The Official DOOM 3 [H]ardware Guide thread.

paraiso36 said:
This is specially for Mike Ryan

Good Day,

I was the one who started the doubt about the test results with an AMD Xp 3200 and P4 3.0 GHZ with the ATI 9800pro 128 MB..First I need to say that I appreciate Mike intention to ask the question again, because until now nobody has been able to answer it..As we are all fanatic of this gaming world and most of all arent expert on the subject, maybe there isnt a person here who knows the answer about why the AMD 3200 deserves a High Quality and P4 3.0 only Medium quality..May be the person (sorry I am writing with my IPAQ so it is difficult to browse other pages ) who answered that it was as it was...is right...But I dont want to offend but maybe this doubt is just because I always try to explain something that called greatly my attention. Someone answered Mike that if he knew about CPU architecture he would get the answer.Please I ask him to teach us with his knowledge, because by now I am at the starting point...A few days ago I bought my P4 3.2 GHZ CPU so I made lot of researchs about P4 3.0, 3.2 and the AMD rivals.From that I learnt,

1. AMD Xp 3200 isnt better than P4 3.0 in terms of performance. Overall is slighty behind it in all areas of tests including games.
2. Amd begin to hit Intel with his AMD 64...P4 3.0 races against AMD 64 3200 and sometimes AMD 64 3000. P4 3.2 races against AMD 64 3400 and sometimes AMD 64 3200..The race is very matched but I admit that in gaming AMD 64 wins and sometimes it isnt a marginal victory..On the other hand Intel P4 wins almost all the tests about video and audio processing.

From the points above I still have my doubts about the result of the test even more if you consider other factors,

1. First as I mention the AMD Xp 3200 isnt better than P4 3.0..If anyone can prove that it is better please tell me...But even suppose they are in a perfect draw,
2. AMD 3200 has only 512 MB of RAM and P4 3.0 1 GB.Yes I know someone can tell me that Doom 3 is not dependent of Ram, but I havent seen a single system runs better with 512 MB ran instead of 1 GB.
3. The result not only gives AMD Xp 3200 High Quality instead of medium one given to the P4 3.0..It also gives 8xAF to the AMD and NO AF to P4...

Really there is something strange about it...or something I dont understand..Maybe we will have to wait until the game is released but my doubt still exist and nobody has been able to give a explanatory answer..If the AMD were a 64 model I would understand it..but not with a XP..I think it is to much punisment to P4 3.0 and I want to understand it...Thanks again and specially to Mike because I still have the same doubt.


I have to wonder what benches you've based your CPU comparisons on. I pray to God it's not THG...
 
Hi Guys,

probably shoudn't be asking questions I know the answers to but I was wondering if anyone knows of the possibility of running Doom 3 on Geforce 2 MX, 64 Mb or am I simply s**t out of luck :(

Thanks,
Tom
 
Hi Guys,

probably shoudn't be asking questions I know the answers to but I was wondering if anyone knows of the possibility of running Doom 3 on Geforce 2 MX, 64 Mb or am I simply s**t out of luck :(

Thanks,
Tom

you sir, are SOL.. imho.
 
davidj said:
What The Wall Street Journal is to the financial world, HardOCP.com is the same for the PC enthusiast/gamer.

I am just floored by all of the work and results that have been produced. You have prepared thousands of gamers for this Special event..

Plus all of this is free

Now to upgrade :D

Thank you again!

davidj

Well then it's not technically "free" now is it? :p
 
Tim said:
Well then it's not technically "free" now is it? :p
Well, many people going into the article thought that they would have a hard time with a 9500 Pro, which is simply not the case. Then again, I game at 10x7 with AA and AF turned off in everything. ;)
 
Ardrid said:
I have to wonder what benches you've based your CPU comparisons on. I pray to God it's not THG...


yeah, its a good thing an overlclocked XP @ 2.5Ghz 476FSB like mine wasnt used or that guy would have had a heart attack, hehe. I guess his "research" was only done at pro-Intel sites. :rolleyes:
 
geediu said:
check out http://www.doomsdayhq.com/screenshots.php?num=98.........? as close as you can get?

google jdoom
It runs on the Doomsday engine, which is just a Win32 port of the DOOM engine. While impressive compared to the original, I really want to see them port the classics to the monstrous D3 engine. Was it Gabe Newell that said VALVe was porting the original HL over to the HL2 engine (or was it just CS)?
 
lorcani said:
Well, many people going into the article thought that they would have a hard time with a 9500 Pro, which is simply not the case. Then again, I game at 10x7 with AA and AF turned off in everything. ;)

I was just joking about the fact that he felt the need to upgrade after the reading the article, which means that he's spending money because of the article.
 
Tim said:
I was just joking about the fact that he felt the need to upgrade after the reading the article, which means that he's spending money because of the article.
I know, but most people here, if they really wanted to, could play Doom 3 on all their relatively new systems, including the ones that don't have 6800GTs. I'd probably still upgrade, though. :D

That was the coolest thing about the guide, IMO: you guys proved that Doom 3 would run on most hard-ware produced in the last 3 years, which is amazing.
 
lorcani said:
I know, but most people here, if they really wanted to, could play Doom 3 on all their relatively new systems, including the ones that don't have 6800GTs. I'd probably still upgrade, though. :D

That was the coolest thing about the guide, IMO: you guys proved that Doom 3 would run on most hard-ware produced in the last 3 years, which is amazing.

Really Kyle and Brent deserve all the kudos. They're the ones that spent all that time at id working on the benchmarks. All I do is come in here and try not to screw stuff up on the weekends. :p
 
Tim said:
Really Kyle and Brent deserve all the kudos. They're the ones that spent all that time at id working on the benchmarks.
Yes, it is incredible that a webmaster and an editor care so much about their fanbase that they spent more than 60 hours huddled around a computer screen, rigorously benchmarking, with very little thanks. You guys deserve a medal. :)
 
lorcani said:
It runs on the Doomsday engine, which is just a Win32 port of the DOOM engine. While impressive compared to the original, I really want to see them port the classics to the monstrous D3 engine. Was it Gabe Newell that said VALVe was porting the original HL over to the HL2 engine (or was it just CS)?
They started by porting hl1 to the source engine, then they worked on cs, and they said that both tfc and dod are coming for source as well.
 
what i would really love to see along with alot of other people
is how doom3 runs on mobile platforms

many many many of us in the [H] group use lappy's or our lappy is our only system

i persoanly use a P-M1.7 (2mb cache)system with 1GB mem and an ATI radeon 9700 mobile 128mb
with a 5400rpm 80gb drive

i know the performance is going to fall somewhere between a 8500 and a 9600XT but i would love to know if i'm going to be able to run smothly while using the eye candy at 800x600

the other thing that was missing from the review was any testing of ATI's SE cards

the 9800SE and 9600SE cards are very popular in the highlevel budget systems, i know from personal experience that a 9800SE 256bit ($130) will outperform a 9600XT($140) in UT2004 and far cry but i would love to know how well it works with doom3

[H] itself has said although the latest and greatest from ati and Nvid will make news for the company and gets its name out there the real fight for supremicy will be in the budget market where companies are made and lost. i feal that although its cool to know how a 6800 is going to preform with doom3 the majority of us out there desperatly want to know how its going to go with out 9500pro, or 9800SE or 5700pro (any card in the $100-200 range)

i really think [H] should go back and add some systems i into the article

a very common system among my friends is a
N-force 2 mobo with 512mb dual (2x256) running a barton XP2500
and using a saphire radeon 9800SE 256bit 128mb

and for lappys
how well does the mobile nvid 5700 and the ati 9600/9700 work using
P-m systems vs p-4m vs amd athlon64mobile

if anything the article has opened more questions that its closed
i now know that my system will run doom3
but i dont know where between running marginaly and running well the systems are going to place
 
Geta said:
yeah, its a good thing an overlclocked XP @ 2.5Ghz 476FSB like mine wasnt used or that guy would have had a heart attack, hehe. I guess his "research" was only done at pro-Intel sites. :rolleyes:


you mean 239 dont you....sheesh...

dont add the multiplyers in makes you look like a noob.
 
IxGOxSOLO said:
Will we be seeing any updates to the guide, or maybe smaller guides based on updates or things you wanted to do but happened to miss, when the game is released? Will this start to be a trend for [H] regarding major game releases in the future? All I have to say is you guys rock, and this is one awesome guide that I will find myself going over for quite some time. This Doom 3 guide is the kind of work that separates the [H] from everybody else and puts [H]ardOCP ahead of the pack.
Thanks for the kind words, but there were too many to quote. (I think that is a first.) :)

Yes, you will be seeing guide updates or at least more articles linked in with more content in the same vein.

We contacted Valve last year about doing the same thing but do not have anything solid with them currently, although we are surely doing our best to secure a guide for Half Life 2.
 
22killer said:
AAHHH Nooooo, please kyle or anyone else,

I have a Geforce3. You guys totally forgot to bench it but listed it like crazy?!!?

Can you please go back and see what the benches were, just incase you forgot to add them to the website when you were puting it up?

Please man i need to know how good it does, compared to the G4mx. How much beter does it do? I only seen a picture of it thats it.
You know, I broght two and we did screens, but I guess it slipped through the cracks for some reason.

The GF3 should do 800x600 on low I would GUESS.
 
Blue Damned said:
Nice guide. But still, you guys haven't described how it feels to play Doom 3 in Matrox's Surround Gaming. That must be mind blowing. Even more then whatever better FPS can give. Please include a little Parhelia.
Have you ever seen a slide show at the IMAX?
 
aces170 said:
Need I say more Hardocp is simply the best. Top in my fav list.
! question though How come you recommend Radeon 9800 pro on high settings when it drops to 6 fps in minimum. Also would the game behave diffrent in the highest mode of difficulty?

And I am planning to buy an A64 newcastle proccy Socket 754, with Radeon 9800 pro how will the game run on that will it be similar to 3000+ barton? Also can you suggest me a board which will help me to oc the proccy ( I am proably out of topic here, kindly ignore me if thats the case)
We based on suggestions for res and texture settings on PLAYING THE GAME, not looking at the data. Apparently that 6fps did not hamper gameplay and was likely in a cut scene.

As for your OC stuff, there is a whole article linked in the guide for that.
 
ne0-reloaded said:
PS hey Mr. Kyle Bennet, what system do you have, just wanna know if yours is the Ultimate one or not?
I will as soon as I pull this 9800Pro 256 out of here. :)
 
Kyle....*bows* I'm not worthy...!


PLEASE answer this:

Do you *think* my current FX-51, 2Gb 3200 with a 6800ultra OC (450mhz) will be able to run doom3 at Ultra Quality @ a *higher res* than 1600x1200 (like your "ultimate system" can)???


An answer would be much appreciated....Sorry for another hardware Q - as if you haven't done enough already. :D


* It's NOT an OCed system btw.
 
all these questions will be answered once we finally do our own friggin installs ,, aiieeee the wait is KILLING ME!!!!!

oops looks like they put up the image by mistake on the [H] FTP server..dlling now..haha..
 
Ok maybe a stupid question or its already asked (31 pages are alot I read 15 so far)
But was the testlevel that was used for the benchmarks one of the most heavy levels (to render I mean) of the game or was it a simple level. Because I own a 2,26 ghz and a ti4600 I wonder If that is the most heavy level to render. If it is I think I might be able to pull of 800*600 in high quality (I very very hope) instead of medium what you advised in the great guide
 
From the Guide:
We sat down with id Software’s Marty Stratton to decide on some levels that would test out the extremes of gameplay and visuals. From that discussion, we chose the “Enpro” level as the focus of our gameplay testing.
I hope that gives you some insight Kamikaas.
 
lopoetve said:
Yeah, I know. I gamed for a long time on budget systems. I was thinking of the older celerons, that were total crap... Or the new ones. And I'm a college student as well, so I know how that goes.
Older Celerons total crap? I'll have you know sir, that the C300A(450) was a god amoung CPU's...at the time :)
 
bob63 said:
Older Celerons total crap? I'll have you know sir, that the C300A(450) was a god amoung CPU's...at the time :)

Well, he may be talking about the original Celerons that were just castrated PII's with no level 2 cache. They could overclock like mad, but the performance still blew because of the lack of level 2 cache.
 
Tim said:
Well, he may be talking about the original Celerons that were just castrated PII's with no level 2 cache. They could overclock like mad, but the performance still blew because of the lack of level 2 cache.


now hold on a minute my 300A ran at 466 and outbenched my buddies 450 pIII in everything possible test....quake two, 3dmark, ut, we tried em all at the time..

it killed him too cuz he paid a ton more for it of course...the FSB more than made up for lack of cache...
 
theelviscerator said:
now hold on a minute my 300A ran at 466 and outbenched my buddies 450 pIII in everything possible test....quake two, 3dmark, ut, we tried em all at the time..

it killed him too cuz he paid a ton more for it of course...the FSB more than made up for lack of cache...

The 300A had 128K of full speed level 2 cache. I was talking about the original Celerons that Intel launched that had no cache whatsoever.
 
Looks like I'll be able to pull it off with my box (see sig for specs) ... thanks, Kyle :cool:
 
Tim said:
The 300A had 128K of full speed level 2 cache. I was talking about the original Celerons that Intel launched that had no cache whatsoever.

you are correct those sucked badly, put them outta mind entirely...

my 466 and my voodoo3 at one time was the stuff for quake2, got 110 fps at 1024 with game maxed....

enough to get me a top 25 railgun on the clq for a time..
 
my current system is:
amd xp 2400 (266fsb)
768mb pc2100 memory (budget)
ati 9500 pro

Im going to upgrade to
amd xp 3000 (333 fsb)
1 gig pc2700 (kingston hyperx)
6800 ultra

my mobo is not very good for o'cing, and maxes out at that fsb.
i wanted to know how much performance i would lose being at a 333fsb with pc2700 memory (as opposed to 400fsb and pc3200). Would more memory, or faster memory be better?

ps. THAT was the most extensive performance guide i have ever seen. Very nice work
 
i was a bit pissed off that they did not use a 5900xt so sorry for this.
My specs
p4 2.4ghz
768mb pc2700
128mb 5900xt

How will it run?
 
Geta said:
yeah, its a good thing an overlclocked XP @ 2.5Ghz 476FSB like mine wasnt used or that guy would have had a heart attack, hehe. I guess his "research" was only done at pro-Intel sites. :rolleyes:
Lol,

My "research" was more simple than that,

1.- www.google.com
2.- In search section I typed or " amd xp 3200" or " amd xp 3200 versus pentium 4 3.0"

You can imagine ....lots of pages appeared...I began reading them...my conclusion was what I mentioned....( or my mouse is allied with Intel and all clicks were made "on Intels pages" ;) ...)

Jokes aside, see is believing...I encourage you to show me pages to help me "get out of my error" ...thanks
 
lopoetve said:
If you've got $700 you might get a 512mb card. GDDR yields as they are, it's gonna be a long time.

No reason for it, doom3 @ high is 99% of ultra.

Good Point.
My system will work for now so I am just going to give it 2 or 3 months and see what happens.
 
theelviscerator said:
you mean 239 dont you....sheesh...

dont add the multiplyers in makes you look like a noob.


uh, sheesh right back at ya. what are you talking about, noob, multiplier...? Hear this, no I mean 476Mhz FSB. This is the correct way to put FSB, DDR. If I was using SDRAM @ 238Mhz, then yes. Were talking DDR, you know? 476Mhz is my FSB. Looks like your the boob, intel boi.
 
foolInc said:
Using ddr2/i925 seems to affect some of the cards strangely. Generally when ddr2 is used it causes an improvement for the x800xt card while causing a slowdown for the 6800 cards, the best example being for the 6800 gt on a p4 3.6:

ddr1/i915:
http://www2.hardocp.com/1090787143TO62AMOEWX_11_2_l.gif

ddr2/i925:
http://www2.hardocp.com/1090787143TO62AMOEWX_14_2_l.gif

and the x800xt on the same systems:

ddr1/i915:
http://www2.hardocp.com/1090787143TO62AMOEWX_11_3_l.gif

ddr2/i925:
http://www2.hardocp.com/1090787143TO62AMOEWX_14_3_l.gif

so my question is, am i missing something obvious or is there some sort of weird conflict/optimisation at work? oh and I think at ddr2 the XT gets its heighest IQ/FR of any system tested, this seems odd...

Oh yeah thanks for the hardware guide, lots of info, being a greedy bastard I would have liked more but given the time limits it was great. btw im thinking i'll be able to get 800*600 at MQ on my athlon 1.33, 512 pc2700 and ti4200 64mb, i was hoping to get 10*7 but i am thinking even 8*6 might be unrealistic :confused:

You missed one thing.... AA

6800GT
915 - 1600x1200 2xAA 8xAF - Avg 51.9
925 - 1600x1200 NoAA 8xAF - Avg 55.9

X800XT
915 - 1600x1200 NoAA 8xAF- Avg 50.7
925 - 1600x1200 2xAA 8xAF - Avg 46.3

So, there's no way to know for sure whether DDR 2 causes a performance hit/gain unless the tests had been run at the same settings.
 
I have an athlonxp 2200 with a gf4ti4200 128mb. Will i see a lot of difference both in IQ and performance with a 9600xt or a 9600pro or should i wait and save for a better card?
What's better? a 9600xt 128mb or a 9600pro 256mb? Thanks
 
uni said:
I have an athlonxp 2200 with a gf4ti4200 128mb. Will i see a lot of difference both in IQ and performance with a 9600xt or a 9600pro or should i wait and save for a better card?
What's better? a 9600xt 128mb or a 9600pro 256mb? Thanks

go for a 6800 LE when its avialable,
 
Hi, I'm a nubbie.

I just got a quick question for all you smart people.

I've got a p4 2.4 on an asus mobo w/ 533 FSB. No overclocking. A G4 Ti 4600 128.

AND

I've got 1GB of 1044Mhz Rambus.

How does rambus fair with games? Will it even matter w/ my vid card being a bit older?


Thanks Guys,
Smotpoker
 
Back
Top