Standard Def Steve wrote:

Skyscraper wrote:

A Willamette P4 is not a good CPU for running Doom 3...

1024*768 Ultra: 31.3 FPS 😁

P4 Willamette 1900@2375MHz, Asus P4B i845, 512MB PC133@166MHz CL3-3-3-7, Geforce 6800GT 92.91, onboard audio. XP-SP3

You know what sucks even more than a Willy P4?
A Northwood Celeron. 25.2 fps at 2.6GHz, and that's with DDR! 🤣

The same CPU running with the same chipset but with PC133 memory is even worse 😁

1024*768 Ultra: 19.3 FPS

Celeron 2600 Northwood 128kb, Asus P4B i845 SDR,512MB PC133 CL2-2-2-6, Geforce 6800GT 92.21, onboard audio, XP-SP3

Skyscraper wrote:

A Willamette P4 is not a good CPU for running Doom 3...
1024*768 Ultra: 31.3 FPS 😁

P4 with SDRAM is very slow - DDR is much better... 😀

P4 Willamette 1.6 @ 1.92 GHz + Asus P4P800 Dlx. + 2x 512 MB DDR 320 CL2-2-2-5 + GF 7800 GS @ 500/1400
39.3 fps

this result (Celeron northwood) is lower than my p3 650@866 with the 8400GS PCI

19.9 FPS, Pentium 3 650@866, Asus CUBX-E, 512MB PC133 C3, Geforce 8400GS G98 340.52, AWE64, Win XP SP3

Skyscraper wrote:

The same CPU running with the same chipset but with PC133 memory is even worse 😁

1024*768 Ultra: 19.3 FPS

Celeron 2600 Northwood 128kb, Asus P4B i845 SDR,512MB PC133 CL2-2-2-6, Geforce 6800GT 92.21, onboard audio, XP-SP3

I...just...no words! They should've sent a poet!

That's gotta be getting dangerously close to Cyrix in terms of IPC. 😁 It's almost unbelievable that Intel of all companies was making CPUs with such piss poor performance in 2003-2004. And of course OEMs like Compaq didn't make things any better by pairing those CPUs with 256MB of shared memory.

SPBHM wrote:

this result (Celeron northwood) is lower than my p3 650@866 with the 8400GS PCI

p38400gs d3.jpg

19.9 FPS, Pentium 3 650@866, Asus CUBX-E, 512MB PC133 C3, Geforce 8400GS G98 340.52, AWE64, Win XP SP3

That's a really good result for an 866. That's right up there with my Coppermine at 1133.

DDR memory helped the Willamette a bit but not very much

1024*768 Ultra: 34.4 FPS
Willamette P4 2.0, Asus P4B533-E i845E, 1x512MB DDR266 CL2-2-2-5, Geforce 6800GT 92.91, onboard audio, XP-SP3

1024*768 Ultra: 43.9 FPS
Willamette P4 2.0@2340, Asus P4B533-E i845E, 1x512MB DDR312 CL2-2-2-5, Geforce 6800GT 92.91, onboard audio, XP-SP3

I did also manage to get a Gallatin running in the old i845 board

1024*768 Ultra: 91.2 FPS
Gallatin P4 EE 3.4 no HT, Asus P4B533-E i845E, 1x1024MB DDR400 CL2-3-3-6, Geforce 6800Ultra 92.91, onboard audio, XP-SP3

1024*768 Ultra: 97.4 FPS
Gallatin P4 EE 3.4@3740 no HT, Asus P4B533-E i845E, 1x1024MB DDR440 CL2.5-3-3-7, Geforce 6800Ultra 92.91, onboard audio, XP-SP3

very nice, have you disabled HT or the board doesn't allow you to enable it? at least with 1.31 I suspect HT might help a little bit

Standard Def Steve wrote:

Skyscraper wrote:

19.9 FPS, Pentium 3 650@866, Asus CUBX-E, 512MB PC133 C3, Geforce 8400GS G98 340.52, AWE64, Win XP SP3

That's a really good result for an 866. That's right up there with my Coppermine at 1133.

I guess the 8400GS pci is good for something... running doom 3 benchmark with sub 1GHz p3s 🤣 because with the Xeon it wasn't really a good result

also it scales pretty well with the clock increase, the stock result (650) was a little under 15FPS

SPBHM wrote:

very nice, have you disabled HT or the board doesn't allow you to enable it? at least with 1.31 I suspect HT might help a little bit

HT was disabled but it works perfectly. The motherboard officially supports the P4 Northwood 3.06 so HT support was added to this board in BIOS 1011 late 2002 when that CPU was released. I did not see any performance difference with or without HT in Doom3 or 3dmark 2001 but as the score without HT was 0.1 FPS higher that is the score for the list 😁.

At overclocked speed I only benched with HT off as this boards isnt designed for running an overclocked Gallatin. HT adds alot of power consumtion and I did not want the VRM mosfets to go up in smoke 😁. The 3dmark 2001 score at stock speed I posted in the 3dmark 2001 thread is with HT enabled.

I will move the Gallatin CPU to my Asus P4C800 motherboard to rebench that system as the old score with the P4C800 was with the slower Doom3 1.3 version. It looks strange with the old P4B533-E i845e system higher up in the list than the dual channel memory Canterwood system using the same CPU so I need to fix that ASAP 😉

Edit:

Here is a real first world problem 😁

I have too many motherboards to keep track of where I put them... I own 3 Asus Canterwood boards, so far I have found my P4C800 Deluxe rev 2.00 as it is the board I use in the system I borrowed the Gallatin from. I did also find my P4C800Gold rev 2.00 in a half built system but not my P4C800-E Deluxe. I prefer my P4C800-E when overclocking as it has all VRM mosfet spots occupied while the other two boards only have 2 mosfets for each power phase.

Not all P4C800-E have all the VRM mosfets as I have seen pics of such boards with mosfets missing, perhaps a previous owner added the 3 extra. I did find my Asus P4P800-E Deluxe and it has all mosfet spots occupied but it is a 865PE board so it wont perform as good as the Canterwood boards.

I have not seen the board Im looking for lately, its probably in its box in a larger box under other boxes in some dark corner of a closet so I think I will have to make do with one of the other boards for now. I have pushed the Gallatin pretty hard before with these boards with missing mosfets but that was with an ambient temperature of 18 - 21C. Its summer now, in my tinkering corner its 30C today 😒

Anything Goes:

Doom3 Timedemo [v1] = 168.6

ECS P4M890T-M V2, C2D/E6700, 2gb PC4300/533 [single channel], 7900GTX/v92.91, XPpro/SP3

Waiting for some low latency dual channel ram to arrive - I'll re-run the benchmark again when it does. Apples to apples this motherboard performs quite a bit better than my default Biostar P4M890-TE [147 vs 168] but unfortunately for the life of me I can not get the 7900GTX to run in anything but VGA mode in W98 with it. So very likely I'll roll back to the Biostar after benching the dual channel ram and sacrifice performance for dual boot. I'm bummed since I've had success with all the other VIA based 890 boards I've tried using the Rloew memory patch and modded v82.69 NV drivers.

Okay, dusted off the ol' period-accurate system again (P4EE 3.2/2 GB DDR-400 2.5-3-3-6/6800 Ultra/XP SP3), decided to connect it to my main desktop's PEXHDCAP to test out the capture card's capabilities for PC capture with both VGA and DVI-D, and ran Doom 3 (1.3.1 Steam) again while I'm at it.

800x600 High = 71.3 FPS
1024x768 High = 71.9 FPS (a slight increase!)
1280x1024 High = 71.3 FPS
800x600 Ultra = 70.5 FPS
1024x768 Ultra = 70.5 FPS
1280x1024 Ultra = 69.3 FPS

This system is frankly baffling me with its performance right now. I mean, now it's performing much closer to what it should be, but it's still about 20 FPS short of similarly-specced systems in here.

NamelessPlayer wrote:

Okay, dusted off the ol' period-accurate system again (P4EE 3.2/2 GB DDR-400 2.5-3-3-6/6800 Ultra/XP SP3), decided to connect it […]
Show full quote

Okay, dusted off the ol' period-accurate system again (P4EE 3.2/2 GB DDR-400 2.5-3-3-6/6800 Ultra/XP SP3), decided to connect it to my main desktop's PEXHDCAP to test out the capture card's capabilities for PC capture with both VGA and DVI-D, and ran Doom 3 (1.3.1 Steam) again while I'm at it.

800x600 High = 71.3 FPS
1024x768 High = 71.9 FPS (a slight increase!)
1280x1024 High = 71.3 FPS
800x600 Ultra = 70.5 FPS
1024x768 Ultra = 70.5 FPS
1280x1024 Ultra = 69.3 FPS

This system is frankly baffling me with its performance right now. I mean, now it's performing much closer to what it should be, but it's still about 20 FPS short of similarly-specced systems in here.

Here is my Gallatin 3.4 period correct system benchmarked with Doom3 v1.0. This is the best non overclocked period correct P4 result so far, only 0.4 FPS from the dual Opteron! 😀

1024*768 Ultra: 105.6 FPS

P4 EE "Gallatin" 3.4. Asus P4C800 Deluxe v2.0. 2x512MB DDR400 CL2,2,2,5, Geforce 6800Ultra 92.91, onboard audio, XP-SP3

Skyscraper wrote:

Here is my Gallatin 3.4 period correct system benchmarked with Doom3 v1.0. This is the best non overclocked period correct P4 result so far, only 0.4 FPS from the dual Opteron! 😀

1024*768 Ultra: 105.6 FPS

P4 EE "Gallatin" 3.4. Asus P4C800 Deluxe v2.0. 2x512MB DDR400 CL2,2,2,5, Geforce 6800Ultra 92.91, onboard audio, XP-SP3

Really nice! Just gotta get my hands on Doom 3 v1.0, I guess.

I'm still scratching my head over the performance differential here, since 200 MHz more on the CPU isn't that much when it's the same architecture. You've got half the RAM, but tighter timings. It's the sort of thing I can't really diagnose without another 875P board to play with, and I was hoping I'd just need this one board due to the ISA slot it offers.

Or maybe I should just reinstall XP altogether and see if anything drastically different happens between the two installations.

NamelessPlayer wrote:

Really nice! Just gotta get my hands on Doom 3 v1.0, I guess. […]
Show full quote

Skyscraper wrote:

Here is my Gallatin 3.4 period correct system benchmarked with Doom3 v1.0. This is the best non overclocked period correct P4 result so far, only 0.4 FPS from the dual Opteron! 😀

1024*768 Ultra: 105.6 FPS

P4 EE "Gallatin" 3.4. Asus P4C800 Deluxe v2.0. 2x512MB DDR400 CL2,2,2,5, Geforce 6800Ultra 92.91, onboard audio, XP-SP3

Really nice! Just gotta get my hands on Doom 3 v1.0, I guess.

I'm still scratching my head over the performance differential here, since 200 MHz more on the CPU isn't that much when it's the same architecture. You've got half the RAM, but tighter timings. It's the sort of thing I can't really diagnose without another 875P board to play with, and I was hoping I'd just need this one board due to the ISA slot it offers.

Or maybe I should just reinstall XP altogether and see if anything drastically different happens between the two installations.

~10 FPS of the difference is the Doom3 version. ~5 FPS is the 200 MHz difference in core speed another ~5 FPS is the difference in memory timings.

Do you run with PAT activated or not? That could be a few more FPS accounted for.

Time for some mobile-on-desktop goodness!
This T5600 actually punches pretty hard for a CPU running at only 1.83GHz/667FSB. I'm really liking the i975x chipset--the memory bandwidth is quite a bit higher than my nForce 650i board. It's too bad I can't use quad core CPUs with this board.

Mobile C2D T5600 ES (1.83GHz/667FSB), 4GB DDR2-667 3-3-3-8, GeForce 9800GT (341.44), AOpen i975xa-YDG (i975x), onboard audio, Win7 x64

Compared to some other CPUs I've tested:

C2D e6600 @ 2.4GHz (Win7): 172.8
Opty 185 @ 3.13GHz (Win7): 172.1
Pentium D 935 @ 3.2GHz (Win7) 114,3
Pentium M @ 2.7GHz (WinXP) - 144.2

Standard Def Steve wrote:

Time for some mobile-on-desktop goodness! This T5600 actually punches pretty hard for a CPU running at only 1.83GHz/667FSB. I'm […]
Show full quote

Time for some mobile-on-desktop goodness!
This T5600 actually punches pretty hard for a CPU running at only 1.83GHz/667FSB. I'm really liking the i975x chipset--the memory bandwidth is quite a bit higher than my nForce 650i board. It's too bad I can't use quad core CPUs with this board.

Mobile C2D T5600 ES (1.83GHz/667FSB), 4GB DDR2-667 3-3-3-8, GeForce 9800GT (341.44), AOpen i975xa-YDG (i975x), onboard audio, Win7 x64

Compared to some other CPUs I've tested:

C2D e6600 @ 2.4GHz (Win7): 172.8
Opty 185 @ 3.13GHz (Win7): 172.1
Pentium D 935 @ 3.2GHz (Win7) 114,3
Pentium M @ 2.7GHz (WinXP) - 144.2

I do also like the i975X chipset, diddnt AOpen release a BIOS with quad CPU support? Asus i975X board do support every socket 775 quad there is, even the socket 771 ones is supposedly working with the well known mod 😀

Skyscraper wrote:

Standard Def Steve wrote:

Time for some mobile-on-desktop goodness! This T5600 actually punches pretty hard for a CPU running at only 1.83GHz/667FSB. I'm […]
Show full quote

Time for some mobile-on-desktop goodness!
This T5600 actually punches pretty hard for a CPU running at only 1.83GHz/667FSB. I'm really liking the i975x chipset--the memory bandwidth is quite a bit higher than my nForce 650i board. It's too bad I can't use quad core CPUs with this board.

Mobile C2D T5600 ES (1.83GHz/667FSB), 4GB DDR2-667 3-3-3-8, GeForce 9800GT (341.44), AOpen i975xa-YDG (i975x), onboard audio, Win7 x64

Compared to some other CPUs I've tested:

C2D e6600 @ 2.4GHz (Win7): 172.8
Opty 185 @ 3.13GHz (Win7): 172.1
Pentium D 935 @ 3.2GHz (Win7) 114,3
Pentium M @ 2.7GHz (WinXP) - 144.2

I do also like the i975X chipset, diddnt AOpen release a BIOS with quad CPU support? Asus i975X board do support every socket 775 quad there is, even the socket 771 ones is supposedly working with the well known mod 😀

This board only takes Socket M mobile CPUs--specifically Core Duo (Yonah) and Core 2 Duo (Merom only; all of the mobile Quads were Socket P and will not work).

Standard Def Steve wrote:

This isn't a 775 board. It only takes Socket M mobile CPUs--specifically Core Duo (Yonah) and Core 2 Duo (Merom only; Penryn is Socket P and will not work).

Ah, I thought you were using some fancy socket adapter 😀. I dont even know if there is such a thing as a s479 CPU ---> s775 motherboard adapter, if not someone should build one!

Skyscraper wrote:

NamelessPlayer wrote:

Really nice! Just gotta get my hands on Doom 3 v1.0, I guess. […]
Show full quote

Skyscraper wrote:

Here is my Gallatin 3.4 period correct system benchmarked with Doom3 v1.0. This is the best non overclocked period correct P4 result so far, only 0.4 FPS from the dual Opteron! 😀

1024*768 Ultra: 105.6 FPS

P4 EE "Gallatin" 3.4. Asus P4C800 Deluxe v2.0. 2x512MB DDR400 CL2,2,2,5, Geforce 6800Ultra 92.91, onboard audio, XP-SP3

Really nice! Just gotta get my hands on Doom 3 v1.0, I guess.

I'm still scratching my head over the performance differential here, since 200 MHz more on the CPU isn't that much when it's the same architecture. You've got half the RAM, but tighter timings. It's the sort of thing I can't really diagnose without another 875P board to play with, and I was hoping I'd just need this one board due to the ISA slot it offers.

Or maybe I should just reinstall XP altogether and see if anything drastically different happens between the two installations.

~10 FPS of the difference is the Doom3 version. ~5 FPS is the 200 MHz difference in core speed another ~5 FPS is the difference in memory timings.

Do you run with PAT activated or not? That could be a few more FPS accounted for.

Everest says PAT's enabled, though the BC875PLG doesn't actually give me a BIOS option to enable or disable it to begin with. This thing needs a better BIOS.

I might tweak the memory timings a little more, but I know I can't bring it down to 2-2-2-6 instead of 2.5-2-2-6 without introducing instability. The DIMMs might just be too dense for it, or the motherboard can't handle it where another just might.

Neither can I bump up the speed 200 MHz since there aren't any multiplier or FSB adjustments in the BIOS (something with the manual curiously mentions, but is completely absent in practice).

Since the biggest difference lies in the Doom 3 version used, however, why not start there?

Despite being an engineering sample, the mobile T5600 overclocks quite well. The performance scaling is just awesome, too. I nearly doubled the framerate by overclocking it to 2.93GHz. And despite only having 2MB of L2, IPC is much greater than the desktop e6600 with 4MB L2. In fact, this is dangerously close to the Q6700 at 3.33GHz. I believe the superior memory performance of the i975x (and higher DRAM frequency) is at play here.
Oh, and I tossed in one of my GTX 560s just for good measure. 😊

Mobile C2D T5600 ES (2.93GHz/1066FSB), 4GB DDR2 @ 887MHz 4-4-4-12, GTX 560 (353.06), AOpen i975xa-YDG (i975x), onboard audio, Win7 x64

So how does this compare with overclocked desktop CPUs on an nForce 650i motherboard?
Pentium E2220 @ 3.2GHz/1066 = 137.6
C2D E6600 @ 3.0GHz/1333 = 180.6
Quad Q6700 @ 3.33/1333 = 265.4

I'm not too sure why the Q6700 was so much faster than the E6600 (the game only uses 2 cores), but whatever...I'm certainly glad that it was! 😀

is it possible to run doom 3 without SSE?
I only have 1.31 and it wont work on my Pentium II

How's about a first-gen i7 based laptop?

Normal clock speed for the 720QM is 1.6GHz, but I saw it boosting to around 2.4GHz during the benchmark.

HP Envy 17, Mobile i7-720QM (1.6GHZ w/ Turbo), 8GB DDR3-1333 9-9-9-24-1T, Mobility Radeon HD 5850, onboard audio, Win7 x64