Bulldozer press kit pictures

[LOL_Wut_Axel]

Just because Intel or some other x86 manufacturer has not made a processor within the same power envelope as ARM does not mean it cannot be done. Also just because ARM hasn't reached the computing performance of x86 also does not mean that it cannot be done. Remember, absence of evidence is not the same as evidence of absence.

Consider that 20 years ago, if you were to tell people that MIPS would be relegated to routers instead of high performance computing, you'd be the laughing stock of everyone around. And yet here we are: MIPS relegated to routers and x86 in the majority of supercomputers.

Because right then and there architectures were in their early inception and most of what made differences were new innovations. Right now, most ways to make architectures fast have already been discovered. As it turns out, more innovation came from RISC in the early days, which gave it a speed advantage. As architectures turned more complex and cost became a bigger factor, x86 took over. PowerPC, for example, hit a wall when it came to performance/watt.

Again, if what you're saying is possible, it'd already be done. Intel has plenty of liquid assets and they'd have made an x86 CPU that can compete with ARM in low power already.

There's no SPARC T4-like Sandy Bridge or vice versa because it's not that simple. The architectures are designed for different purposes. In this case, again, SPARC T4 is made for hugely parallel workloads but compared to something like Sandy Bridge it's left in the dust in single-threaded, which is a very important factor for desktops and laptops.

 

[podspi]

http://www.tigerdirect.com/applicati...?EdpNo=1375893

Saw it at overclock.net

Sabertooth 990 FX bundled with an FX-6100

not shipping for 7-21 days

I can't believe nobody is commenting on this. Good find! Looks like the rumored prices for BD are rightish. The launch must be soon if we can preorder from somewhere that isn't BLT.

 

[Voo]

Again, if what you're saying is possible, it'd already be done. Intel has plenty of liquid assets and they'd have made an x86 CPU that can compete with ARM in low power already.

Yep, I'm sure Intel is horribly worried about not competing in the smartphone market - I mean who wouldn't love a field with extremely slim margins and strong competition? Just look at how much love Atom has gotten so far - no large engineering team, nodes behind their more important architectures and just look at how much they worry about MeeGo. Not the usual signs of an extremely focused approach to the low power market. And then developing a completely new architecture and not just trying to scale down an existing one, takes its time and costs lots of money and the other solution just doesn't work (IBM showcased that nicely enough).
For whatever reason many people are just way too focused on the used ISA when looking at power figures - there are decisions that influence power consumption much more than any ISA ever could and those are basically independent of the ISA (and the ARM ISAs aren't that great from a energy efficiency pov either..)


And a simple fact remains: ARM still has to show that they can develop a high performance chip that can compete with a modern x86 architecture such as SB. Their engineering team is much smaller than those of Intel or IBM, with less experience and even Intel, IBM, AMD all had failures on their part - chips that looked great on paper but didn't turn out well at all.

PS: And there's no SPARC like x86 architecture mostly because the whole market for such CPUs is minimal - you don't compete with niche products. Why would intel want to do that?

 

[classy]

http://www.tigerdirect.com/applicati...?EdpNo=1375893

Saw it at overclock.net

Sabertooth 990 FX bundled with an FX-6100

not shipping for 7-21 days

That's quite a bit more than rumored. The 6100 was suppose to be like $155. Maybe its a premium for early adoption.

 

[Edrick]

Intel were using an IMC back in 386 days.

Really? How do you figure that?

 

[jacktesterson]

Im buying one

 

[sm625]

Why is it that people expect AMD's CPU's to rock?

Because they own a gpu and have an experienced driver team at their disposal. There is no reason and no excuse for BD to not completely dominate in gaming when combined with HD6XXX (or HD7XXXX) gpu and catalyst driver. People need to look at the big picture. It may be a server chip but there is no reason they cannot add a few instructions to help with DirectX, and to give the Catalyst team something to play with. Now with 8 cores at their disposal, they can do all sorts of prerendering and pre filtering. Hell they could figure out a way to cheat and double or even triple your framerate without it being noticeable to the end user, ie with no loss of detail. Its been done before. You guys dont know your history if you think it aint. Look at normal mapping or hardware occlusion. I'm not saying I know for a fact they've done anything like this, I'm just saying they should be doing this. It is what I expect. I dont care if AMD cannot beat intel at cinebench. But they absolutely should dominate at BF3. If they dont then they need to sell off ATI because they wasted billions of dollars.

 

[sequoia464]

I can't believe nobody is commenting on this. Good find! Looks like the rumored prices for BD are rightish. The launch must be soon if we can preorder from somewhere that isn't BLT.

Thanks, but I didn't actually find this - a guy by the name of Evil Penguin at overclock posted it initially.

http://www.overclock.net/amd-cpus/791495-bulldozer-blog-live-717.html

 

[Lonbjerg]

Because they own a gpu and have an experienced driver team at their disposal. There is no reason and no excuse for BD to not completely dominate in gaming when combined with HD6XXX (or HD7XXXX) gpu and catalyst driver. People need to look at the big picture. It may be a server chip but there is no reason they cannot add a few instructions to help with DirectX, and to give the Catalyst team something to play with. Now with 8 cores at their disposal, they can do all sorts of prerendering and pre filtering. Hell they could figure out a way to cheat and double or even triple your framerate without it being noticeable to the end user, ie with no loss of detail. Its been done before. You guys dont know your history if you think it aint. Look at normal mapping or hardware occlusion. I'm not saying I know for a fact they've done anything like this, I'm just saying they should be doing this. It is what I expect. I dont care if AMD cannot beat intel at cinebench. But they absolutely should dominate at BF3. If they dont then they need to sell off ATI because they wasted billions of dollars.

Oh I know the history of how AMD wrestled/cheated their way to an x86 license, how their CPU trailed Intel (untill they aquired NexGen and Intel slipped up with Netburst.)...so don't go there, will only backfire.

But I am amazed at your stance about just "addeding a few instructions to help DirectX"...it is not that simple.

 

[The Ultimate]

Wait. Let me get this straight. AMD does the hard work of coming up with a viable 64-bit extension to x86, and you chide them for innovating? Even Intel had to follow suit and copy
AMD's extensions.

Would you have preferred the entire industry to move to Itanium? (shudder) It has a whole host of other issues with architecture.

It's my opinion that you are nothing but a troll.

You are right, with such posts, he showed his true fanboy colors, seems that he would love to pay $300 for a Pentium G9650. I think that competition drives innovation, thanks to that, we had Core 2 instead of Netbust, we had Phenom II instead Smithfield, we had Nehalem/Sandy Bridge instead of Thuban, plus x64, IMC, Fusion and Intel's equivalent, Zacate instead of Atom etc etc.

We shall see what Bulldozer is gonna bring, I think that it will be competitive against the 2600K, it will not be faster for now, evenly matched with the slight edge for 2600K in overall performance and the edge to Bulldozer in heavily multithreaded apps.

 

[sm625]

Oh I know the history of how AMD wrestled/cheated their way to an x86 license, how their CPU trailed Intel (untill they aquired NexGen and Intel slipped up with Netburst.)...so don't go there, will only backfire.

But I am amazed at your stance about just "addeding a few instructions to help DirectX"...it is not that simple.

So AMD acquires NexGen and then dominates the cpu world for half a decade, after a 5-7 year delay. Simple. But AMD acquiring ATI and then dominating the gaming world after a 5-7 year delay? Now that is inexplicably complicated and impossible. Is that what you're trying to say? Or are you saying that it is not possible to remove any of the millions of cycles per second of video card driver overhead by using just a few well placed instructions and / or a few idle cpu cores?

 

[Lonbjerg]

Wait. Let me get this straight. AMD does the hard work of coming up with a viable 64-bit extension to x86, and you chide them for innovating? Even Intel had to follow suit and copy
AMD's extensions.

Would you have preferred the entire industry to move to Itanium? (shudder) It has a whole host of other issues with architecture.

It's my opinion that you are nothing but a troll.

If you thinking dragging all of x86's legacy along to the future is a really, really, really, really, really, really, really, really bright thing to do, I can understand why you don't get it.

But x86 was slated to die...and be replaced by a better ISA.

IA64 would have been much better for the futue than X86-64.
Both in terms of ISA(cluttter)...but also in CPU design.

The problem with x86 will only grow...all because AMD did us an "favour" :|

 

[drizek]

Really? How do you figure that?

The northbridge was soldered onto the motherboard. Integrated graphics, integrated memory controller.

 

[jhu]

If you thinking dragging all of x86's legacy along to the future is a really, really, really, really, really, really, really, really bright thing to do, I can understand why you don't get it.

But x86 was slated to die...and be replaced by a better ISA.

IA64 would have been much better for the futue than X86-64.
Both in terms of ISA(cluttter)...but also in CPU design.

Hahahaha... oh wait, you're serious.

The problem with x86 will only grow...all because AMD did us an "favour" :|

What problems would those be exactly? Fast computing at affordable prices? Let me ask you. What kind of processor is in your main computer at home?

 

[LondonBurning]

Heck yeaaah, been waiting a long time.

 

[Edrick]

The northbridge was soldered onto the motherboard. Integrated graphics, integrated memory controller.

Oh dear god. IMC as in on die as per nehalam. Alpha was doing this in the 90s.

 

[LOL_Wut_Axel]

Yep, I'm sure Intel is horribly worried about not competing in the smartphone market - I mean who wouldn't love a field with extremely slim margins and strong competition? Just look at how much love Atom has gotten so far - no large engineering team, nodes behind their more important architectures and just look at how much they worry about MeeGo. Not the usual signs of an extremely focused approach to the low power market. And then developing a completely new architecture and not just trying to scale down an existing one, takes its time and costs lots of money and the other solution just doesn't work (IBM showcased that nicely enough).
For whatever reason many people are just way too focused on the used ISA when looking at power figures - there are decisions that influence power consumption much more than any ISA ever could and those are basically independent of the ISA (and the ARM ISAs aren't that great from a energy efficiency pov either..)


And a simple fact remains: ARM still has to show that they can develop a high performance chip that can compete with a modern x86 architecture such as SB. Their engineering team is much smaller than those of Intel or IBM, with less experience and even Intel, IBM, AMD all had failures on their part - chips that looked great on paper but didn't turn out well at all.

PS: And there's no SPARC like x86 architecture mostly because the whole market for such CPUs is minimal - you don't compete with niche products. Why would intel want to do that?

http://en.wikipedia.org/wiki/Mobile_Internet_device#Medfield_platform_.282011.29

 

[Lonbjerg]

Hahahaha... oh wait, you're serious.

Someone has to be, the internet is full of clowns.

What problems would those be exactly? Fast computing at affordable prices?

The x86 architecture is disastrous.
Why?
Mainly it's because it's old.
x86 started its life as something fairly simple in Intel's 8086 chip, when IBM only had interest in 8088.
And has been growing ever since.
To it's defence it kept fairly straight line of backward compatibility but also picked up new "baggage". in each era.
Add to that the multiple of coprocessor extentions(MMX/SSE, x64 ect.).
With all of this "baggage" the x86 instruction set is now a beast, a monster that requires a decent amount of effort just to tame before you can even get to processing. Every new x86 chip must live with all the sins of every past x86 chip.
I could go on and on, but I fear it's wasted.
You are entranched in x86...just like some people were entrenched in RISC some years ago...and the problem will only grow with time.
So I don't care what x86 did in the past...I care about the future.

Let me ask you. What kind of processor is in your main computer at home?

Nice fallacy...but irrelevant.

 

[psolord]

Because they own a gpu and have an experienced driver team at their disposal. There is no reason and no excuse for BD to not completely dominate in gaming when combined with HD6XXX (or HD7XXXX) gpu and catalyst driver. People need to look at the big picture. It may be a server chip but there is no reason they cannot add a few instructions to help with DirectX, and to give the Catalyst team something to play with. Now with 8 cores at their disposal, they can do all sorts of prerendering and pre filtering. Hell they could figure out a way to cheat and double or even triple your framerate without it being noticeable to the end user, ie with no loss of detail. Its been done before. You guys dont know your history if you think it aint. Look at normal mapping or hardware occlusion. I'm not saying I know for a fact they've done anything like this, I'm just saying they should be doing this. It is what I expect. I dont care if AMD cannot beat intel at cinebench. But they absolutely should dominate at BF3. If they dont then they need to sell off ATI because they wasted billions of dollars.

What could they do with the FX 4100 then?

And what about the Nvidia users?

 

[Lonbjerg]

So AMD acquires NexGen and then dominates the cpu world for half a decade, after a 5-7 year delay. Simple. But AMD acquiring ATI and then dominating the gaming world after a 5-7 year delay? Now that is inexplicably complicated and impossible. Is that what you're trying to say? Or are you saying that it is not possible to remove any of the millions of cycles per second of video card driver overhead by using just a few well placed instructions and / or a few idle cpu cores?

Domnitated?
Then Intel must be killing AMD right now, because the gap today is larger than the Netburst era?

And a 5 year (I'll be generous and give you 5 years) anomely over a period from 1969 to 2001 isn't impressive, quite the opposite...and it only happend becuase AMD got NexGen og Intel fumbled the ball themselfes.

For AMD to gain headway in the GPU market, NVIDIA needs to fumble...and the 5xx series was no fumble.

So AMD needs it's competation to fail in order to gain...that is not a sound strategy.

And don't compare that of making a CPU with making a GPU...diffrent beasts...different rules.

 

[Lonbjerg]

You are right, with such posts, he showed his true fanboy colors, seems that he would love to pay $300 for a Pentium G9650. I think that competition drives innovation, thanks to that, we had Core 2 instead of Netbust, we had Phenom II instead Smithfield, we had Nehalem/Sandy Bridge instead of Thuban, plus x64, IMC, Fusion and Intel's equivalent, Zacate instead of Atom etc etc.

We shall see what Bulldozer is gonna bring, I think that it will be competitive against the 2600K, it will not be faster for now, evenly matched with the slight edge for 2600K in overall performance and the edge to Bulldozer in heavily multithreaded apps.

Besides you cheap trolling, did you have anything relevant to say about the x86 ISA?

 

[busydude]

Besides you cheap trolling, did you have anything relevant to say about the x86 ISA?

Does it have anything to do with the thread at hand? You are the first to derail this thread.. LOL.

 

[Despoiler]

Does it have anything to do with the thread at hand? You are the first to derail this thread.. LOL.

This.

 

[Lonbjerg]

Does it have anything to do with the thread at hand? You are the first to derail this thread.. LOL.

Bulldozer runs x86-64...and beltbuckles...you post adresses neither.

 

[jhu]

Someone has to be, the internet is full of clowns.

The x86 architecture is disastrous.
Why?
Mainly it's because it's old.
x86 started its life as something fairly simple in Intel's 8086 chip, when IBM only had interest in 8088.
And has been growing ever since.
To it's defence it kept fairly straight line of backward compatibility but also picked up new "baggage". in each era.
Add to that the multiple of coprocessor extentions(MMX/SSE, x64 ect.).
With all of this "baggage" the x86 instruction set is now a beast, a monster that requires a decent amount of effort just to tame before you can even get to processing. Every new x86 chip must live with all the sins of every past x86 chip.
I could go on and on, but I fear it's wasted.
You are entranched in x86...just like some people were entrenched in RISC some years ago...and the problem will only grow with time.
So I don't care what x86 did in the past...I care about the future.

So what alternative are you proposing? I ask because POWER, SPARC, and MIPS have the same issues you mentioned. IA64 is a nightmare, and ARM also has plenty of idiosyncrasies.

Nice fallacy...but irrelevant.

It is neither a fallacy nor irrelevant. If you're going to sit there saying how absolutely aweful x86 is, then you need to put your money where your mouth is and use non-x86 as your main home computer. It's easily done. But are you doing it, or are you too ashamed to admit you use x86?