Doom 3: CPU Battlegrounds - Guess who wins it ;)

[alexruiz]

Originally posted by: mikecel79

HardOCP reported something similar crowning the FX-53 as the ultimate CPU for D3. Now what AMD needs is to fully support DDR500 or better.... no need for DDR-II yet.

DDR500 is not going to be a JEDEC standard. They officially stopped at DDR400. Very slim chance that AMD will support that speed unless it's standardized. There wasn't going to be a DDR400 but Intel pushed for it so they could have dual channel DDR400 to match the speeds of their 800Mhz FSB. Besides AMD would need to redesign the memory controller which would mean a new verison of the K8 chip family.

You are right in the fact that JEDEC certified only DDR400, but running DDR500 natively requires only a BIOS update IF the memory controller is capable to reach and sustain the speed. I am assuming that all the CG version K8s are capable to run at that memory speed.

 

[fkloster]

Good Times

 

[iwantanewcomputer]

Originally posted by: viivo
You have to feel at least some pity for the new Celeron. Poor little thing.

are you serious? when is the last time anybody wanted to see a celeron alive? Why do you think people around here call it "celery"?

Quite happy with my a64

 

[imported_NoGodForMe]

I thought it was interesting seeing the low end CPUs running the 6800 Ultra.
Many have asked if it is worth upgrading their video card to a 6800GT, and then buying a better CPU/MB/Memory in the future. The answer is clearly Yes.
If someone pays the $400 and puts a 6800GT into an older computer, it may be CPU limited, but it will be much faster than what they have, and they will see the eye candy.

 

[Cobalt]

Where did they get the S939 FX-51?

 

[RaynorWolfcastle]

Originally posted by: alexruiz

Originally posted by: RaynorWolfcastle
Was there any doubt that AMD would be winning these benchmarks? It's been established for awhile already that the K8 architecture is more suited to gaming than Netburst.

I can't believe the replies in this thread and how attached people are to a specific company; maybe someone should make a thread showing off some Lightwave/media encoding benchmarks so that Intel fanboys can show off their huge e-penis. :roll:

edit: Just so that the fanboys don't flame me, I own computers with both AMD and Intel CPUs; I ambivalent towards the companies and buy what processor suits my needs best.

StrangerGuy

Now the A64 beats the P4 in every benchmark, including media encoding.

Really? Kind of like in this benchmark here? Text
or how about Lightwave and 3DS Max? Text

Try to encode whatever you want in whatever format you feel using a commercial application (Ulead VideoStudio or Roxio videowave) and tell me what happens.... In fact, try to encode to DV (the most used
codec at consumer level because of digital camcorders) using Ulead and tell me who wins. The applications/codecs/containers where AMD beats intel in video encoding are 2:1 in favor of AMD. Obviously, when most of the internet sites show XMPEG encoding DivX you think the P4 is unstoppable. Do the same DivX encoding in DVD2AVI or virtualdubmod and tell me who wins...... If you encode DivX with XMPEG and MPEG with TMPGEnc, then the P4 is for you. If you use something different then the K8 is your solution.

Don't contribute to the myth that the P4 is better for media encoding because that is not true. It wins if the application if heavily optimized for SSE2/HT. Fortunately, most of the applications focus more on features than optimizations for the P4, so my K8 is unstopabble creating titles, adding transitions or layering clips. I will believe the P4 is better when it wins the majority of video encoding tasks, not just the XMPEG+DiVx or WM9..... But when it gets beaten encoding MPEG2 in CCE or encoding sorenson in quicktime or encoding Xvid in virtualdub then you know that is not the case.


Alex

You don't get it. I don't care if the apps *are* optimized for the P4 or not, I just care about the encoding time. I don't care about having fair benchmarks, I care about getting my encoding done. I don't use CCE, I use tmpenc which is heavily P4 optimized (I don't have the money to spend on CCE, maybe you do). If the apps I needed were all K8 optimized, then I'd use an A64 to do my encoding.

Also, I see you didn't care to comment on the 3DS and Lightwave benchies. I was replying to the post that the K8 is faster in all respects to the P4, which it clearly isn't. Most, yes; all, no.

With the replies on this board, you'd think that by proclaming anything other than that the A64 is the be-all and end-all CPU, you'd just raped their mother. :roll:

 

[RaNDoMMAI]

I have a question

I am using a mobile barton 2400 OC to 2.4ghz(200 x 12)

Should i just consider this like a normal 3200 barton and place it in the same area as the barton 3200 on the chart? Or is there something different because it is a mobile chip? arent moblie chips just very good regular chips?

thx
~RaNDoM

 

[JackHawksmoor]

I'm just glad Prescott trounced Northwood on something meaningful to me. (Since I'm getting one tommorow and all. ) That's Prescott on socket 478 and at "low" clock speeds. Prescott on socket 775 will be even better (I doubt it'll equal an Athlon 64, but at least it'll FINALLY make Prescott and a 925 board a clear performance improvement over Northwood on 875).

 

[DotheDamnTHing]

Originally posted by: mikecel79

Originally posted by: DAPUNISHER
RaynorWolfcastle,

You are exactly right, but your logic is wasted on most the members here. I swear this place is turning into AMDMB v2.0 Intel is in the drivers seat everywhere but in the tech junkie demographic, so their perception is irrelevant, while Intel market share and profit margins most certainly are not.

I do have to say that Intel has some work to do or some day AMD will wake up and start effectively marketing their products costing Intc market&mind share.

So true!

props to raynor for challenging the a64 bias on general hardware

 

[SickBeast]

Originally posted by: RaNDoMMAI
I have a question

I am using a mobile barton 2400 OC to 2.4ghz(200 x 12)

Should i just consider this like a normal 3200 barton and place it in the same area as the barton 3200 on the chart? Or is there something different because it is a mobile chip? arent moblie chips just very good regular chips?

thx
~RaNDoM

It compares very close to the A64 3200+.

 

[SickBeast]

Originally posted by: RaynorWolfcastle

Originally posted by: alexruiz

Originally posted by: RaynorWolfcastle
Was there any doubt that AMD would be winning these benchmarks? It's been established for awhile already that the K8 architecture is more suited to gaming than Netburst.

I can't believe the replies in this thread and how attached people are to a specific company; maybe someone should make a thread showing off some Lightwave/media encoding benchmarks so that Intel fanboys can show off their huge e-penis. :roll:

edit: Just so that the fanboys don't flame me, I own computers with both AMD and Intel CPUs; I ambivalent towards the companies and buy what processor suits my needs best.

StrangerGuy

Now the A64 beats the P4 in every benchmark, including media encoding.

Really? Kind of like in this benchmark here? Text
or how about Lightwave and 3DS Max? Text

Try to encode whatever you want in whatever format you feel using a commercial application (Ulead VideoStudio or Roxio videowave) and tell me what happens.... In fact, try to encode to DV (the most used
codec at consumer level because of digital camcorders) using Ulead and tell me who wins. The applications/codecs/containers where AMD beats intel in video encoding are 2:1 in favor of AMD. Obviously, when most of the internet sites show XMPEG encoding DivX you think the P4 is unstoppable. Do the same DivX encoding in DVD2AVI or virtualdubmod and tell me who wins...... If you encode DivX with XMPEG and MPEG with TMPGEnc, then the P4 is for you. If you use something different then the K8 is your solution.

Don't contribute to the myth that the P4 is better for media encoding because that is not true. It wins if the application if heavily optimized for SSE2/HT. Fortunately, most of the applications focus more on features than optimizations for the P4, so my K8 is unstopabble creating titles, adding transitions or layering clips. I will believe the P4 is better when it wins the majority of video encoding tasks, not just the XMPEG+DiVx or WM9..... But when it gets beaten encoding MPEG2 in CCE or encoding sorenson in quicktime or encoding Xvid in virtualdub then you know that is not the case.


Alex

You don't get it. I don't care if the apps *are* optimized for the P4 or not, I just care about the encoding time. I don't care about having fair benchmarks, I care about getting my encoding done. I don't use CCE, I use tmpenc which is heavily P4 optimized (I don't have the money to spend on CCE, maybe you do). If the apps I needed were all K8 optimized, then I'd use an A64 to do my encoding.

Also, I see you didn't care to comment on the 3DS and Lightwave benchies. I was replying to the post that the K8 is faster in all respects to the P4, which it clearly isn't. Most, yes; all, no.

With the replies on this board, you'd think that by proclaming anything other than that the A64 is the be-all and end-all CPU, you'd just raped their mother. :roll:

Raynor,

The benchmarks you posted are outdated by over a month compared with these.

Read the commentary on workstation performance:

On the whole. we see a pattern of even workstation performance between the top Intel 560 on 775 and the top AMD FX53 on 939. The only exceptions are Intel topping Light as it usually does, and the FX53 topping 3dsMax. The nForce and VIA chipset boards were fairly even in SPECviewperf except for the standout performance of the K8N Neo2. Otherwise, there is little separating the performance of any of these boards in Workstation Performance.

People reccomend A64s like hotcakes on here because as a whole they are the superior processor when compared with the P4, plus they offer future 64-bit support. If the A64 can match the P4 in these 32-bit benchmarks, it will annihilate it once 64-bit apps are around.

I'm sure people were pimping the Celeron 300's back in the day just as much as this if not worse.

 

[Lyfer]

Nice benchmarks! Another Win for Clawhammer!

 

[earthling30]

I'm suprised that the XP 2000+ I have is still being used in these benchies, and look, it's hanging at the bottom just below the new Celeron D that is suppose to be a whole gigaflop faster (1.8 vs. 2.8)! Hmm.






Sigh.........Mine is sooooo outdated.......almost time for an upgrade! Thinking of changing over to an XP-Mobile and adding another 512 of ram, but I've never overclocked! Oh well!

Don't get me wrong I own AMD and Intel machines all the same. Just all depends on what I'm doing, but I'm a sucker when it comes to buying a sleeper!

 

[earthling30]

Originally posted by: DAPUNISHER
RaynorWolfcastle,

You are exactly right, but your logic is wasted on most the members here. I swear this place is turning into AMDMB v2.0 Intel is in the drivers seat everywhere but in the tech junkie demographic, so their perception is irrelevant, while Intel market share and profit margins most certainly are not.

I do have to say that Intel has some work to do or some day AMD will wake up and start effectively marketing their products costing Intc market&mind share.

Outstanding Point!

PorBleemo, now that's a good quote to have as a sig from DAPUNISHER.

Originally posted by: DAPUNISHER
You are new so I'll clue you in. There are members here with elite status that do nothing but spam in off topic. Rank means little here consequently. I've seen lifers asking for simple hardware advice because they are clueless n00bs about tech, they just know how to post OMGWTFBBQ really well

 

[RaynorWolfcastle]

Originally posted by: SickBeast

Originally posted by: RaynorWolfcastle
You don't get it. I don't care if the apps *are* optimized for the P4 or not, I just care about the encoding time. I don't care about having fair benchmarks, I care about getting my encoding done. I don't use CCE, I use tmpenc which is heavily P4 optimized (I don't have the money to spend on CCE, maybe you do). If the apps I needed were all K8 optimized, then I'd use an A64 to do my encoding.

Also, I see you didn't care to comment on the 3DS and Lightwave benchies. I was replying to the post that the K8 is faster in all respects to the P4, which it clearly isn't. Most, yes; all, no.

With the replies on this board, you'd think that by proclaming anything other than that the A64 is the be-all and end-all CPU, you'd just raped their mother. :roll:

Raynor,

The benchmarks you posted are outdated by over a month compared with these.

Read the commentary on workstation performance:

On the whole. we see a pattern of even workstation performance between the top Intel 560 on 775 and the top AMD FX53 on 939. The only exceptions are Intel topping Light as it usually does, and the FX53 topping 3dsMax. The nForce and VIA chipset boards were fairly even in SPECviewperf except for the standout performance of the K8N Neo2. Otherwise, there is little separating the performance of any of these boards in Workstation Performance.

People reccomend A64s like hotcakes on here because as a whole they are the superior processor when compared with the P4, plus they offer future 64-bit support. If the A64 can match the P4 in these 32-bit benchmarks, it will annihilate it once 64-bit apps are around.

I'm sure people were pimping the Celeron 300's back in the day just as much as this if not worse.

I bolded the section in my comments relating to my point.

Also, the second set of benchies compares the Athlon FX with the P4 560, my comment related to the P4 560 vs the Athlon64s, if you want to compare the FX53, compare it with the P4 EE. I should certainly hope that the S939 FX53 beats out the the 560, it's much more expensive (and it's also available in the market, but that's a different story altogether). If you can find benches on the new boards that show that the 3800+ is faster than the P4 560 in 3DS Max, I'll be glad to concede the point (despite the fact that the 560 should only be equivalent to a 3600+).

 

[stickybytes]

Where is the amd 64 3200+ in the middle of all this? Of course it could be assumed that it would be somewhere below the 3400+ and better then the p4's 3.2 and one can generally assume where it would rank but it would still be nice to see some concrete numbers.

 

[CraigRT]

Originally posted by: FullRoast
Anyone else surprised that the 3400+ (socket 754) beat out the 3500+ (socket 939)?

Yeah, I noticed that too... oh well.. its not like it's anything other than negligible

 

[mechBgon]

Also, I see you didn't care to comment on the 3DS and Lightwave benchies. I was replying to the post that the K8 is faster in all respects to the P4, which it clearly isn't. Most, yes; all, no.

Do you actually use a 3D modelling/animation app yourself? My experience with my budget 3D app (trueSpace) has been that the actual rendering phase of the job is the tip of the iceberg. I was working on a scene that took about three days of processing time to reach a ~85% radiosity solution. It rendered in about 30 seconds.

See what I'm getting at? Fast rendering is good in its own right, especially if you have lots of frames to render, but there's a heck of a lot more to using a 3D modelling app than the rendering. Suggestion: pick a simpler target where the rendering speed/encoding time/framerate really IS the only thing that matters. Like Doom 3, maybe :evil:

(trivia fact: I've been using the trueSpace family since buying my first PC, a refurb Cyrix PR166 with 16MB of EDO and Win95. I got it at Egghead's B&M store for $700 and found trueSpace2/SE in the bargain bin for $10 or something, and down the slippery slope I went )

 

[CraigRT]

Originally posted by: RaynorWolfcastle
Was there any doubt that AMD would be winning these benchmarks? It's been established for awhile already that the K8 architecture is more suited to gaming than Netburst.

I can't believe the replies in this thread and how attached people are to a specific company; maybe someone should make a thread showing off some Lightwave/media encoding benchmarks so that Intel fanboys can show off their huge e-penis. :roll:

edit: Just so that the fanboys don't flame me, I own computers with both AMD and Intel CPUs; I ambivalent towards the companies and buy what processor suits my needs best.

StrangerGuy

Now the A64 beats the P4 in every benchmark, including media encoding.

Really? Kind of like in this benchmark here? Text
or how about Lightwave and 3DS Max? Text

Who really cares. both CPU's can do the job, and do the job well.. fact is one is slightly faster than the other, and in some (not many cases) a decent amount faster. I personally don't care, but I stick to AMD cause I love the low cost of their performance... but if it were Intel, I'd be buying their CPU's. But you honestly gotta love the performance number difference in Doom3... wow! it's weird though how this is NOT the case with engines like Quake3... made by the same person (carmack)... Intel smackdown with the Q3 engine... maybe someone who knows more about how games work can explain that.. seems kinda strange to me!

 

[magomago]

Originally posted by: RaynorWolfcastle

Originally posted by: SickBeast

Originally posted by: RaynorWolfcastle
You don't get it. I don't care if the apps *are* optimized for the P4 or not, I just care about the encoding time. I don't care about having fair benchmarks, I care about getting my encoding done. I don't use CCE, I use tmpenc which is heavily P4 optimized (I don't have the money to spend on CCE, maybe you do). If the apps I needed were all K8 optimized, then I'd use an A64 to do my encoding.

Also, I see you didn't care to comment on the 3DS and Lightwave benchies. I was replying to the post that the K8 is faster in all respects to the P4, which it clearly isn't. Most, yes; all, no.

With the replies on this board, you'd think that by proclaming anything other than that the A64 is the be-all and end-all CPU, you'd just raped their mother. :roll:

Raynor,

The benchmarks you posted are outdated by over a month compared with these.

Read the commentary on workstation performance:

On the whole. we see a pattern of even workstation performance between the top Intel 560 on 775 and the top AMD FX53 on 939. The only exceptions are Intel topping Light as it usually does, and the FX53 topping 3dsMax. The nForce and VIA chipset boards were fairly even in SPECviewperf except for the standout performance of the K8N Neo2. Otherwise, there is little separating the performance of any of these boards in Workstation Performance.

People reccomend A64s like hotcakes on here because as a whole they are the superior processor when compared with the P4, plus they offer future 64-bit support. If the A64 can match the P4 in these 32-bit benchmarks, it will annihilate it once 64-bit apps are around.

I'm sure people were pimping the Celeron 300's back in the day just as much as this if not worse.

I bolded the section in my comments relating to my point.

Also, the second set of benchies compares the Athlon FX with the P4 560, my comment related to the P4 560 vs the Athlon64s, if you want to compare the FX53, compare it with the P4 EE. I should certainly hope that the S939 FX53 beats out the the 560, it's much more expensive (and it's also available in the market, but that's a different story altogether). If you can find benches on the new boards that show that the 3800+ is faster than the P4 560 in 3DS Max, I'll be glad to concede the point (despite the fact that the 560 should only be equivalent to a 3600+).

Doesn't matter if its 560 vs 3800+ If Intel doesn't put out anything faster it is their fault. You can't say "pit the fastest against eachother" and then decide to limit AMD's top processor (Though I do agree with the FX vs EE arguements)

Now let us just see AMD market well~ they haven't had it this good in a long time

 

[Connoisseur]

sickbeast there's no way in hell a mobile barton even at 2.4 ghz comes close to an A64 3200+ in the performance chart. Even tho a mobile barton has 512k L2 cache, on the chart i'd probably put it slightly higher than the P4 3.2C. Randommai i'd think our mobile bartons (I have one at 2.4ghz too ), would probably perform somewhere between the 3.2C P4 and the K8 sempron 3100+. This is of course speculation based on the benchmark and what the article said about the game being dependent on cache. It's not going to touch any of the Athlon 64's as far as performance goes (except MAYBE the 2800+). My two cents.

P.S. if I had a Geforce 6800U I wouldn't care one damn bit If I ONLY got 67+ fps on 1280x1024 with my mobile barton.

 

[caz67]

Good to see Nvidia rocking in Doom3.

 

[Cerb]

AXPs aren't fairing badly, either.
Yay, no more upgrading for me this year!

 

[Calin]

Originally posted by: SonicIce
Looks like AMD is the cats meow. But man this is the worst ive seen inten beaten by AMD, usually its a closer battle.

Since the AMD Slot A Athlon was launched - and even then the Athlon's advantage was lower than now
And man, the Slot A Athlon processors looked really cool :brokenheart:, much better than the Pentium II and !!! slot 1

Calin

 

[SonicIce]

Originally posted by: Calin

Originally posted by: SonicIce
Looks like AMD is the cats meow. But man this is the worst ive seen inten beaten by AMD, usually its a closer battle.

Since the AMD Slot A Athlon was launched - and even then the Athlon's advantage was lower than now
And man, the Slot A Athlon processors looked really cool :brokenheart:, much better than the Pentium II and !!! slot 1

Calin

I have a slot 1 300mhz celeron