• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

AMD Cancels 10-Core, 20-Core Server Processors

I don't think it is all that bad as it might indicate that AMD is focusing on improving per core performance rather than cramming more cores as they could. They could cram in 10-20 cores but with the same 32nm, they're probably going to have to reduce the stock clockspeed even lower than now which is just going to make matters worse if BD/PD arch relies heavily on its clockspeed advantage.
 
Basically shelving everything as they work on R&D. Hopefully its to avoid over spending on slower parts, and just to invest in a new architecture (though I don't think the BD architecture itself is bad, just needs an overhaul).

I am hoping that they haven't completely abandoned the performance end, just put it on the back burner.
 
dma0991 said:
I don't think it is all that bad as it might indicate that AMD is focusing on improving per core performance rather than cramming more cores as they could. They could cram in 10-20 cores but with the same 32nm, they're probably going to have to reduce the stock clockspeed even lower than now which is just going to make matters worse if BD/PD arch relies heavily on its clockspeed advantage.
Click to expand...
I do hope they are doing so. Because it seems at even at some good scaling programs BD has a lower IPC pre squaremeter compared to Sandy Bridge, due to the wrong design of cache level and the high latency within the dispatch unit and the execution unit.
 
Topweasel said:
I am hoping that they haven't completely abandoned the performance end, just put it on the back burner.
Click to expand...

They have a brand-new CEO who is from the outside as far as CPU design is concerned, its not "in his blood" to blindlessly chase after Intel.

Take away the drive and you are left with analyzing the numbers. What CEO would look at Intel and say "you might be able to outspend me 6:1, but I'm going to take my 1 and piss it away trying to best your 6"?

That's a pretty tough sell in the boardroom, an even tougher sell in front of your analysts.

So what would any CEO do? Stop throwing good money after bad, find a way to avoid competing toe-to-toe with Intel.
 
Idontcare said:
They have a brand-new CEO who is from the outside as far as CPU design is concerned, its not "in his blood" to blindlessly chase after Intel.

Take away the drive and you are left with analyzing the numbers. What CEO would look at Intel and say "you might be able to outspend me 6:1, but I'm going to take my 1 and piss it away trying to best your 6"?

That's a pretty tough sell in the boardroom, an even tougher sell in front of your analysts.

So what would any CEO do? Stop throwing good money after bad, find a way to avoid competing toe-to-toe with Intel.
Click to expand...


I agree, and this is also consistent with what I think AMD is attempting to do in the next few years....move itself out of the Intel/AMD camp and instead become part of multiple camps including AMD/Intel, ARM, Embedded etc. I think it is probably the only way AMD could maintain long term viability. As you say, throwing yourself against Intel forever, when outspent by multiple factors, is a long term losing strategy.
 
CHADBOGA said:
AMD are beginning the long slow march to becoming the next VIA.
Click to expand...


They were headed this way the way they were already going....they were not going to get back ahead of Intel. They know their own internal road map over the next five years and likely know a lot about Intel's road map, and likely made some hard choices to change the direction of the company.
 
Well at least they hit a point of deminishing returns, I think its great theyre less willing to go with their more cores method. Hopefully amd can figure out a way to invest in people more capeable of making a balanced cpu.
 
Idontcare said:
They have a brand-new CEO who is from the outside as far as CPU design is concerned, its not "in his blood" to blindlessly chase after Intel.

Take away the drive and you are left with analyzing the numbers. What CEO would look at Intel and say "you might be able to outspend me 6:1, but I'm going to take my 1 and piss it away trying to best your 6"?

That's a pretty tough sell in the boardroom, an even tougher sell in front of your analysts.

So what would any CEO do? Stop throwing good money after bad, find a way to avoid competing toe-to-toe with Intel.
Click to expand...


But is it the right move to start battling a miniscule margin end of mainstream and below?

The money is in Servers, period.
Give up servers and then what?

Intel might not succeed so well with first low end/mobile iterations, but who cares.

The losses from those divisions are miniscule to the profits from server end.



Won't this pretty much make Intel have freepasses for the next generation(and probably the coming generations) in the highest margin and largest revengue/profit market at all?

Pretty much giving them all the time in the world to tackle ARM ultraportable/mobile.
 
busydude said:
I think the days(or hope) of AMD competing with Intel on the high end.. are over.
Click to expand...

of course not, intel will slumber on its current lead, maybe even hibernate or fall asleep to the point where the compition 1up's them again ... it is history you know ..

It will take a few years of course, so get your new rig now, it's gonna last.
 
RavenSEAL said:
Which means get ready for $1200 quad cores :thumbsdown:
Click to expand...

AMD hasn't competed directly with Intel in performance since Conroe released.

Despite the feeling from many AMD users that Intel is the Lord Satan, they've continued to innovate, produce more performance per MHz, more MHz, and more performance per watt all the while keeping the exact same pricing structure they've had since before Conroe.

The sad thing about all of this is the simple fact that AMD is the problem. Their awful showing for the last seven years has done nothing to innovate the market and their consistent overpricing of new products has done little to nothing to affect Intels yearly price price structure.
 
Intel competes against its older products too, they cant "hibernate". Yes, they might slowdown a bit though. Ivy Bridge with its ~162mm² die and lower TDPs would probably allow Intel to release a hexa-core version (for LGA1155 users) if they really needed.
 
Intel has to compete with themselves now unlike the P4 days. A nearly 6 year old Core 2 Duo is still enough for most users. Intel needs to create something compelling to have someone upgrade. We're the exception in that we want the latest and greatest on a frequent basis.
 
MisterMac said:
But is it the right move to start battling a miniscule margin end of mainstream and below?

The money is in Servers, period.
Give up servers and then what?

Intel might not succeed so well with first low end/mobile iterations, but who cares.

The losses from those divisions are miniscule to the profits from server end.



Won't this pretty much make Intel have freepasses for the next generation(and probably the coming generations) in the highest margin and largest revengue/profit market at all?

Pretty much giving them all the time in the world to tackle ARM ultraportable/mobile.
Click to expand...

The money, as you say, is in all the various market areas. Small margins just mean you have to sell more volume, which is a legitimate business strategy. Besides, it really did not make much difference whether they wanted to "give up" or not. My feeling is that they know that Intel's R&D and fab lead was only going to widen the gap over time, so why keep fighting for a market you are fairly confident you cannot maintain?

Besides, we don't know that Intel will continue to have dominance in the server markets of the future. It appears to be the case today, but disruptive technologies are right around the corner and you never know what may happen.

And further, AMD may be able to move into custom IP arenas that Intel cannot do (yet). Say google wants to order 100,000 servers and decided it wants custom cores with a few Piledriver cores, a few ARM cores, and a few custom cores for security and encryption. If Google has to buy servers with all of that right now, it can get pretty pricey, especially if not all of those things are necessary all of the time. Instead, AMD could offer all of that on one CPU package and sell it at a nice markup..making them money and saving Google money at the same time. This is just an example, and that is just one company, so you never know. Have to remember, only Intel and AMD have potent x86 cores ATM and that could be a huge selling point for AMD in the custom IP market!
 
Zor Prime said:
Hold on and wait for someone to mention that Intel competes with itself which keeps prices down.
Click to expand...

Not sure if you'd be willing to listen to him since you seem rather cynical from the outset, but here is what Intel's own Mark Bohr had to say:
Mark Bohr said:
Not only does Intel want to remain ahead of its MPU competition, its computer customers need faster MPUs every two years so they can sell new systems to their customers.
Click to expand...

(He is obviously referring to Intel's customers as the OEM's who buy their chips and then resell them to the end-user, we are not Intel customers, we are Newegg customers, DELL customers, etc...but he pretty much says it there in black and white)
 
Smartazz said:
Intel has to compete with themselves now unlike the P4 days. A nearly 6 year old Core 2 Duo is still enough for most users. Intel needs to create something compelling to have someone upgrade. We're the exception in that we want the latest and greatest on a frequent basis.
Click to expand...


There is also the issue that Intel does not want to get caught with its pants down so to speak. If they keep innovating and firing on all cylinders, while executing a sound business strategy, they should be able to make a lot of money whether there are competitors or not. If they keep up a solid roadmap of improvement, it also keeps others from getting curious and doing things like looking into designing and using ARM cores for the server environment. If Intel had a processor that could fill that market, it makes it a lot less likely someone else is going to come in and cause headaches in the future.

Further, many of Intels products are coming out of design cycles that have been in the works for years, so slowing it down at this point does not help reduce R&D costs etc...since those massive teams need to be doing something. Longer term, they could cut R&D budgets, but I would guess for the time being they have a pretty fixed R&D cost and pipeline and it would not make much sense to slow down now.
 
Idontcare said:
Not sure if you'd be willing to listen to him since you seem rather cynical from the outset, but here is what Intel's own Mark Bohr had to say:


(He is obviously referring to Intel's customers as the OEM's who buy their chips and then resell them to the end-user, we are not Intel customers, we are Newegg customers, DELL customers, etc...but he pretty much says it there in black and white)
Click to expand...

Of course.

But Intel doesn't do it for the purpose of keeping prices down which is some insane garbage I read on here from time to time.
 
CHADBOGA said:
Why is Ivy Bridge going to slot into Sandy Bridge's pricing, after the dud that is Bulldozer got exposed?
Click to expand...

So people actually upgrade? Believe it or not, you do compete with yourself when you bring out a new product. You have to give people enough value so that they want to upgrade from what they already have. Sure prices are probably going to go up without competition, but a 1200 dollar quad core will not fly in the market.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top