Will AMD's Bulldozer plow through Intel's Sandy Bridge?

Yeah, I wasn't understanding that nit pick either. "similar performance.., possibly with minor performance improvements." It's still a K10.5 with no L3 cache, and miracles are not expected.

Llano is not going to become much more competitive clock for clock, but as I also mentioned, a speed bump may help close the gap. Still, Llano will only be competitive with Core 2 Duo and Core 2 Quad and I only posted that to dispel the fantasy speculated that Llano was going to compete with Sandy Bridge. Llano isn't much of a mystery on the CPU side. It's apparently nothing to brag about either since AMD has been oddly quiet about performance (contrast with handling of Zacate 3rd party testing and information, which is SOP when AMD has something competitive or better).

It's worth repeating: in the year 2011, AMD's mainstream CPUs are still at Athlon II X2 and Athlon II X4 performance. Doesn't anyone see the problem here? Maybe that's what's so hard to understand, how this is still happening, a nightmare to those who hoped AMD would have turned it around by now.


Athlon II's are super cheap to produce.
 
Its a very small IPC increase, the big deal is the die shink and the fact that Llano is very inexpensive to produce.
Llano is ~220mm^2 for the quad core version. It's a bit more expensive for AMD to produce* dual and quad core Llano models than it is for Intel to produce corresponding core count SB dual or quad core models. Llano has more metal layers (if it follows K10.5, which it seems to), smaller production scale and GF as a middleman adds its own overhead and profit to prices.

I think TI pioneered the model of losing money on every sale, but making it up in volume with the TI-99/4A. Strangely that never caught on with other computer companies. :p

* not counting AMD's 32nm HiK/MG production problems it cited last year as the reason for delaying Llano
 
Do we actually have production cost numbers on the products, or is this just wild speculation?
 
It's not just showing benchmarks or performance figures, they haven't even said anything about the performance at all, except one comparison to Magny Cours processors that was extremely vague. There is literally no information at all.

Obviously you can draw your own conclusions as to what this might mean, but I would think that they would be dropping hints in order to stave off some of the Sandy Bridge purchasing if they had something worth showing off, no?

Is it possible they have something amazing that they want to actually release before Intel finds out what it can do? AMD did that with the 5xxx series GPUs + eyefinity. Maybe they are doing something similar with Buldozer- keep everything hushed until launch and really blow people away. This scenario may not be likely, but I like to hope that the competition will stay tough and keep prices lower and performance improving quickly.
 
Do we actually have production cost numbers on the products, or is this just wild speculation?
Not wild, but yes speculation based on what is known about both chips, including size and manufacturing steps. The point was to counter that somehow the Llano chips with a low price ceiling are somehow "super cheap for AMD to produce" or more profitable for AMD. The assumption in my post was that both GF and Intel are competent manufacturers using similar processes: 32nm HiK-MG, immersion lithography (of course there are still differences like gate first/last, number of steps using immersion lithography and SOI/bulk). Do you have a different starting assumption or something that would throw a monkey wrench in my understanding? To top it off, and not to get snarky, one manufacturer isn't delaying products due to trouble with its manufacturing process.
 
Is it possible they have something amazing that they want to actually release before Intel finds out what it can do? AMD did that with the 5xxx series GPUs + eyefinity. Maybe they are doing something similar with Buldozer- keep everything hushed until launch and really blow people away. This scenario may not be likely, but I like to hope that the competition will stay tough and keep prices lower and performance improving quickly.
Sure, this is possible too of course, I'm just saying that I find it unlikely given how the last 5 years have played out. I would really love to be wrong. I skipped the entire P4 and Core2 generations from Intel using various Athlon processors, but the Core i-series have been dominant.
 
Do you have a different starting assumption or something that would throw a monkey wrench in my understanding?
No I don't have any monkey wrenches, nor would I be revealing them here.

It wouldn't surprise me if Intel can produce an equivalent sized chip for less. I'd say this for historical reasons and that they've already been selling 32nm chips for a year, rather than some vague process analysis; even semiconductor engineers aren't able to accurately predict yields, let alone the internet.
 
Last edited:
I know I'm being optimistic here, but what if AMD is keeping Bulldozer performance numbers so quiet because giving Intel any clues on its performance could make Intel get even more aggressive with their timelines? Intel isn't the same arrogant company it was when Sledgehammer came out. Maybe if they get even hint of a threat from AMD they would be ready to leave them in the dust never to be seen again?
There is also the fact that letting loose some performance numbers now could mean they may lose some sales now as well. If the performance numbers were good, then people (like enterprise customers) may hold out for the new product. If the performance numbers were bad, they just might give up on AMD altogether and go with Intel. The way AMD is keeping its head above water now financially, they probably don't want to risk either to prevent money from flowing in.
 
I know I'm being optimistic here, but what if AMD is keeping Bulldozer performance numbers so quiet because giving Intel any clues on its performance could make Intel get even more aggressive with their timelines? Intel isn't the same arrogant company it was when Sledgehammer came out. Maybe if they get even hint of a threat from AMD they would be ready to leave them in the dust never to be seen again?

Could be. I don't discount that as a possibility.
 
I know I'm being optimistic here, but what if AMD is keeping Bulldozer performance numbers so quiet because giving Intel any clues on its performance could make Intel get even more aggressive with their timelines? Intel isn't the same arrogant company it was when Sledgehammer came out. Maybe if they get even hint of a threat from AMD they would be ready to leave them in the dust never to be seen again?
There is also the fact that letting loose some performance numbers now could mean they may lose some sales now as well. If the performance numbers were good, then people (like enterprise customers) may hold out for the new product. If the performance numbers were bad, they just might give up on AMD altogether and go with Intel. The way AMD is keeping its head above water now financially, they probably don't want to risk either to prevent money from flowing in.

You know when you say that you're probably right. I hope AMD kicks Intel's balls up between their tonsils with Bulldozer and Bobcat and they never knew what hit them. We were seeing all the Phenom II 940 X4 test months before it's release on youtube and then Maki and Kingpin broke a world record with it in late Dec08- early Jan09 in Finland. Intel had alot to study and make up for then.
 
Llano is ~220mm^2 for the quad core version. It's a bit more expensive for AMD to produce* dual and quad core Llano models than it is for Intel to produce corresponding core count SB dual or quad core models. Llano has more metal layers (if it follows K10.5, which it seems to), smaller production scale and GF as a middleman adds its own overhead and profit to prices.

I think TI pioneered the model of losing money on every sale, but making it up in volume with the TI-99/4A. Strangely that never caught on with other computer companies. :p

* not counting AMD's 32nm HiK/MG production problems it cited last year as the reason for delaying Llano


I swallowed my foot on that one,so let me clarify what I should have said. Llano is not super cheap to manufacture. It was relatively inexpensive to design or put another way did not require anywhere near the design investment cost of bulldozer. Keeping a competitive presence in the market is often more important than making great profit margins on a small quantity of chips.
 
Personally, if I was in AMD's position right now, I would hold back numbers as well. Especially if Bulldozer is as fast or faster than the new Sandy Bridge cpu's.

As mentioned before, if they tipped their hat too early, that would give Intel plenty of time to readjust their cpu release timeline or massively adjust their cpu pricing. Or both.
 
Conspiracy theories aside, we just don't release benchmarks prior to launch. And I personally know that I have never approved a "leaked" benchmark.

I can't speak for the other divisions, but I know that it is not policy to release benchmarks prior to launch. There are very specific business reasons why letting the cat out of the bag, regardless of what the performance is, can be a bad thing.

Obviously there are those who keep pushing becase they want to know, but from a business standpoint there is more reason not to release benchmarks prior. Did intel release any offical benchmarks on SB prior to launch? No. Were there "leaked" benchmarks? Probably.

We do occasionally do some demos prior to launch (like with the Zacate last September) but that was not a benchmark, it was a comparison, nothing was audited.
 
Conspiracy theories aside, we just don't release benchmarks prior to launch. And I personally know that I have never approved a "leaked" benchmark.

Not trying to attack you personally on this but when there's good news to report it tends to leak. If Bulldozer ends up being another Phenom it's going to bulldoze itself.
 
Last edited:
Not trying to attack you personally on this but when there's good news to report it tends to leak. If Bulldozer ends up being another Phemon it's going to bulldoze itself.

Then i'll just call it "Plow Boy" and keep running my Phenom II's :p
 
Then i'll just call it "Plow Boy" and keep running my Phenom II's :p

I've bought plenty of AMD CPUs, I have nothing against them. But the last time I bought an AMD CPU it commanded $800 and in those days AMD was acutally make a little cash there for a while. AMD DESPERATELY needs to raise it's ASPs.
 
I've bought plenty of AMD CPUs, I have nothing against them. But the last time I bought an AMD CPU it commanded $800 and in those days AMD was acutally make a little cash there for a while. AMD DESPERATELY needs to raise it's ASPs.

I remember those days, I had an FX-60 from back then. The NF4 days. You know that was only like five-six years ago now that I think about it!
 
I remember those days, I had an FX-60 from back then. The NF4 days. You know that was only like five-six years ago now that I think about it!

Yeah, I got the Athlon 64 X2 4800+ in December 2005 and replaced it with a Q6600 about in the middle of 2007. I do have to say the jump from the 4800+ to the Q6600 was HUGE. That was one of the greatest upgrades I ever saw in terms of performance in only 18 months.
 
I've bought plenty of AMD CPUs, I have nothing against them. But the last time I bought an AMD CPU it commanded $800 and in those days AMD was acutally make a little cash there for a while. AMD DESPERATELY needs to raise it's ASPs.

Back then I had an Athlon 3200+, 90nm, Athlon 3800+, dual Opteron 246's and dual Opteron 254's.
 
I just got rid of a FX60/x2 4800/x2 3800 and upgraded to i7 930 and to be honest, apart from gaming, I dont notice the difference between the 4800/fx60 and the i7,only when I game do i notice the difference.

The pc boots just as fast as it did before, it opens up my apps just as quickly as it did before, it plays 1080p vids just as good as it did before, so although I know the jump between fx60 to i7 930 is huge, I simpyl dont notice it often because I rarely game now, even though I went nuts during the steam holiday sale.

For what I do with exception of gaming, the 48/60 were perfect for me, but I had to upgrade as it was too cheap not too.
 
Back then I had an Athlon 3200+, 90nm, Athlon 3800+, dual Opteron 246's and dual Opteron 254's.

These were all good CPUs and AMD was the CPU of choice for high end gaming. No doubt AMD would LOVE to be in that position again.
 
We do occasionally do some demos prior to launch (like with the Zacate last September) but that was not a benchmark, it was a comparison, nothing was audited.
Will there be a public Demo of Bulldozer like that of Zacate (either desktop or server versions) before the Q2 release? It would be really cool if AMD would give us another taste.
 
My last 2 systems before my current one were both Core 2 Duo chips. I loved them But one thing I truly missed was being about to just let my PC run for weeks on end when I owned my last AMD 3700+. I have no Idea how the Core i series Chips are as I have never owned one. but both of those C2D chips couldn't run more then a couple days without blue screening or needing a restart. There are probably more factors to all of this madness, could be windows 7, could be the memory I chose this time. I have used like brand products in all of my build recently. MSI mainboards, G.skill memory, WD hard drives. It just seems to be that EVERY AMD CPU I have ever owned has been much more stable over longer periods. All of them have been overclock(comparatively). Coming from a stand point of what works best for me is WHAT my money can do for me. Not what is the best possible performance. I'm a full time college student and I work a part time job, but on those budgets being able to enjoy the things I love.

Now on topic, No I don't believe that AMD with their bulldozer will blow straight pass sandy bridge. Although I do believe(could be wrong) that it will be a big ass speed bump for intel to hit a 65mph.

Fusion Is what I feel AMD is putting the most work into. That is a huge market to fill, not only will they be used in netbooks but possibly notebooks if the performance is great enough. The Atom chips that are out now just plan suck.

I forgot something, however I cannot remember what it was. But I don't think :):) is going to bash on me for stating an opinion as his opinions got him in the shitter.
 
My last 2 systems before my current one were both Core 2 Duo chips. I loved them But one thing I truly missed was being about to just let my PC run for weeks on end when I owned my last AMD 3700+. I have no Idea how the Core i series Chips are as I have never owned one. but both of those C2D chips couldn't run more then a couple days without blue screening or needing a restart. There are probably more factors to all of this madness, could be windows 7, could be the memory I chose this time. I have used like brand products in all of my build recently. MSI mainboards, G.skill memory, WD hard drives. It just seems to be that EVERY AMD CPU I have ever owned has been much more stable over longer periods. All of them have been overclock(comparatively). Coming from a stand point of what works best for me is WHAT my money can do for me. Not what is the best possible performance. I'm a full time college student and I work a part time job, but on those budgets being able to enjoy the things I love.

Eh, I've owned my fair share of AMD cpu's (see sig), but the general consensus on the forums seems to be the polar opposite of what you said. There was something very wrong with your C2D systems and I have no idea how you simply lived with what was obviously a hardware problem.:confused:
 
Like I said, there were other factors. I could have been the two processors I purchase, E6550 and E6300. I don't know. I have just had more luck with the two AMD processors around the two intel one. I won't blame the E6300 on it, Ill blame it on the EVGA 680i board which died prematurely. The E6550 had no excuse, I had it on three total board, 1 Gigabyte, 1 MSI, 1 Asus. All three just didn't seem to make it any more stable. Multiple memory changes. It ran just fine other wise, I sold it to a buddy and he couldn't be more pleased with the system overall. He turns it off nightly and all that jazz.

When I build this system I whole heartedly contemplated getting the I5-750, it just didn't make sense for my money as I would have had 350 in the board and processor. I got the MSI/AMD combo for 300 it would have been similar performance but I was tired of intel and tried something new. I like it and have had zero issues and that is fine. The only brand I'm a complete fanboy to is LG which has no game on this subject. So the next system I build could be intel depending on where prices are in comparison to performance. Also where I want to be. It changes every build. Do I debate and wait three years instead of two and build a much higher end system. 2010 was the year of get myself a new desktop and stop using my laptop for main use.
 
Will there be a public Demo of Bulldozer like that of Zacate (either desktop or server versions) before the Q2 release? It would be really cool if AMD would give us another taste.

Not sure what the client plans are. I have plans to do some server demos before launch.
 
You know when you say that you're probably right. I hope AMD kicks Intel's balls up between their tonsils with Bulldozer and Bobcat and they never knew what hit them. We were seeing all the Phenom II 940 X4 test months before it's release on youtube and then Maki and Kingpin broke a world record with it in late Dec08- early Jan09 in Finland. Intel had alot to study and make up for then.

I hope not. Hopefully we'll see really similar performance. This way they'll be forced to compete with one another on price. Then we win.
 
That clears it up for me at least. AMD does NOT release benchmarks or allow testers from sites to preview performance at AMD, except when they DO allow it. And by strange coincidence, AMD just happens to release benchmark figures or allow testers from sites access to hardware when it has something competitive or better, and do neither when that's not the case. :p

I remembered something that TR or AT or ET (4 years ago, forget the exact site) wrote a little into the Conroe era about getting pre-release access to hardware. The review mentioned that during the A64/X2 era when Intel was often very uncompetitive with P4 and PD, that it (Intel) still sent out hardware and press stuff as it always had, when it full well knew its products would lose the benchmarks. I mean by the time the higher speed A64 and mid-range X2 were released, Intel was totally outclassed in its corresponding products. The writer lamented that AMD had started restricting info and products for review when it knew it would lose. You can go back through the history of product previews and see this yourself. To borrow from Charlie, "something changed."

BTW, Llano was shown yesterday at CES, no performance data was released. No access to the hardware was given either. I bet Dirk is still "super excited." :p
 
I hope not. Hopefully we'll see really similar performance. This way they'll be forced to compete with one another on price. Then we win.

I guess.... but id find it halarious if the proposed situation occures. Teach intel a thing or two about the K series chips, and the dual socket bullshit.
 
I hope it does well, I want it to do well. But I can't help but feel it won't be competitive to Sandybridge. If they keep power consumption and heat down, even if it's a little slower I would seriously consider it.
 
That clears it up for me at least. AMD does NOT release benchmarks or allow testers from sites to preview performance at AMD, except when they DO allow it. And by strange coincidence, AMD just happens to release benchmark figures or allow testers from sites access to hardware when it has something competitive or better, and do neither when that's not the case. :p

I remembered something that TR or AT or ET (4 years ago, forget the exact site) wrote a little into the Conroe era about getting pre-release access to hardware. The review mentioned that during the A64/X2 era when Intel was often very uncompetitive with P4 and PD, that it (Intel) still sent out hardware and press stuff as it always had, when it full well knew its products would lose the benchmarks. I mean by the time the higher speed A64 and mid-range X2 were released, Intel was totally outclassed in its corresponding products. The writer lamented that AMD had started restricting info and products for review when it knew it would lose. You can go back through the history of product previews and see this yourself. To borrow from Charlie, "something changed."

BTW, Llano was shown yesterday at CES, no performance data was released. No access to the hardware was given either. I bet Dirk is still "super excited." :p

I think AMD stands to loose more than Intel due to early release benchmarks. At least 8 out of 10 people will buy (a computer powered by) Intel over AMD regardless of the benchmarks. I am talking general population, not enthusiasts.

In other words people who buy AMD are more likely to read benchmarks and releasing Bulldozer figures now if they show great performance could cannibalize current Phenom II/MagniCore sales. I want to see the figures as much as the next guy but I can understanding why AMD is holding them back.

Also even Q2 production is a rumor. There is conflicting information that Llano may be released before Bulldozer which could come out later this year. GloFlo's 32nm SOI process is brand new, question is what kind of benchmarks could they even deliver this many months before the product is finalized, even if they wanted to?
 
Last edited:
@sirmo: I know, its fascinating how uninformed customers are. For instance, I helped a friend build a pc recently, and when picking out parts he told me, "I want the fastest quad core Pentium that Intel makes." I just stared at him lol. Intel made a good name for itself, so much that its fighting its own branding.
 
I guess.... but id find it halarious if the proposed situation occures. Teach intel a thing or two about the K series chips, and the dual socket bullshit.

The dual socket thing is no different than having separate sockets for the Opteron and the Athlon or the Phenom and the Opteron. The LGA1366 based setups aren't really designed for anything but server and workstation use. However the motherboard manufacturers marketed them to us and built them with enthusiasts in mind. I don't think that was Intel's intent.
 
The dual socket thing is no different than having separate sockets for the Opteron and the Athlon or the Phenom and the Opteron. The LGA1366 based setups aren't really designed for anything but server and workstation use. However the motherboard manufacturers marketed them to us and built them with enthusiasts in mind. I don't think that was Intel's intent.
I'm not sure I agree with this. They have offered the Nehalem architecture in both consumer (Core i7) and server (Xeon) lines, I can't imagine they didn't intend for the Core i7 series to be picked up by enthusiasts. If they had intended it to be for workstation and server setups, why both releasing two separate products and not just go with the Xeons?

I think the people who got screwed honestly were the LGA1156 adopters. How long was that socket viable? A year? I've had my 920 now for 2 years, by the time LGA2011 it will be closer to three years of availability which isn't bad at all honestly.
 
tbh i think it will always stay at the rate of save more money for decent performance=AMD, best performance for more $$ = intel
 
tbh i think it will always stay at the rate of save more money for decent performance=AMD, best performance for more $$ = intel
Except that's only true in very specific instances now. Intel has a good budget line as well. AMD is cheaper if you don't want to replace a motherboard, otherwise it's mostly a wash.
 
Except that's only true in very specific instances now. Intel has a good budget line as well. AMD is cheaper if you don't want to replace a motherboard, otherwise it's mostly a wash.

Your right, "at this time" being the key phrase though. For the majority of last year, and I would say at least 9 months out of 12 you could say confidently AMD was the better budget build. Right now though we're in lingo while we wait for AMD to release there next chip, not a higher clock speed on same chip. So it's still safe to say usually AMD is the better budget build but for the next couple month's Intel for the most part is better in both areas.
 
I'm not sure I agree with this. They have offered the Nehalem architecture in both consumer (Core i7) and server (Xeon) lines, I can't imagine they didn't intend for the Core i7 series to be picked up by enthusiasts. If they had intended it to be for workstation and server setups, why both releasing two separate products and not just go with the Xeons?

I think the people who got screwed honestly were the LGA1156 adopters. How long was that socket viable? A year? I've had my 920 now for 2 years, by the time LGA2011 it will be closer to three years of availability which isn't bad at all honestly.

I think they were intended for upper end enthusiast boxes. Honestly I think the architecture was really intended for servers with Intel deciding they could capitalize on the enthusiasts by providing it to them. I think we were more of an after thought and LGA1366 was never intended to be a mainstream part. Core i7 branded chips sort of bridge the gap between the Core ix line and the Xeon's in high end workstations and enthusiast boxes.
 
Your right, "at this time" being the key phrase though. For the majority of last year, and I would say at least 9 months out of 12 you could say confidently AMD was the better budget build. Right now though we're in lingo while we wait for AMD to release there next chip, not a higher clock speed on same chip. So it's still safe to say usually AMD is the better budget build but for the next couple month's Intel for the most part is better in both areas.

Intel generally provides their processors at competitive pricing these days. The only thing that really pushes things into AMD's favor are slightly cheaper motherboard chipsets and therefore slightly cheaper boards. Though that gap has been closing. Of course now there is the Phenom II X6 which gives you a boost in some applications due to the extra cores but this advantage isn't really leveraged enough by applications.
 
Yeah, Intel essentially licenses the chipset with the server or workstation as its focus, leaving the actual "workstation," "server", or "gaming rig" branding up to the motherboard manufacturers' discretion.
Of course now there is the Phenom II X6 which gives you a boost in some applications due to the extra cores but this advantage isn't really leveraged enough by applications.
Which is pretty much matched by Intel's hyper-threaded quads. Which leaves the likes of the 980x commanding such a premium.
 
Back
Top