Rumour: Bulldozer 50% Faster than Core i7 and Phenom II.

[Idontcare]

That could be the writer's perception on the situation, a little bit of pessimism along with the good news. Any more delays to Bulldozer and it might ruin their chances of competing against Intel. I am more interested in this part of the article which suggests that there will be more problems for AMD if they were to delay their launch.

In a "delay to third quarter 2011" scenario, AMD would face an aggravated competitive situation, with Intel having released speed bump versions of their high-end CPUs and the upcoming enthusiast platform featuring current generation sexa- and possibly octa-code CPUs based on socket 2011 around the corner. Now you might understand why motherboard manufacturers might not be very happy about a possible delay. It would make AMD motherboard sales go down the drain.

The logic here does not make sense to me.

AMD intends to compete with Intel in the third quarter regardless whether they launch BD in the 2nd or 3rd quarter.

It's not like they get to avoid the 3rd quarter or Intel's speedier chips if they release in Q2, the 3rd quarter is going to happen regardless what happens in Q2.

 

[classy]

Man this athlon stuff all over again. I don't care if its 2nd or 4th for that matter. When they release the product if its good then people will buy it. God you would think after 10+ years all these AMD death rattles would cease. I still see people arguing over stuff like IPC and clock for clock, which hasn't meant much of anything since AMD and Intel each began to use different sockets. This is the way AMD is as a company and always has been.

 

[Idontcare]

God you would think after 10+ years all these AMD death rattles would cease.

You will notice who, or rather "what", is doing the death rattle.

Meaning its a business. BSN, Fudz, S/A, etc etc...they all make $$$ to churn out baseless crap that people click on and read.

Its no conspiracy, I'm convinced neither Intel nor AMD actively seed this kind of subterfuge; rather, I really do think it is simply a cast of enterprising individuals such as Fudo, Theo, Charlie and others who, like the Paparazzi and The SUN, are simply churning this BS for the sake of pocketing a little coin from the ad revenues.

The death rattles will never cease so long as there is money to be made in shaking that rattle. :\

 

[dma0991]

The logic here does not make sense to me.

AMD intends to compete with Intel in the third quarter regardless whether they launch BD in the 2nd or 3rd quarter.

It's not like they get to avoid the 3rd quarter or Intel's speedier chips if they release in Q2, the 3rd quarter is going to happen regardless what happens in Q2.

I understand that eventually Intel will take the lead whether AMD decides to release it early or not but if AMD is a direct competitor to SB then a slightly early launch might just improve AMDs chances of selling as it might be the top dog at that moment.

Now if AMD were to wait till Q3, enthusiasts would have a choice to go with SB-E as an alternative to the 8 core Bulldozer. Any product that is fit for the market without any problems should be launched ASAP as waiting for no reason will not increase their sales revenue anyway.

Page 100 already.

 

[classy]

You will notice who, or rather "what", is doing the death rattle.

Meaning its a business. BSN, Fudz, S/A, etc etc...they all make $$$ to churn out baseless crap that people click on and read.

Its no conspiracy, I'm convinced neither Intel nor AMD actively seed this kind of subterfuge; rather, I really do think it is simply a cast of enterprising individuals such as Fudo, Theo, Charlie and others who, like the Paparazzi and The SUN, are simply churning this BS for the sake of pocketing a little coin from the ad revenues.

The death rattles will never cease so long as there is money to be made in shaking that rattle. :\

ID some of these discussions to me just don't make sense. Comparing AMDs new architecture to its previous one makes sense. But to compare AMD to Intel is completely irrelvant because the only thing that seperates them is the End Result and how much is it. They are different and have been for a very very long time. Each company believes in a certain philosophy so of course they are going to be different. Then you read these death rattles, but AMD has always been a company who wants a peice of the pie, not own the whole thing.

They have always been happy with providing a competitive product. Athlon gave them the performance crown, but the company appears to be happy being competitive even if not leading. So all these notions AMD is going to die without winning the crown are silly. And in the 12-14 years of following the changes in cpus and the directions that these two have taken cements the fact that each has different goals.

 

[Idontcare]

I understand that eventually Intel will take the lead whether AMD decides to release it early or not but if AMD is a direct competitor to SB then a slightly early launch might just improve AMDs chances of selling as it might be the top dog at that moment.

Now if AMD were to wait till Q3, enthusiasts would have a choice to go with SB-E as an alternative to the 8 core Bulldozer. Any product that is fit for the market without any problems should be launched ASAP as waiting for no reason will not increase their sales revenue anyway.

Page 100 already.

Well, tis true, tis always true, that time-to-market is money and to be sure AMD would have stood better chance of making more money on Bulldozer had they released it in Q1 than Q2 than Q3, etc etc.

So in that holistic sense of a marketing analysis, to be sure a delay to Q3 is undesired and counter-productive, but such a delay in and of itself is not the reason for the undesirableness of such an outcome.

Such a delay would not bring with it any more of a loss in profits in Q3, they'd merely not have earned the stacks of coin that were available to be earned in Q2.

Or to put it differently, all that is lost is the opportunity to make money selling BD's in Q2 should a delay to Q3 occur. The money they stood to make in Q3 is not suddenly put in jeopardy just because of a delay in BD to Q3.

In other words I think some of these "delay is uber bad" articles are misplacing where the bad comes into the accounting. The bad is merely lost opportunity on sales that could have occurred in the past, it does not miraculously steal opportunity and sales from the future.

ID some of these discussions to me just don't make sense. Comparing AMDs new architecture to its previous one makes sense. But to compare AMD to Intel is completely irrelvant because the only thing that seperates them is the End Result and how much is it. They are different and have been for a very very long time. Each company believes in a certain philosophy so of course they are going to be different. Then you read these death rattles, but AMD has always been a company who wants a peice of the pie, not own the whole thing.

They have always been happy with providing a competitive product. Athlon gave them the performance crown, but the company appears to be happy being competitive even if not leading. So all these notions AMD is going to die without winning the crown are silly. And in the 12-14 years of following the changes in cpus and the directions that these two have taken cements the fact that each has different goals.

The death rattles were always unrealistic anyways. Worst case scenario for AMD is they do a DEC. A more likely scenario is they do a Cyrix. And even more likely still is that they'd do a Hynix.

As I see it, there are asymptotic limits to the boundary conditions of where AMD's future may lie and Via + Matrox represents the lower-bounds of that limit.

To whatever extent Via has a viable x86 business, and to whatever extent Matrox has a viable graphics business, so too does AMD as a worst case end-game.

 

[Soleron]

Or to put it differently, all that is lost is the opportunity to make money selling BD's in Q2 should a delay to Q3 occur. The money they stood to make in Q3 is not suddenly put in jeopardy just because of a delay in BD to Q3.

I agree there's no automatic loss. But OEMs may have planned products for several quarters and with multiple designs in case one supplier fails, and AMD might have needed to hit Q2 to win the design. I don't know how much of a factor this will be; I imagine it was worse for Llano since that is more in the OEM's mainstream price range.

If Llano had launched against SB maybe some of the Intel laptops out now would be AMD laptops, and hence would be making money for AMD for longer than the delay length alone.

 

[Idontcare]

I agree there's no automatic loss. But OEMs may have planned products for several quarters and with multiple designs in case one supplier fails, and AMD might have needed to hit Q2 to win the design. I don't know how much of a factor this will be; I imagine it was worse for Llano since that is more in the OEM's mainstream price range.

If Llano had launched against SB maybe some of the Intel laptops out now would be AMD laptops, and hence would be making money for AMD for longer than the delay length alone.

Definitely there are longer term ramifications that will come from a delay, if one occurs, in terms of credibility and confidence with and from the business partners in future products. Nvidia had that happen with their delayed Fermi release.

Don't get me wrong, I am not trying to claim that a delay in BD is without its downsides, near term and long term, but rather I am merely attempting to dispel this notion that the competitiveness of Bulldozer in the 3rd quarter is somehow dependent on its release occurring in the 2nd quarter.

This claimed cause-and-effect, and only this specific claim of it, is a fallacy and I merely wanted to entertain a discussion exploring such. But all the rest of the accounting downsides to a delay in volume shipments of bulldozer SKU's still applies, without question.

 

[formulav8]

Just like IDC said, a delay to 3rd quarter would still have similar challenges as if BD launched right now and then 3rd quarter came when BD has already been launched and ready.

 

[piesquared]

If in fact there is a delay, which AMD has refuted, then that will give them time to ramp 32nm volume as they need to displace 45nm volume rather quickly. Hitting the market when volume out strips demand is beneficial. Right now there volume is focused on Llano and Interlagos/Valencia, and rightly so. It makes sense to target the high volume and high margin markets first, gaining market share in both notebook and server. AMD already has a decent market presence in desktop with existing offereings, so their primary goal is and should be where the big growth potential exists.

 

[Cerb]

Why is it important to handle peaks

That leaving execution starved, with code that offers high ILP, seems like a silly thing to do, compared to matching the capability end to end, to prevent inefficient execution of efficient code.

at the cost of a less power efficient wider implementation?

A need to guarantee a certain level of performance per independent CPU core, as has been the case up through their current desktop and server CPUs, I would imagine.

 

[chihlidog]

Why has JFAMD not chimed in on this "Delay"?

I'll say it again, just tell us if BD is worth the wait for those of us considering going with SB.

 

[Dresdenboy]

That leaving execution starved, with code that offers high ILP, seems like a silly thing to do, compared to matching the capability end to end, to prevent inefficient execution of efficient code.

The CPU also has to make sure that there is data to process and branches correctly predicted. Otherwise it'll have enough bubbles here and there to catch up with decoding. Also limited x86-code ILP, suboptimally filled dispatch windows and some dispatch/execution constraints (1 branch, 2 loads, 1 store, 1 mul, 1 div per dispatch group etc.) will also cause a little stall here and there.

A need to guarantee a certain level of performance per independent CPU core, as has been the case up through their current desktop and server CPUs, I would imagine.

Sacrificing power efficiency means that the design will use more power to have the same or a bit more performance. In the end there is still the same TDP limit so the whole thing has to be clocked lower (due to those optimizations for peak cases) and perform at lower levels on average.

 

[Dresdenboy]

Why has JFAMD not chimed in on this "Delay"?

I'll say it again, just tell us if BD is worth the wait for those of us considering going with SB.

Which delay? Never heard of one. But I heard of misinterpreted model listings

 

[Phynaz]

Why has JFAMD not chimed in on this "Delay"?

I'll say it again, just tell us if BD is worth the wait for those of us considering going with SB.

You don't really expect AMD to tell you BD isn't worth waiting for, do you?

 

[Cogman]

Why has JFAMD not chimed in on this "Delay"?

I'll say it again, just tell us if BD is worth the wait for those of us considering going with SB.

Part of the reason might be because he can't.

 

[chihlidog]

Which delay? Never heard of one. But I heard of misinterpreted model listings

Hence why I quoted "delay". None of the info we have is official.

You don't really expect AMD to tell you BD isn't worth waiting for, do you?

Not in so many words. But he's here on his own time, as he makes very clear. I'd just really like to hear something from him that tells me I have a reason to go back to AMD with Bulldozer, as a gamer. If he cant, he cant, but I think its wise for them to give us SOMEthing and hopefully before too long he can.

Let me make clear, I do not mean my posts to be in any way antagonistic. Id really like to go back to AMD and would just love to be thrown a bone here. Im not demanding anything, AMD owes me squat. Im just still hoping.

 

[Tuna-Fish]

That leaving execution starved, with code that offers high ILP, seems like a silly thing to do, compared to matching the capability end to end, to prevent inefficient execution of efficient code.

The different parts of the pipeline have different utilization. The newest Itanium has 12-wide execute and 6-wide decode, is this a silly design? (Well, it is, but for other reasons )

Data stalls stop execution, but don't stop decoding. They also tend to be at the start of code segments (right after failed predictions). This means that in a perfectly balanced design going for maximum utilization of resources, the decode is narrower than the execute.

The widths of parts of the core are results of lots of tradeoffs. Making things wider costs transistors and clock speed -- if they wanted to maximize peak throughput, they could have made the whole pipeline 8 wide. That obviously makes no sense. The engineers designing the chip decided that a 4 wide shared frontend makes more sense than a wider one. They might have been wrong, but we certainly have no way to know.

 

[piesquared]

[edit]

Nevermind-silly

 

[skipsneeky2]

AMD has claimed very little. Everything else are from other people - leaks, assumptions, hopes & dreams, musings, and what have you.

As far as AMD is concerned: benchmarks only at launch.

The good thing most assumptions have been correct the gtx580 is a good example.

If the leaked benchmarks are to be true,time to upgrade my i7 950

If not ill hold out on for ivy bridge :thumbsup:

 

[Tuna-Fish]

The good thing most assumptions have been correct the gtx580 is a good example.

If the leaked benchmarks are to be true,time to upgrade my i7 950

If not ill hold out on for ivy bridge :thumbsup:

Honestly, unless you 3d model or encode video, there probably won't be enough difference between BD and i7 950 to justify upgrading. The early birds who bought i7 920 or likes when they came out really got worth for their money this time around.

 

[skipsneeky2]

Honestly, unless you 3d model or encode video, there probably won't be enough difference between BD and i7 950 to justify upgrading. The early birds who bought i7 920 or likes when they came out really got worth for their money this time around.

True i was one of them i bought a core i7 940 and paired it with a gtx295 .

Had the system for over a year with absolutely no problems then i sold it cause i lost interesting in gaming.

Just now am i getting back into the full swing of things again .

When core 2 duo came out i was the only guy still using a pentium 4 while playing bf2 and i was teased about it so i try and not get behind with technology.

 

[Tuna-Fish]

When core 2 duo came out i was the only guy still using a pentium 4 while playing bf2 and i was teased about it so i try and not get behind with technology.

That's because P4 was utter shit when core 2 duo was out, and had been for 3 years or so.

But i7 950 won't be that bad, even when BD is released. There simply won't be the same kind of performance differential that you had. Unless you are upgrading for the sake of the upgrade, if I was you, I'd look at what good SSD disks are out, and save for a 28nm GPU when they come out either very late this year, or early next year.

 

[skipsneeky2]

That's because P4 was utter shit when core 2 duo was out, and had been for 3 years or so.

But i7 950 won't be that bad, even when BD is released. There simply won't be the same kind of performance differential that you had. Unless you are upgrading for the sake of the upgrade, if I was you, I'd look at what good SSD disks are out, and save for a 28nm GPU when they come out either very late this year, or early next year.

While i welcome the advancement in technology i hope this 28nm part comes out after bf3

Picked up my gtx580 in anticipation for bf3

Had a ssd in a previous build,they are nice but price per gb is still kinda up there,but my next build will have one.

If i do upgrade,the 950 will be a full 24/7 dedicated folding rig for team 198 while the bd platform will be strickly for gaming.

 

[JFAMD]

Why has JFAMD not chimed in on this "Delay"?

Because it is company policy not to comment on rumors. I can only talk about facts.