PCgameshardware : Bulldozer? Please. Intel Confirms 8 Core SB-E For Q3

[JoJoman88]

It will come down to how "much" better the 8 core BD will be vs a 4 core SB as to where prices for BD come in at. If SB falls behind to much, Intel will drop prices on SB to answer AMD. Just remember that AMD is using the doing more with less thing big time with BD and that you are getting 8 cores from less die space than 8 Intel cores. There will be a few times that the missing die will hurt BD(not too many times I think). So on matching core count to matching core count, unless AMD has made big improvements to other the parts of the processor, I do not think an 8 core BD out doing an 8 core SB or IB if Intel has such up their sleves. This is going to put AMD back into the game, what will be just as important will be Intel's response to BD. Lower prices or better processors yet to come or both. Both is what we need!

 

[eternalone]

Wow Didnt know the Phenom X6 actually beats the SB 2500k in performance, then Im pretty sure BullDozer might outperform SB.

 

[drizek]

So, when you take 40%, of the cleint market, layer on the 20% that build their own systems and then look at the 5% (overly generous) that are concerned about having the top speed at all costs, you are really looking at about .4% of the market that is focused on those top bin desktop SKUs. So, the corollary is that 99.6% of the market is determined outside of that top performance SKU.

But it's not just about the top performance SKU. It's also about cheap chips that overclock to match them. AMD used to be company that built those chips 6 years ago, but that changed. I just saw a thread the other day about a Q6600 overclocked to 4GHz. That's what Intel had competing with Phenom at 2.3GHz. The Q6600 was selling for around $200, not $1500, and it is where AMD got killed.

Fast forward to today and the 2500K is the Q6600s spiritual successor. $200, overclocks like crazy. That's what enthusiasts want.

Is there a Bulldozer version of that?

Edit: I just noticed in the graph above that a stock 2.4GHz Q6600 outperforms a 3.5GHz Phenom II X4 in Monte Carlo. That's just wrong.

 

[Castiel]

Looking at this graph and just imagining an 8 Core SB-E pulling 45,750 in Cinebench;s multithreaded test

 

[Riek]

Looking at this graph and just imagining an 8 Core SB-E pulling 45,750 in Cinebench;s multithreaded test

Looking at the graph and assuming SB 8core can reach 3.4GHz you might see something in the line of 37781. e.g. far from the 45,750.

In that respect an hypotetical 16core thuban would reach 46,481

 

[JFAMD]

Here's a thought:

People will look at their budget and they will buy the product that best meets their needs.

Today in x86 you can buy everything from Atom and Brazos up to 4-way Xeon and Opteron.

People will buy what they need.

Where is this desire (on both sides) for one product to be named "ruler of them all"?

The CPU market is a diverse market, those that argue one product is going to be better than the other must just like to argue.

 

[ed29a]

But it's not just about the top performance SKU. It's also about cheap chips that overclock to match them. AMD used to be company that built those chips 6 years ago, but that changed. I just saw a thread the other day about a Q6600 overclocked to 4GHz. That's what Intel had competing with Phenom at 2.3GHz. The Q6600 was selling for around $200, not $1500, and it is where AMD got killed.

Fast forward to today and the 2500K is the Q6600s spiritual successor. $200, overclocks like crazy. That's what enthusiasts want.

Is there a Bulldozer version of that?

Edit: I just noticed in the graph above that a stock 2.4GHz Q6600 outperforms a 3.5GHz Phenom II X4 in Monte Carlo. That's just wrong.

You didn't bother reading JFAMD's post or didn't understand it. All I get from your post is: Overclock, overclock, overclock! That fals into the 1% of people who buy computers (and I am being very generous) category. The people who buy computers from Dell and HP don't overclock, they don't even understand what the hell is a CPU. These people alone are the vast majority of the consumers who buy computers. If AMD doesn't want to cater to the rest, the enthusiasts, it's their choice and it makes perfect business sense.

 

[Makaveli]

Wow Didnt know the Phenom X6 actually beats the SB 2500k in performance, then Im pretty sure BullDozer might outperform SB.

lol the only thing that graph shows is what we already know in test where amd can use all their cores example encoding they generally do well. When you look at overall performance across most apps tho the Phenom doesn't look so hot.

And if I was building a system tomorrow I would not choose a Phenom X6 over a SB board once you overclock both chips the 2500k chip should eat the phenom regardless of core count.

It will also do it using less power.

 

[bandgit]

Where is this desire (on both sides) for one product to be named "ruler of them all"?

Mr. Fruehe, again with all due respect to an executive of your experience and standing, I would like to explain at least "my opinion" on that statement.

It is progress, sir. The primary driving force of computer technology should be to "go where no CPU has gone before", not fill in between the waves in the wake of a faster boat. It is an inexhaustible desire of humanity which barely five years after the courageous challenge of a President landed two men on the moon at a time when a calculator weighed four pounds and NASA's fastest computer would barely drive a modern phone. Benefits to humanity are measured by historians only by the leaders as they are the ones which shape our times. Does it really matter and does anyone really care who lost to FDR in his various elections? It only matters that he motivated America to save the world and keep the stars on the stripes rather than a swastika.

A private corporation such as AMD only has responsibilities to its shareholders to make a buck, and thus it can be argued that a "fill in the blanks Intel missed" is a valid company financial strategy. However, it is a profoundly flawed technological strategy. I propose to you that CPU companies are not shoemakers or t-shirt manufacturers, sir. They have a responsibility to humanity to lead the way in demonstrating how computer technology can save us from our dire mistakes (one of which may be currently in the process of rendering half of the great nation of Japan uninhabitable for millions of years) and lead the way to a brighter and more hopeful future for ourselves and our children.

Abandoning the technological cutting-edge leadership mantle to Intel does a disservice not just to enthusiast prosumer gamers and the like, it does a disservice to humanity. AMD is the only company in the world which is positioned to save us all from decades of technological monopoly by Intel, which like any other unchallenged monopoly will have stultifying effects on its entire field. The world is depending on computer technology to write our future. If the future is written by a monopoly which can charge what it wants for whatever it feels like producing, then AMD will go down in history not as a competitor who failed, but as a villain who postponed and hindered the future of humanity.

 

[Martimus]

Mr. Fruehe, again with all due respect to an executive of your experience and standing, I would like to explain at least "my opinion" on that statement.

It is progress, sir. The primary driving force of computer technology should be to "go where no CPU has gone before", not fill in between the waves in the wake of a faster boat. It is an inexhaustible desire of humanity which barely five years after the courageous challenge of a President landed two men on the moon at a time when a calculator weighed four pounds and NASA's fastest computer would barely drive a modern phone. Benefits to humanity are measured by historians only by the leaders as they are the ones which shape our times. Does it really matter and does anyone really care who lost to FDR in his various elections? It only matters that he motivated America to save the world and keep the stars on the stripes rather than a swastika.

A private corporation such as AMD only has responsibilities to its shareholders to make a buck, and thus it can be argued that a "fill in the blanks Intel missed" is a valid company financial strategy. However, it is a profoundly flawed technological strategy. I propose to you that CPU companies are not shoemakers or t-shirt manufacturers, sir. They have a responsibility to humanity to lead the way in demonstrating how computer technology can save us from our dire mistakes (one of which may be currently in the process of rendering half of the great nation of Japan uninhabitable for millions of years) and lead the way to a brighter and more hopeful future for ourselves and our children.

Abandoning the technological cutting-edge leadership mantle to Intel does a disservice not just to enthusiast prosumer gamers and the like, it does a disservice to humanity. AMD is the only company in the world which is positioned to save us all from decades of technological monopoly by Intel, which like any other unchallenged monopoly will have stultifying effects on its entire field. The world is depending on computer technology to write our future. If the future is written by a monopoly which can charge what it wants for whatever it feels like producing, then AMD will go down in history not as a competitor who failed, but as a villain who postponed and hindered the future of humanity.

Did you honestly say that?

If you feel that CPUs are so important to the advancements of the world, I suggest you put in the work to advance the technology to the level you feel is right.

If you feel that strongly about it, make sure it is done yourself and stop relying on others to do it for you. Yes, it is complicated and difficult to understand, but it is far from impossible, and you can drive a change if that is really what you want.

 

[Idontcare]

Mr. Fruehe, again with all due respect to an executive of your experience and standing, I would like to explain at least "my opinion" on that statement.

It is progress, sir. The primary driving force of computer technology should be to "go where no CPU has gone before", not fill in between the waves in the wake of a faster boat. It is an inexhaustible desire of humanity which barely five years after the courageous challenge of a President landed two men on the moon at a time when a calculator weighed four pounds and NASA's fastest computer would barely drive a modern phone. Benefits to humanity are measured by historians only by the leaders as they are the ones which shape our times. Does it really matter and does anyone really care who lost to FDR in his various elections? It only matters that he motivated America to save the world and keep the stars on the stripes rather than a swastika.

A private corporation such as AMD only has responsibilities to its shareholders to make a buck, and thus it can be argued that a "fill in the blanks Intel missed" is a valid company financial strategy. However, it is a profoundly flawed technological strategy. I propose to you that CPU companies are not shoemakers or t-shirt manufacturers, sir. They have a responsibility to humanity to lead the way in demonstrating how computer technology can save us from our dire mistakes (one of which may be currently in the process of rendering half of the great nation of Japan uninhabitable for millions of years) and lead the way to a brighter and more hopeful future for ourselves and our children.

Abandoning the technological cutting-edge leadership mantle to Intel does a disservice not just to enthusiast prosumer gamers and the like, it does a disservice to humanity. AMD is the only company in the world which is positioned to save us all from decades of technological monopoly by Intel, which like any other unchallenged monopoly will have stultifying effects on its entire field. The world is depending on computer technology to write our future. If the future is written by a monopoly which can charge what it wants for whatever it feels like producing, then AMD will go down in history not as a competitor who failed, but as a villain who postponed and hindered the future of humanity.

It can only be a disservice to humanity, within the context of the logic you are embodying here, if Intel follows suit and likewise abandons pursuit of cutting-edge advancements.

Along with IBM, and Oracle (SUN) and every ARM licensee.

Time to draft that letter to Otellini! It would be a disservice to humanity should Intel's rate of innovation slow down owing to the absence of competition. Right?

 

[bandgit]

Did you honestly say that?

If you feel that CPUs are so important to the advancements of the world, I suggest you put in the work to advance the technology to the level you feel is right.

If you feel that strongly about it, make sure it is done yourself and stop relying on others to do it for you. Yes, it is complicated and difficult to understand, but it is far from impossible, and you can drive a change if that is really what you want.

I am trying to drive the change in the best way I know how. Trying to get the point across to an AMD executive who is in a position to affect in some way the direction of the only company in the world today which has the technological basis to be able to compete with Intel.

If you want me to go raise a few billion dollars to build my own fab, I regret to inform you that is not exactly a likely scenario. You go right ahead and let me know how you fare.

It can only be a disservice to humanity, within the context of the logic you are embodying here, if Intel follows suit and likewise abandons pursuit of cutting-edge advancements.

Along with IBM, and Oracle (SUN) and every ARM licensee.

Time to draft that letter to Otellini! It would be a disservice to humanity should Intel's rate of innovation slow down owing to the absence of competition. Right?

Check your history books. Has there ever been in the history of free enterprise a single monopoly which has performed and innovated to the degree that it would have if it had a truly worthy competitor nipping at its heels? IBM, Oracle, and ARM are not players in the "personal computer" arena and likely will not be in the foreseeable future. I'm not talking about the enterprise or the smartphone here, I'm discussing the meat and potatoes PC market. And as for the letter to Paul O, I think he's far too busy laughing his a$$ off at AMD's (alleged) abject phail on the high end to pay much attention to me.

 

[Idontcare]

Check your history books. Has there ever been in the history of free enterprise a single monopoly which has performed and innovated to the degree that it would have if it had a truly worthy competitor nipping at its heels? IBM, Oracle, and ARM are not players in the "personal computer" arena and likely will not be in the foreseeable future. I'm not talking about the enterprise or the smartphone here, I'm discussing the meat and potatoes PC market. And as for the letter to Paul O, I think he's far too busy laughing his a$$ off at AMD's (alleged) abject phail on the high end to pay much attention to me.

I'm glad to see you agree with me.

So why are you blaming the competition that is bowing out of the race when it is clearly the monopoly that is to blame for forcing the competition out of the market only to then proceed with slowing down their own rate of development?

I see a common villain in your story of crimes against humanity, and it is not the cash-poor debt-rich David that is the culprit.

If you insist on AMD doggedly pursuing the cutting edge for the sake of humanity then surely you hold Intel to no less of a standard, even in the absence of competition should AMD's story parallel Via/Cyrix.

And what of these other technology titans shirking their moral imperatives to serve mankind? Exactly IBM and Oracle are not in the "personal computer arena" and why not? Don't they know that by willfully not pursuing Intel into this marketspace, at all cost, they are doing a disservice to humanity?

(mind you that I don't actually agree with the sentiment you are espousing, am merely pointing out the logical inconsistencies in your own application of this flawed sentiment...what is good for the goose need be good for the gander and if AMD is letting down mankind then by the same logic so too is Intel and Via and IBM and any other technology company in existence by virtue of them not doggedly pursuing a technology race that you'd have us believe is so imperative to humanity)

 

[Idontcare]

If you want me to go raise a few billion dollars to build my own fab, I regret to inform you that is not exactly a likely scenario. You go right ahead and let me know how you fare.

You may not believe me, but truth be told when I was 12 yrs-old I decided that semiconductor technology and atomic engineering (Feynman and "there's plenty of room at the bottom") was the thing that was going to take mankind to the next era in all things technology so I intentionally set about on a career path with the specific goal in mind of becoming a process development engineer so I could first-hand have a role in shaping the pace of the leading edge in process technology.

(I went on to become a process development engineer at Texas Instruments)

I did some back of envelope projections and came to the conclusion that to do the ab initio work needed to support the development of materials science and atomic-based engineering in silico I needed computer processing capability that would not be wholesale available until circa 2030. I wasn't about to sit idly by and wait for it to come to me, I decided I would actively engage the development cycle and in the meantime I'd do what I could to expedite the timeline and bring in that 2030 target date as much as possible so I get to work on the stuff I have a real passion for (alternative energy).

Martimus is right, if you believe and have passion you can make a first-person critical impact to the very subject you are addressing. The lazy path, and substantially less effective too, is to resort to attempting to influence development through nothing more than posts in forums.

edit: here's that computational capability projection:

 

[bandgit]

I can certainly see your point, Idontcare, and I would be more than happy to join with like minded individuals such as yourself in taking a hands-on approach to "changing the CPU world." I would have to get over my fundamental disgust at the ways that some fabs seem to be financed, with Dr. Ruiz deftly sidestepping the alleged charges of insider trading and terrorist group support. But hey, if I can play a role in helping to raise a few billion of ethical dollars, I'm in like Flynn!

 

[Idontcare]

I can certainly see your point, Idontcare, and I would be more than happy to join with like minded individuals such as yourself in taking a hands-on approach to "changing the CPU world." I would have to get over my fundamental disgust at the ways that some fabs seem to be financed, with Dr. Ruiz deftly sidestepping the alleged charges of insider trading and terrorist group support. But hey, if I can play a role in helping to raise a few billion of ethical dollars, I'm in like Flynn!

Ha ha, don't bring ethics into it though ...your space-race example above is fraught with ethical dilemnas regarding tax-dollars spent sending a few men to a moon while homeless adults and malnourished school children languished here on planet earth for lack of funding in social programs.

But I will make a generous offer, you bring your billion dollars (preferably 2 or 5 :sneaky, I'll bring my experience and know-how, and Martimus will bring his no-bull sense of focus and together we'll take on Intel and innovate like mad for the sake of mankind!

Deal?

 

[Edrick]

bandgit,

What is exactly wrong with the CPU world today? Are they not fast enough for you? Are they holding mankind back? The truth is that the advances we have made in this field n the last 20 years is quite astonishing. GPU HPC technology and what we can do with DC applications was unthinkable just one decade ago.

 

[bandgit]

Idontcare: I have about $50 in my savings account, and I'm more than happy to contribute it all to the "cause." I live in British Columbia which is suffering from chronic homelessness and a virtual collapse of our medical system but it didn't stop the province from losing a few billion dollars on throwing an Olympic party last year. "There will be poor always, pathetically struggling, look at the good things you've got." - Jesus Christ Superstar

Edrick: come on. We're using x86, the same user interface as Xerox Parc, speech recognition is still effectively a joke, and I'm still seeing DOS-style ASCII on my i7 when I boot up. 35 years have passed since the Apple I went on sale and to an outside observer it would be difficult to discern the difference between it and my current computer. 35 years before the Apple I, Richard Feynman was using a room full of female mathematicians as his computer in the Manhattan Project! The progress of 1941 to 1976 absolutely dwarfs the relative stagnation of 1976 to today. The same is evident in airplane and automotive design, etc. We "think" we live in an age of progress when instead the overall rate of engineering amelioration has stalled out!

 

[Phynaz]

But I will make a generous offer, you bring your billion dollars (preferably 2 or 5 :sneaky, I'll bring my experience and know-how, and Martimus will bring his no-bull sense of focus and together we'll take on Intel and innovate like mad for the sake of mankind!

Deal?

Lol. You'll get more enjoyment out of spending that money on beer :awe:

And if you give away a few billion dollars of beer you will benefit mankind.

 

[bandgit]

Lol. You'll get more enjoyment out of spending that money on beer :awe:

And if you give away a few billion dollars of beer you will benefit mankind.

OK, you talked me into it. I'm heading off to the beer store to spend that $50 on some of that fine Canadian brew!

 

[Idontcare]

OK, you talked me into it. I'm heading off to the beer store to spend that $50 on some of that fine Canadian brew!

Eh? You mean my beer, right? I'm afraid I'm going to have to re-purpose your beer fund in the name of humanity, yeah that, and uhm, science, yeah in the name of engineering I insist you beer-me post-haste! C'mon man, for the sake of the children! Will no one beer-me for the sake of the children!? :'(

 

[Edrick]

The progress of 1941 to 1976 absolutely dwarfs the relative stagnation of 1976 to today. The same is evident in airplane and automotive design, etc. We "think" we live in an age of progress when instead the overall rate of engineering amelioration has stalled out!

I do not share your views here. We went from vacuum tubes and punch cards (1975) to where we are today. Cell phones went from a brick (1985) to iphones of today. Processing power of supercomputers of 10 years ago now fit on a $500 graphic card. We have gone from 0.74 MIPS (1971) to 159,000 MIPS (2010) per CPU. (http://en.wikipedia.org/wiki/Instructions_per_second).

I do agree with you that it could be better. If the world depended on us creating CPUs 3 times as fast as today, I am sure we would have them. But it doesn't. Business plays a part as well. Not to mention that software lags behind hardware all the time as well. So while you are creating faster CPU, you also need to do something about us developers.

 

[HW2050Plus]

Looking at this graph and just imagining an 8 Core SB-E pulling 45,750 in Cinebench;s multithreaded test

If it can do that, but let's assume. But especially in Cinebench Bulldozer will show tremendous performance increase. So I would expect while Sandy Bridge EN 16T is likly faster in several other benchmarks to a 8C/8T Bulldozer, having the double logical core count, it might not be faster in Cinebench.

With the assumption of 2 speedup (per core per clock) and double of cores compared to 4 core Thuban you will come to a value of ~57 000 in Cinebench. A value even the SB-E cannot reach. Cinebench is likly a benchmark where Bulldozer will excel, while in others we will surely do not see such strong improvements and Sandy Bridge EN will surely be faster (namely all integer only stuff).

So this will surely be the wrong benchmark to demonstrate the supremacy of Sandy Bridge EN (Core i7 Extreme) vs. Bulldozer (Vision FX8000).

 

[bandgit]

Eh? You mean my beer, right? I'm afraid I'm going to have to re-purpose your beer fund in the name of humanity, yeah that, and uhm, science, yeah in the name of engineering I insist you beer-me post-haste! C'mon man, for the sake of the children! Will no one beer-me for the sake of the children!? :'(

Consider yourself beered! I'll email you a two-four of Molson Canadian right away. But remember that I, like the late great MJ, am doing it for the children!

I do not share your views here. We went from vacuum tubes and punch cards (1975) to where we are today. Cell phones went from a brick (1985) to iphones of today. Processing power of supercomputers of 10 years ago now fit on a $500 graphic card. We have gone from 0.74 MIPS (1971) to 159,000 MIPS (2010) per CPU. (http://en.wikipedia.org/wiki/Instructions_per_second).

I do agree with you that it could be better. If the world depended on us creating CPUs 3 times as fast as today, I am sure we would have them. But it doesn't. Business plays a part as well. Not to mention that software lags behind hardware all the time as well. So while you are creating faster CPU, you also need to do something about us developers.

My beef is not so much with MIPS or gigaflops. It's with the effective slowdown of overall engineering progress. The first flight of the Boeing 707 was in 1957, a full 54 years ago. 54 years earlier was 1903 and that was the year of the first Wright Brothers' flight! The 707 is different from today's jetliners to an infinitesimal degree as compared to what took off from Kitty Hawk! What we see as such great progress since 1957 really isn't that great. We had color tv then, but now we have color tv that's flatter. Whoopdee do. In 1903, Marconi hadn't yet received his patent for radio! I'm sure that if there was a way to graph "overall effective" progress since the Industrial Revolution, the last half century would be a slow droop down on the curve.

As you very correctly state, business plays a part. How many millions of manhours are wasted programming new and different ways for 13 year olds to chop heads off monsters all night long which could have been redirected to improving the state of humanity? But we're in a very strange world where the benefit to the community is secondary to making a buck no matter what violent or sexual perversions have to be portrayed. Everything is sacred as long as it turns a profit and we all play merrily along because we have bills to pay. And, of course, no one has the cojones to stand up and scream from the rooftops that we're all going to hell in a handbasket. "The greatest trick the devil ever played was convincing the world that he doesn't exist" - Baudelaire

HW2050Plus: What I am anxiously waiting for is the SB-EN vs BD Photoshop benchmarks. I wanna see how those filters grind on 1GB+ files. To me (and possibly me alone) that's pretty well the only benchy that I give a good darn about!

 

[Martimus]

Consider yourself beered! I'll email you a two-four of Molson Canadian right away. But remember that I, like the late great MJ, am doing it for the children!



My beef is not so much with MIPS or gigaflops. It's with the effective slowdown of overall engineering progress. The first flight of the Boeing 707 was in 1957, a full 54 years ago. 54 years earlier was 1903 and that was the year of the first Wright Brothers' flight! The 707 is different from today's jetliners to an infinitesimal degree as compared to what took off from Kitty Hawk! What we see as such great progress since 1957 really isn't that great. We had color tv then, but now we have color tv that's flatter. Whoopdee do. In 1903, Marconi hadn't yet received his patent for radio! I'm sure that if there was a way to graph "overall effective" progress since the Industrial Revolution, the last half century would be a slow droop down on the curve.

As you very correctly state, business plays a part. How many millions of manhours are wasted programming new and different ways for 13 year olds to chop heads off monsters all night long which could have been redirected to improving the state of humanity? But we're in a very strange world where the benefit to the community is secondary to making a buck no matter what violent or sexual perversions have to be portrayed. Everything is sacred as long as it turns a profit and we all play merrily along because we have bills to pay. And, of course, no one has the cojones to stand up and scream from the rooftops that we're all going to hell in a handbasket. "The greatest trick the devil ever played was convincing the world that he doesn't exist" - Baudelaire

HW2050Plus: What I am anxiously waiting for is the SB-EN vs BD Photoshop benchmarks. I wanna see how those filters grind on 1GB+ files. To me (and possibly me alone) that's pretty well the only benchy that I give a good darn about!

I have to admit that I don't agree with your assessments of humanity. Humanity was fine 2 million years ago when we didn't even have written language or highly organized civilization. There was nothing fundamentally wrong with it then, and there is nothing fundamentally wrong with it now.

We are drievn to suceed for the same reasons we were back then. We are inherently unhappy, and nothing we have is ever going to be good enough. There is a reason we aren't still hunting and gathering our food, and living in caves. That would be fine, and we would survive doing so, but our most defining evolutionary trait is that we are inherently unhappy with the status quo. It is a trait that pushes us to make better sources of food, better sources of shelter, better sources of pleasure, better sources of anything really. It is also a trait that punishes those who are not unhappy, as they are looked down on as hippys, vagrants, or just underacheivers.

Humankind not striving to "better" itself is not a step in the wrong direction, nor do I see that as even an option after millions of years of evolution have made striving for something "better" what we are all about.

I laugh every time I see someone make a statement that we are slowing down development or regressing because of what happened "back in the day" when they are talking about maybe 20-50 years ago. Mankind has been around for millions of years, and that small fraction of it is not an accurate depiction of what mankind can or should do.