Curtesy of Ace's Hardware
These are inhouse benchmarks as well, not supplied by Intel or using Intel configured systems.
These are inhouse benchmarks as well, not supplied by Intel or using Intel configured systems.
Pentium 4 Extreme Edition Benchmarks! At 3.4GHz too!!
- Thread starter andreasl
- Start date
quote from link: "Is this kind of performance good enough to counter the Athlon 64 FX?"
I am thinking not ... Athlon 64s I think will still hold the "gamer" tittle.
I am thinking not ... Athlon 64s I think will still hold the "gamer" tittle.
dullard
Elite Member
- May 21, 2001
- 25,765
- 4,293
- 126
There was one site claiming that the Athlon FX is 40%-50% faster in some games than the 3.06 GHz P4. According to Ace's results the 3.4 GHz P4EE is 18%-31% faster than the 3.06 GHz P4 (assuming the typical 10% average boost the 3.2 GHz P4 gets over the 3.06 GHz P4). So this takes a big chunk out of the Athlon FX lead - but it still will be in the lead. The question now becomes: will Prescott also have an Extreme Edition? If so then the 10%-15% boost from Prescott when added to the EE boost will make a Prescott EE basically tied with the Athlon FX.
It seems that intel must be deeply concerned over the performance of the Athlon FX. This EE edition really looks like a desparation move, increasing die size over 3x while not really changing anything internally (besides the cache) just doesnt make sense economically. It really shows how much power intel has over the cpu market as no other mfg could afford to sell this cpu at a competative price. (just my opinion I could be wrong 
)
)on the 23rd is the official lauch of the amd64 and i presume ( ) that reviews will be available at that point. so in four days you know for sure
) that reviews will be available at that point. so in four days you know for sure 
Adul
Elite Member
what was the name of one of faith no more's albums? king for a day...Originally posted by: Alkali
My Birthday is on the 23rd, and I'm 23.... I feel so Special
:beer:
Regs
Lifer
- Aug 9, 2002
- 16,665
- 21
- 81
The assumptions made by dullard are a little bit fanciful, but so are the specs from the Athlon FX using Opterons. Actually, what Anand did was make a educated guess. Saying that the Athlon FX should be 30-50% faster then a top of the line P4.
What I get from he Extreme Edition, is that the flopped on more L2 cache without changing the actual architecture of the cache itself. Should be no big deal .
What I get from he Extreme Edition, is that the flopped on more L2 cache without changing the actual architecture of the cache itself. Should be no big deal .
ok, well i was on the front page and read that someone quoted xbitlabs, or someone from idf and it went something like
" the first prescott 3.2ghz will not perform as well as the 3.2ghz EE"
although that person might have been wrong.
but intel i think is just trying to keep the performance crown, they arent on their heels, but just trying to play the numbers game. LARGER CACHE = markeneering 64bit= semi markeneering
MIKE
" the first prescott 3.2ghz will not perform as well as the 3.2ghz EE"
although that person might have been wrong.
but intel i think is just trying to keep the performance crown, they arent on their heels, but just trying to play the numbers game. LARGER CACHE = markeneering 64bit= semi markeneering
MIKE
The Prescott will have 1 MB on die L2 cache, not L3 cache. More importantly, it will have a revamped L1 cache and it will retire 4 uops per cycle, IIRC.
64bit = marketting. Larger cache = marketting. Every little bit of advantage = marketting.
dullard
Elite Member
- May 21, 2001
- 25,765
- 4,293
- 126
If I can't do that then I'm nothing. You took the one thing I'm an expert at away from me. Oh well. We can talk about just the hard facts then.
I don't have the page in cache, nor the individual results - but I saved my final numbers at least:
The EE added a 1.5% to 14% boost when going from 3.2 GHz P4 to 3.2 GHz P4 EE. Average boost was 7.4%.
The 3.4 GHz P4EE has a 7.1% to 19.4% boost over the 3.2 GHz P4. Average boost was 12.7%.
Or just comparing the 3.4 GHz P4EE to the 3.2 GHz P4EE it was about exactly 5% faster on all programs. Since 3.4/3.2 =1.0625 it did not come close to the full 6.25% faster it could have gotten. Performance isn't ramping up as much as clockspeed for Intel anymore.
I have the benchies in cache:
Comanche 1024x768x32
P4 3.4 GHz EE : 73.8
P4 3.2 GHz EE : 70.4
P4 3.2 GHz : 61.8
Unreal Tournament Asbestos Botmatch 1280x1024x32
P4 3.4 GHz EE : 104.4
P4 3.2 GHz EE : 99.7
P4 3.2 GHz : 90.4
Wolfenstein Ennemy Territory 1024x768x32 High quality. Many thanks to our hardcore online gamer, Hawski, alias Uffe Merrild.
P4 3.4 GHz EE : 142.6
P4 3.2 GHz EE : 135.9
P4 3.2 GHz : 130.8
Unreal II (patched to 1.03)
P4 3.4 GHz EE : 68.9
P4 3.2 GHz EE : 65.6
P4 3.2 GHz : 61.2
Comanche 1024x768x32
P4 3.4 GHz EE : 73.8
P4 3.2 GHz EE : 70.4
P4 3.2 GHz : 61.8
Unreal Tournament Asbestos Botmatch 1280x1024x32
P4 3.4 GHz EE : 104.4
P4 3.2 GHz EE : 99.7
P4 3.2 GHz : 90.4
Wolfenstein Ennemy Territory 1024x768x32 High quality. Many thanks to our hardcore online gamer, Hawski, alias Uffe Merrild.
P4 3.4 GHz EE : 142.6
P4 3.2 GHz EE : 135.9
P4 3.2 GHz : 130.8
Unreal II (patched to 1.03)
P4 3.4 GHz EE : 68.9
P4 3.2 GHz EE : 65.6
P4 3.2 GHz : 61.2
am i blind? where are the benchies? i spent 15 minutes on that site and still cant find the numbers
Perhaps you should try reading the thread.
They were pulled because Aceshardware wasnt aware of an NDA.
The benchmarks are, of course, very preliminary, just indicative of the results we should be getting. Some of them were done at 1280x1024, certainly not the optimal resolution for a processor benchmark.
A little off topic but who cares?THe only benchies I wanna see are the guy on this forum who had an overclocked afx to 2.8. However he will get in mucho trouble if those pics leak out so it looks like im waiting a week
Originally posted by: dexvx
Perhaps you should try reading the thread.
They were pulled because Aceshardware wasnt aware of an NDA.
The benchmarks are, of course, very preliminary, just indicative of the results we should be getting. Some of them were done at 1280x1024, certainly not the optimal resolution for a processor benchmark.
were those the only ones they did?
BD231
Lifer
- Feb 26, 2001
- 10,568
- 138
- 106
Competition dosen't mean one opponent is tucking their tail between their legs at all times. It seems like every CPU release brings about posts of fear between AMD or Intel which is getting quite irritating.
Intels move is no different than AMD releasing Barton, they're simply buying time.
It's pretty much a "Hey everyone!, we're still here and we're not sitting on our asses, here's a little gift for you while we add the finishing touchs to our new next best thing that'll be out before you even consider buying this!"
Pffft, extra cache..... like putting sprinkles on cake.
Intels move is no different than AMD releasing Barton, they're simply buying time.
It's pretty much a "Hey everyone!, we're still here and we're not sitting on our asses, here's a little gift for you while we add the finishing touchs to our new next best thing that'll be out before you even consider buying this!"
Pffft, extra cache..... like putting sprinkles on cake.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 24K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 20K
-
-
-
Facebook
Twitter