AMD Bulldozer benchmarks leaked

[Soleron]

.

I meant per-core, and I was saying literally worst case as in they copy pasted it from Greyhound with no improvements. Which of course they did not.

But it won't be anywhere near "double" or "half" current per-core performance, as I've seen it argued.

 

[HW2050Plus]

Gigabyte has confirmed this to be a fake:

http://www. rumorpedia/exclusive-amd-bulldozer-details-from-gigabyte/
Remove space in link

 

[bstaley]

FYI, the Z68 boards are available at NewEgg:
http://www.newegg.com/Product/ProductList.aspx?Submit=ENE&DEPA=0&Order=BESTMATCH&Description=z68&x=17&y=34

 

[veri745]

FYI, the Z68 boards are available at NewEgg:
http://www.newegg.com/Product/ProductList.aspx?Submit=ENE&DEPA=0&Order=BESTMATCH&Description=z68&x=17&y=34

What does this have to do with BD leaks?

 

[bstaley]

The benchmarks were proven fake and I posted earlier in the thread that I was leaning towards the Z68 since there is still no info on BD out there. Is this a problem?

 

[GlacierFreeze]

The benchmarks were proven fake and I posted earlier in the thread that I was leaning towards the Z68 since there is still no info on BD out there. Is this a problem?

Yes, it's called thread derailment. Stay on topic. It's makes forums a better place.

 

[aigomorla]

This will not look good on AMD.

Even tho they had no intensions of doing it, the fact someone did it, is making them look like what they did during the Phenom launch.

Really sad... I hope AMD comes out with an offical saying its fake.
Becuases people thinking its real, finding out the hardway its fake, is only lighting a fuse to a nuclear bomb for AMD.

 

[Morg.]

Well .. Do you really think anyone from Gigabyte has any right to say anything but: 'this is a fake' ?

NDA's anyone ?

I'm still waiting for a good reason that this would be a fake tbh ... gigabyte saying it is is just normal PR.

What is interesting however, is that everyone thinks that Intel stretching the core architecture from the core1 to the sandy bridge is an eternal victory.

It seems reasonable that a new CPU based on a new architecture that has been in the works for 4+ years would at least match a CPU made in the same process, with an outdated architecture (yes, core1 to sandy bridge is very much the same arch.).

We'll see pretty soon anyway, but those numbers *would* make sense.

AMD talked about bulldozer somewhere around the end of the athlon XP era iirc, and the core architecture from Intel has been dominating ever since.

So for those who are too young to remember :

Anything between 486 and Athlon = Intel win by far
Athlon XP from start to finish = AMD win by far
Core arch. from start to finish = Intel win by far

What's next ? I'd be betting on AMD this time tbh... they are already better than Intel in some areas, and by a strong margin:

Like I said in the comments on the latest Westmere-EX review, a 32nm CPU from Intel barely beats (1% or so) a 45nm CPU from AMD in performance per watt.

 

[bunnyfubbles]

So for those who are too young to remember :

Anything between 486 and Athlon = Intel win by far
Athlon XP from start to finish = AMD win by far
Core arch. from start to finish = Intel win by far

What's next ? I'd be betting on AMD this time tbh... they are already better than Intel in some areas, and by a strong margin:

I really don't remember Athlon XP being a hands down winner like you seem to be (and don't forget that AMD won the race to 1GHz before that). Maybe XP was better early on in the P4 vs. XP battle when Intel was struggling with Willamette, but it wasn't like the first XP chips were blazing fast either. And after both XP and P4 experienced process shirnks they both blew up in clock speed and even then there were distinct advantages to both, with the Pentium 4 having a decisive edge with media encoding and multi tasking. In fact I'd go as far as saying Pentium 4 really was the better overall chip and platform, while the XP was a winner particularly for gamers and budget oriented builds

It really wasn't until the Athlon 64 and particularly AMD beating Intel to dualcore that AMD truly ever had a clear advantage over Intel in just about every feasible category as far as desktop CPUs (on the flip side this was also the age of Centrino and the Pentium M dominating laptops)

Like I said in the comments on the latest Westmere-EX review, a 32nm CPU from Intel barely beats (1% or so) a 45nm CPU from AMD in performance per watt.

Server side of things is a whole different animal

http://www.anandtech.com/bench/Product/287?vs=362

Intel's fastest quad vs. AMD's fastest quad

Right now it would take a small miracle for AMD to even match that performance/watt let alone beat Intel strictly by performance.

I'm pulling for Bulldozer to be a success, I really am, but I'm very skeptical. I can see it being yet another XP vs. P4 situation where Bulldozer is a hands down better option when it comes to things like media encoding and multi tasking but I don't expect it to be a grand slam where its better across the board like when AMD enjoyed their X2s vs. high end P4s (and even PDs)

 

[Soleron]

Intel's fastest quad vs. AMD's fastest quad

Right now it would take a small miracle for AMD to even match that performance/watt let alone beat Intel strictly by performance.

If AMD's quads were on the improved low-k 45nm process that Thuban is on the perf/watt would be a lot closer. Then add the 32nm shrink, HKMG, power gating and proper Turbo. That's five advantages Intel has that will be negated with BD even ignoring architecture.

 

[Morg.]

I really don't remember Athlon XP being a hands down winner like you seem to be (and don't forget that AMD won the race to 1GHz before that). Maybe XP was better early on in the P4 vs. XP battle when Intel was struggling with Willamette, but it wasn't like the first XP chips were blazing fast either. And after both XP and P4 experienced process shirnks they both blew up in clock speed and even then there were distinct advantages to both, with the Pentium 4 having a decisive edge with media encoding and multi tasking. In fact I'd go as far as saying Pentium 4 really was the better overall chip and platform, while the XP was a winner particularly for gamers and budget oriented builds

Well ... you missed a lot.

The XP beat the P4 at every single edition.
The P4 never was better at multithreading than the XP.
The P4 was a clear loser for all of its existence, but it was justified by Intel marketing, which also convinced you obviously.

Now maybe some of the 1000 bucks retarded models had no match on the AMD side because AMD did not care (and I'm not even sure of that) - but the sure thing is there was no good reason to buy Intel.

You also have to remember that Intel started the P4 with that dumb RDRAM idea, which they eventually dropped - but for a while it was RDRAM or SDRAM and nothing else, all the while Athlons enjoyed DDR and then DC-DDR etc.

The fact of the matter is, for the whole athlon XP era, there was no competitive offering from Intel and every single pc sold should have been built with Athlons. (Oh and for those who OC, you know how the intel core series OC'd decently ? well some guy did a +100% OC on Athlon -- it WAS a really good chip - in it's time the equivalent of the C2D when it started )

Here some links from Anandtech (disregard the conclusion, as Anand clearly overweights "content creation" and "encoding", which we all know are somewhat irrelevant for 99% of the population, unlike gaming and office productivity):
http://www.anandtech.com/show/835/13
http://www.anandtech.com/show/1066/30

And yet .. that is all without mention of the price, which was a huge win in favor of AMD.

 

[bunnyfubbles]

Well ... you missed a lot.

The XP beat the P4 at every single edition.

I don't think I missed anything, whereas you are clearly remembering things not as they were but the way you want to remember them

The P4 never was better at multithreading than the XP.

I guess you missed the Hyper Threading boat.

The P4 was a clear loser for all of its existence

I clearly stipulated the P4's early shortcomings and lategoings, but the Pentium 4 Northwood was a winner, especially when AMD at the time was struggling with Thoroughbred A and Intel eventually enabling HT for the P4.

but it was justified by Intel marketing, which also convinced you obviously.

Yeah, Intel certainly convinced me into buying two Athlon XP rigs (a 1700+ Palomino and a 2400+ Mobile Barton with its unlocked multiplier)

Although those evil bastards somehow managed to trick me into buying an ever so inferior 2.4GHz Pentium 4C which was such a dog @ 3.3GHz

Now maybe some of the 1000 bucks retarded models had no match on the AMD side because AMD did not care (and I'm not even sure of that) - but the sure thing is there was no good reason to buy Intel.

Yeah, having a clearly faster processor for multimedia content creation wasn't good enough.

You also have to remember that Intel started the P4 with that dumb RDRAM idea, which they eventually dropped - but for a while it was RDRAM or SDRAM and nothing else, all the while Athlons enjoyed DDR and then DC-DDR etc.

Again, I clearly stipulated that Intel struggled with the Pentium 4 with Willamette where their first 1.5GHz Pentium 4 struggled to assert superior performance to Intel's own 1GHz Pentium 3.

The fact of the matter is, for the whole athlon XP era, there was no competitive offering from Intel and every single pc sold should have been built with Athlons.

(Oh and for those who OC, you know how the intel core series OC'd decently ? well some guy did a +100% OC on Athlon -- it WAS a really good chip - in it's time the equivalent of the C2D when it started )

Yes, lets take the super rare successes from AMD and blow them out of proportion when in reality we had just as much successful real-world overclocks on the Intel side with gems such as the P4 1.6A that could regularly pull off a 50% overclock and hit 2.4GHz at a time when AMD was struggling with clockrate scaling.

Here some links from Anandtech (disregard the conclusion, as Anand clearly overweights "content creation" and "encoding", which we all know are somewhat irrelevant for 99% of the population, unlike gaming and office productivity):
http://www.anandtech.com/show/835/13
http://www.anandtech.com/show/1066/30

And yet .. that is all without mention of the price, which was a huge win in favor of AMD.

Yup, if it doesn't work in your favor, just ignore it. Typical fanboy zealotry. How dare any of us actually use our computers for actual work.


Intel really screwed up with Netburst, there no argument from me there, but to say that AMD completely dominated with Athlon XP or that Intel couldn't even put out a competitive product is completely ridiculous

And at the end of the day the proof is all in the pudding: AMD still had to undercut Intel when it came to pricing. If AMD truly had a superior product they would have been able to price their XP line higher. And proof of that lies with the Athlon 64 and X2 days when AMD was able to sell their own "extreme edition" FX chips for $1000 a pop when Intel was struggling during the waning days of Netburst.

 

[Morg.]

Meh. Whatever I won't convince you and this is pointless. (yes I followed the p4 1.6a with ridiculously easy fsb change and I did advise people to buy it)

However, the cited benchmarks show clearly a big lead by AMD in everything around Xp1500+ and then a decent lead by AMD in gaming and office productivity around barton 3000+.

Now if you think something that's priced above for the same performance is competitive, I guess we don't give the same meaning to words.

AMD had to undercut Intel because Customer perception was that Intel > AMD and all OEM's played that game (no high-end computers from PB/HP/others with AMD CPU's in the Athlon era that I remember, all P4 although linked benchmarks show that was counter productive) because Intel asked them to.

And of course, any company whose product is recognized as the best will take advantage of it and go for retarded pricing, just as Intel did and does, as AMD did and will do, as even ATI/AMD did and will do again and as nVidia is currently doing and will do again.

Pricing is only related to market, never to technology. Here's a good example you cannot ignore :
iPhone : more expensive than anything else, limited abilities, limited screen, limited 4g etc..
VS
an army of Android phones that are lesst han half the price, have better screens, more functionalities, 4g since before the iPhone 3g etc.

BUT does it matter ?
No because customers think iPhone great, customer wantz iPhone cuz customer don't know what iPhone is anyway.

Just like anything remotely technological, customers are completely ignorant of what they are buying and as long as they're told it's the best they'll buy it and be happy.

 

[soccerballtux]

Well ... you missed a lot.

The XP beat the P4 at every single edition.
The P4 never was better at multithreading than the XP.
The P4 was a clear loser for all of its existence, but it was justified by Intel marketing, which also convinced you obviously.

Now maybe some of the 1000 bucks retarded models had no match on the AMD side because AMD did not care (and I'm not even sure of that) - but the sure thing is there was no good reason to buy Intel.

You also have to remember that Intel started the P4 with that dumb RDRAM idea, which they eventually dropped - but for a while it was RDRAM or SDRAM and nothing else, all the while Athlons enjoyed DDR and then DC-DDR etc.

The fact of the matter is, for the whole athlon XP era, there was no competitive offering from Intel and every single pc sold should have been built with Athlons. (Oh and for those who OC, you know how the intel core series OC'd decently ? well some guy did a +100% OC on Athlon -- it WAS a really good chip - in it's time the equivalent of the C2D when it started )

Here some links from Anandtech (disregard the conclusion, as Anand clearly overweights "content creation" and "encoding", which we all know are somewhat irrelevant for 99% of the population, unlike gaming and office productivity):
http://www.anandtech.com/show/835/13
http://www.anandtech.com/show/1066/30

And yet .. that is all without mention of the price, which was a huge win in favor of AMD.

welcome to the internet sir.

 

[Morg.]

Arguing over the internet is like the special olympics, even if you win, you're still retarded

 

[bunnyfubbles]

Meh. Whatever I won't convince you and this is pointless. (yes I followed the p4 1.6a with ridiculously easy fsb change and I did advise people to buy it)

Why bother starting something you knew you couldn't finish in the first place?

However, the cited benchmarks show clearly a big lead by AMD in everything around Xp1500+ and then a decent lead by AMD in gaming and office productivity around barton 3000+.

A. those bench marks don't show any sort of significant lead for the XP over P4, certainly nowhere near the level of ass-kicking that the Athlon X2 enjoyed over Pentium D or what we see now with Core i5/i7 Vs. Phenom II

B. there are several instances where Intel had the fastest chip (just go back and look at your linked benches which show Intel had the fastest chip for Quake 3 regardless of if it was a 2GHz P4 vs. an XP 1800+ or P4 3.06GHz HT vs. Athlon XP 3000+ (2.167GHz) Barton

Of course you have your AMD blinders on and refuse to see that. Yes, the Intel chip was hotter and more expensive and the performance differential often not worth it for the situations where it was faster, particularly in games, but it was still there, you can't ignore it and pretend like it isn't there, it makes you look crazy

Now if you think something that's priced above for the same performance is competitive, I guess we don't give the same meaning to words.

Well it should be obvious we don't share the same views when you will simply ignore facts that work against what you want to believe. Just go back to the P4 vs. Barton and see the plethora of benches where the P4 trounces the XP in just about any CPU intensive benchmark one would actually need a fast processor for in a work related task.

the Athlon XP was an excellent chip for enthusiasts because it offered value, it offered competitive performance for a low cost.

AMD had to undercut Intel because Customer perception was that Intel > AMD and all OEM's played that game (no high-end computers from PB/HP/others with AMD CPU's in the Athlon era that I remember, all P4 although linked benchmarks show that was counter productive) because Intel asked them to.

tinfoil hat much? If AMD had a clearly superior processor the truth wouldn't be withheld, they would have enjoyed sales just as good if not better than the days of the Athlon 64 and X2 vs. P4/PD

And of course, any company whose product is recognized as the best will take advantage of it and go for retarded pricing, just as Intel did and does, as AMD did and will do, as even ATI/AMD did and will do again and as nVidia is currently doing and will do again.

So wait a minute, aren't you now arguing in my favor? That the XP line wasn't perceived as the best and thus priced accordingly? Pretty sure it wasn't perceived as the best because it was obvious that it wasn't the best. Having the best price/performance doesn't always win you market dominance.

Pricing is only related to market, never to technology. Here's a good example you cannot ignore :
iPhone : more expensive than anything else, limited abilities, limited screen, limited 4g etc..
VS
an army of Android phones that are lesst han half the price, have better screens, more functionalities, 4g since before the iPhone 3g etc.

Yup, we cannot ignore that: http://www.dailytech.com/article.aspx?newsid=21687

over twice the market share of their competition

If you're analogizing the Athlon XP to the Andriod phones and the Pentium 4 to the iPhone then why didn't the Athlon XP capture twice the market share?

ever heard of apples to oranges?

Just like anything remotely technological, customers are completely ignorant of what they are buying and as long as they're told it's the best they'll buy it and be happy.

And just like anything in the history of competitive products there will always be irrational fanboys of one over the other

 

[amenx]

Well it should be obvious we don't share the same views when you will simply ignore facts that work against what you want to believe. Just go back to the P4 vs. Barton and see the plethora of benches where the P4 trounces the XP in just about any CPU intensive benchmark one would actually need a fast processor for in a work related task.

I remember that. I bought a P4 Northwood 2.8 based on this review:

http://www.anandtech.com/show/977/1

Intel was pwning AMD back then. It was only a year or so later when AMD struck back with the A64.

 

[KompuKare]

Ah the fun of arguing!
Actually for a while even the K6 was quite competitive.
The P4 was a pretty poor design (Intel had to quickly bury the P3 past 1GHz because it made the P4 look silly).

But Intel had and mostly still has two advantages:

- Better fabs. If you make billions in profit you can afford the best fab. Because x86 won the CPU wars (mainly because Motorola wouldn't do the deal IBM required), making x86 was very, very profitable. So even with a poor design like the P4, Intel could compete since they were able to run their power hungry design on a smaller process.

- The money and willingness to offer illegal deals to keep the competition out. Hopefully if by some miracle AMD once again have the better design Intel will play fair but I think it is unlikely since Intel have executed really well since Core2, although the Atom team hasn't really delivered their potential.

 

[tweakboy]

Wow 8 cores and 8 HT. that is 16. wow , amazing.

Ivy bridge will come soon

 

[Broheim]

Pricing is only related to market, never to technology. Here's a good example you cannot ignore :
iPhone : more expensive than anything else, limited abilities, limited screen, limited 4g etc..
VS
an army of Android phones that are lesst han half the price, have better screens, more functionalities, 4g since before the iPhone 3g etc.

BUT does it matter ?
No because customers think iPhone great, customer wantz iPhone cuz customer don't know what iPhone is anyway.

@bolded part: I laughed pretty hard, you should have your eyes checked.

all your other iPhone vs. Android points are completely false, it sounds like you have penis envy.

 

[KingstonU]

Anybody else catch the note that CPU-Z is now supporting "DDR4"?

 

[poohbear]

if Bulldozer was better than SB, AMD would be screaming that at the top of their lungs. Their silence on the issue shows its sub par or at best matches the SB, so if u're gonna upgrade, upgrade now.

 

[jimbo75]

if Bulldozer was better than SB, AMD would be screaming that at the top of their lungs. Their silence on the issue shows its sub par or at best matches the SB, so if u're gonna upgrade, upgrade now.

If I buy SB now and BD ends up faster, will you buy my SB off me for what it cost?

 

[tweakboy]

We did them already didnt we. nvidia drivers latest WHQL ?

 

[podspi]

This thread needs to die :biggrin:


Every time I check the CPU subforum I (irrationally) think that someone has finally revealed all, and instead there is arguing about the Athlon XP and the Pentium 4...

As someone who still uses a Pentium 4 HT @3ghz (at work) on a semi-regular basis, I have to say, it is frankly amazing how much faster CPUs have gotten. Even my N270 feels faster...