlon, P3, P4, 775, to current Ryzen 5600x.

Reply 8 of 8, by nforce4max

Posted on 2016-11-20, 01:46

nforce4max Offline

Rank l33t

Rank: l33t
Posts: 2451
Joined: 2012-05-05, 22:55

Old news to me 🤣 but yes better higher clocked FSB gave a nice boost, not much bandwidth at 60mhz and 66 isn't great but 75 and 83mhz is the sweet spot for a lot of socket 5/7 boards when the option to do so is available. SS7 Pentium and Pentium MMX have a FSB wall at 100mhz (tired 105 on a few boards and no post) but are pretty snappy on those boards.

On a far away planet reading your posts in the year 10,191.

Go to top of page Go to top of page

Back to General Old Hardware