Doom 3 Timedemo
Oct 12, 2012 Re: Doom 3 timedemo shootout with period correct hardware. By Skyscraper » 2014-9-27 @ 08:42 I have not even loaded up.
Buckeye wrote:I'd like to post my numbers but I'm a 'rookie' when it comes to this stuff. Can someone help me with these questions: 1. What screenshot utility is everyone using?
Tried using Shotty but it won't work with WinXp. How do you get all the different screenshots to show up on the desktop at one time. At the end of the timedemo, can't seem to minimize it to add CPUZ, GPUZ info and etc. Print Screen ('Print Scrn' key on my keyboard) + MS Paint (saved file as.png or.jpg) Ad.
Atl + Tab in game (or Ctrl + Alt + Del and NOT clicking on Doom 3 window ). To go fullscreen again, use 'Atl + Enter' shortcut. Oldbie Posts: 807 Joined: 2016-1-19 @ 11:06. Here comes my first Socket A score. It's funny that this system with an AMD 760 chipset (released in October 2000) motherboard sporting a 2133 MHz Athlon XP-M 2800+ running at a 133 Mz FSB with 1GB single channal DRR266 memory which isn't even interleaved is the fastest non overclocked period correct Socket A system so far.
I think we need more Socket A scores! I think 70+ FPS should be possible with a non overclocked period correct system. 55.9 FPS @ 1024 Ultra Athlon XP-M 2800+ (16x133), Abit KG7 (AMD 760 chipset released Oct 2000), 1GB DDR266 2225, GF6800Ultra 92.91, Audigy, XP-SP3. Asus P4GPL-X (i915), CT-479, Pentium-M 780 (highest stock clocked Dothan) acceptable for stock period-correct? IIRC, PM-780 was released early 2005, but technically the i915 (that is acceptable) is also released during the same time-period. Edit: According to Wiki, the PM-780 was released June-2005, so that's probably too late. The PM-770 was released Jan 2005, which is right on the cut-off line.
In fact, looks like all the native 533 FSB Dothans were released in Jan 2005. Looks like I'd have to work with a 400 FSB Dothan. Oldbie Posts: 725 Joined: 2017-3-07 @ 03:32 Location: USA. Dexvx wrote:Asus P4GPL-X (i915), CT-479, Pentium-M 780 (highest stock clocked Dothan) acceptable for stock period-correct? IIRC, PM-780 was released early 2005, but technically the i915 (that is acceptable) is also released during the same time-period.
Edit: According to Wiki, the PM-780 was released June-2005, so that's probably too late. The PM-770 was released Jan 2005, which is right on the cut-off line. In fact, looks like all the native 533 FSB Dothans were released in Jan 2005. Looks like I'd have to work with a 400 FSB Dothan.
Pentum M Dothan CPUs wth 400 MHz FSB runnng on any version of the i915 chipset is fine even though the Pentum M 'native' version of the i915 chipset is from early 2005. I would guess that you will get much better scores usng an Asus P4C800 i875P motherboard though. Vetz wrote:I've been thinking the same thing, but the CT-479 adapter wasnt released until after the cutoff point. Let's pretend that he made the adapter himself. A CPU from 2004 (or earlier) running on a motherboard with a chipset from 2004 (or earlier) is enough in my book.
I do not think a stock Pentium M 765 is likely to be able to top the stock settings period correct list anyhow but if it does it would give me a good reason to rebench my Socket 939 systems using Doom 3 v1.0. There are i865 chipset motherboards with Socket 479, they are rare and mostly suck though. Vetz wrote:I've been thinking the same thing, but the CT-479 adapter wasnt released until after the cutoff point.
Let's pretend that he made the adapter himself. A CPU from 2004 (or earlier) running on a motherboard with a chipset from 2004 (or earlier) is enough in my book. I do not think a stock Pentium M 765 is likely to be able to top the stock settings period correct list anyhow but if it does it would give me a good reason to rebench my Socket 939 systems using Doom 3 v1.0. PM-765 (2.1 GHz) at stock will not top due to clock speed deficit compared to FX Claw/Sledgehammer. But if we go for technicals, the San Diego/Venice was released mid-2005. Hence I asked specifically for rules.
Regardless, I'll be looking for a PM-765 to post results hopefully. Overclocking 400 FSB Dothan is more difficult due to higher multiplier and the fact you need to scour for a C stepping chip. I have a 915PL, which I think is slower than P4C800/P4P800. But then again, the advantage of the 915PL is that it is PCI-e, so I can use my 6800U/X850XT PE. Finding AGP versions of those are a PITA. Also the first post wasn't that clear. Timedemo doesn't work on Doom3 demo?
So we'd have to buy a hard copy of Doom3 (to back-versioning). Is that correct? Oldbie Posts: 725 Joined: 2017-3-07 @ 03:32 Location: USA. Vetz wrote:I've been thinking the same thing, but the CT-479 adapter wasnt released until after the cutoff point. Let's pretend that he made the adapter himself. A CPU from 2004 (or earlier) running on a motherboard with a chipset from 2004 (or earlier) is enough in my book. I do not think a stock Pentium M 765 is likely to be able to top the stock settings period correct list anyhow but if it does it would give me a good reason to rebench my Socket 939 systems using Doom 3 v1.0.
PM-765 (2.1 GHz) at stock will not top due to clock speed deficit compared to FX Claw/Sledgehammer. But if we go for technicals, the San Diego/Venice was released mid-2005. Hence I asked specifically for rules. Regardless, I'll be looking for a PM-765 to post results hopefully. Overclocking 400 FSB Dothan is more difficult due to higher multiplier and the fact you need to scour for a C stepping chip. I have a 915PL, which I think is slower than P4C800/P4P800.
But then again, the advantage of the 915PL is that it is PCI-e, so I can use my 6800U/X850XT PE. Finding AGP versions of those are a PITA. Also the first post wasn't that clear. Timedemo doesn't work on Doom3 demo? So we'd have to buy a hard copy of Doom3 (to back-versioning).
Doom 3 Full Movie
Is that correct? The Doom 3 demo will sadly not run the Timedemo and the new versions sold online are a little bit slower than Doom 3 v1.0 from the orginal disc. If I did not own a Doom 3 disc and could not find one I would buy the Steam version and then acquire the 1.0 version 'by other means'. When it comes to San Diego (February 2005) and Venice they are only accepted when they are CPU models. that also exist with Claw/Sledgehammer cores released in 2003 and 2004. The different (1MB) s939 cores performance in Doom 3 are the same (I have tested both FX55 versions using the same system) so it doesn't affect the non overclocked list at all. It seemed practical to not care about core revisions and just focus on CPU models., sockets.
and chipsets. The overclocked list does get a bit skewed but the difference in max overclock is only 3-4% comparing a good Clawhammer to a good San Diego.
Serial number search. Sage Software, Inc.
If we start caring about AMD core revisions within the same CPU models it would make sense to also care about Intel steppings as they affect overclocking just as much. As long as we don't use sub zero phase change cooling we will not get better scores than the 'Xtremesystems crowd' did back in year 2004 anyhow.Simulating a FX55 for the period correct list by running a single Opteron 252 on Socket 940 or an Opteron 152 on Socket 939 would be okey but will be noted as 'Opteron 152 (FX55)'. Even using a faster single core Socket 939/940 CPU like the FX57 to simulate the FX55 for the stock list would be okey as long as 13x200 MHz is used, the performance will be identical to the real thing.No AM2 with nForce4 + Athlon 64 4000+.When it comes to chipsets we have not been very strict, nForce4 Pro/32 is okey as the only real benefit over nForce4 is non gimped SLI and SLI is not allowed in the period correct list anyhow. Likewise I think all versions of the Intel 915 chipset should be okey as they are basicly the same when it comes to both performance and overclockability. Another thing that is allowed is using a Geforce 6800 GT to simulate a Geforce 6800 Ultra or even an 'Ultra+' or 'Extreme' version as soon as we have a proper score on the list using one of those with known GPU and memory clocks. Nvidia them selves did not talk about an 'Ultra Extreme' until early 2005 and they never released an official spec but higher clocked 6800 versions existed 2004 so we will go with whatever someone manages to find.
The rules aren't very strict but things like the Prescott 2M, Socket AM2 and the Dothan with 533 MHz FSB are stretching the period correct definition a bit too far. Many of these things have been covered before in this thread but I can understand that it's a bit too much to sift through. I will clearify stuff in the first post.