Is AMD a sinking ship?

[Molester]

this notion that, if AMD could produce 3.0ghz in the necessary thermal envelope they would, is wrong

it's called scaling

 

[hUMANbEATbOX]

Intel isn't selling higher speed Core 2 Duos at the moment because they have no reason to. There is no competition from AMD on that front, why release faster processors and push their profit margins down. Before Penryn they probably had power issues with high speed quads. Right now though they are sitting on the ability to release higher clocked chips, and they will whenever AMD gets near their clock speeds and performance.

I think the E8500 could easily be launching at 3.8ghz instead of the 3.16ghz it is slated at. Good for us overclockers, I would say bad for average users, but most average users wouldn't make good use of a 3.16ghz Wolfdale anyways.

 

[FlahsMemory]

Intel isn't selling higher speed Core 2 Duos at the moment because they have no reason to. There is no competition from AMD on that front, why release faster processors and push their profit margins down. Before Penryn they probably had power issues with high speed quads. Right now though they are sitting on the ability to release higher clocked chips, and they will whenever AMD gets near their clock speeds and performance.

Well I dont consider this profit maximizing on highend. If they are able to offer higher speeds then paying 1500 bucks for a 3.6ghz part is hell of alot better than shelling it out for 3ghz (or 3.2ghz processor.) We are talking about a 18% increase in performance from a part that is 1/3rd the price. The prices between a chip that is worth 400 at 2.6ghz and chip that is worth close to $1500 at 3ghz at initial product launch is less appealing to highend buyers that do not overclock considering the marginal gains. It isnt maximizing highend profits in this case then.

Intel really doesnt have problems making money and maximizing profits because they can offer massive variation in CPU products like 1600mhz FSB and half cache removed or added etc. Not to mention they already control 80% + of the industry. I know that they are banking on AMD to maximize profit in someway but they also realize that if they push AMD far out the market they are going to run into issues themselves. As the products they make wont sell very well because lack of performance increases and no competitive pricing. If this happens intel is going to charge more for their product which hurts the consumer and they will also be forced to even fire more of their staff because lack of need for them.

If anything the only time intel really has a good opportunity to push AMD out the market is if they offer various product with various needs. Sort of what AMD is trying with fusion. Offering these various products and opening up a few factories and R &D to cater to those products will cause a boom in growth as they have many directions to focus on and gain profits from. Right now intel main profit is the CPU and until they get more various sources of income they cant push AMD out just yet or else like I said they will only be competing with themselves and suffer for it.

 

[knightwolf6543]

Unlikely, Not only does Via have an x86 License (Purchased from National Semiconductor who had purchased Cyrix) but IBM would swoop up AMD or at least it's IP if it was to go under. Remember that IBM is trying to become a CPU company these days and the CELL hasn't exactly taken the world by storm as quickly as they'd like .

your talking that 9 core cpu in the PS3 correct?
if IBM does end up buying up AMD all i ask is Samuel would please punch hector in the face and call him the biggest idiot in the world.
am i correct in saying that IBM is bigger then intel? i know their stock is alot more.

 

[vanibanez]

All in all AMD must be doing something right if their supply can't keep up with demand. However, a few years back the philosophy was that clock for clock an AMD could whoop an Intel hands down. Would that still be true? Could a 2.4Ghz AMD still keep up with a 3Ghz Intel? Doing calculations/second, etc. I'm not quite up on all the technical voodoo, sorry.

This may be BS but I have heard past rumors of AMD creating a new onboard video chip that could keep up with PCI-E cards. I guess that would reduce power consumption, card purchasing, etc. A reason why they bought ATI possibly?

Just my .0001 cents worth.

 

[Enablingwolf]

This may be BS but I have heard past rumors of AMD creating a new onboard video chip that could keep up with PCI-E cards. I guess that would reduce power consumption, card purchasing, etc. A reason why they bought ATI possibly?

Regardless of where anything is located in a computer system. If it sucks power, it is a burden on the PSU.

While it would be nice to see a socketed/fixed powerhouse GPU, That is a ways off. The ATX design, has really no room left for anything besides add-on expansion cards.

The BTX design, while novel. Did not take off. I want to see a new standard pop out. We are ready for it. Since the xxTX is getting VERY long in the tooth. It would allow for non-80's tech to flourish.

 

[aaronjb]

Here's the truth, because I said so :

AMD will not exist in its current form by YE 2009. As a publicly traded company, their shareholders won't sit through many more dismal quarters without calling for a significant restructuring.

ATI is still very relevant, as the majority of their income is derived through OEM deals (which are the bread-and-butter of a chipset company), and they still have hardware in plenty of non-PC consumer devices.

 

[AlabamaCajun]

Onboard video would reduce some power by reducing the interface chips same as todays onboards. This is in part why AMD had s lower power factor for so long as intel needs a lot of power to pump up the FSB to the NB. Todays Kents and Yorks use less power for the CPU internals making up the difference. Neha will greatly shift that paradigm for non-overclocked systems.

All AMDs spec for K10 are based on 90 and 65nm Intel hardward before mid 2006. The new 45nm chips are going to leap ahead within a month as far as high end systems. I see AMD still doing great in the Laptop, home pc, server and chipset business. As said, the gfx cards will hold their own.

 

[th3]

am i correct in saying that IBM is bigger then intel? i know their stock is alot more.

I think your right about that, but when looking at semiconductors only Intel is way bigger. If semiconductors was a religion D1D is where pilgrims would go, Hillsboro OR would be what Jerusalem was to the crusaders (and Intel even have a fab in Jerusalem, lol). IBM is using the same process tech as AMD and had similar problems with 65nm, Cell was supposed to be faster but the yields were not as expected, and only recently was a PS3 with lower power consumption launched, so at least for the scaling IBM isnt any better than AMD and couldnt do anything to bring AMD back into the high end game anytime soon.

What i think is that no matter what happens to AMD as a company, Intel will continue to shift towards higher prices and lower performance. In the interview INQ did with the chief of Penryn arch it was confirmed that Intel is holding back on clockspeeds to allow for a mid-life clockspeed bump, at the same time the extreme edition is significantly bumped up in price. My best guess is that it will get worse with Nehalem, i dont expect any cheap nehalems for the average enthusiast, only high end high price parts with IMC and then simple old fashion northbridges for the rest of us. IMO, the only way Intel will push IMC for the higher midrange is if AMD becomes more competitive.

 

[funnyperson1]

this notion that, if AMD could produce 3.0ghz in the necessary thermal envelope they would, is wrong

it's called scaling

Well obviously they would need to be able to produce these with decent yields, but I'm not quite sure what you mean by scaling in the current context.

 

[AlabamaCajun]

~
What i think is that no matter what happens to AMD as a company, Intel will continue to shift towards higher prices and lower performance. In the interview INQ did with the chief of Penryn arch it was confirmed that Intel is holding back on clockspeeds to allow for a mid-life clockspeed bump, at the same time the extreme edition is significantly bumped up in price. My best guess is that it will get worse with Nehalem, i dont expect any cheap nehalems for the average enthusiast, only high end high price parts with IMC and then simple old fashion northbridges for the rest of us. IMO, the only way Intel will push IMC for the higher midrange is if AMD becomes more competitive.

Back in the day! Sky-high priced Prescotts and Itaniums (which now sell for as low as $50 bucks.

 

[Molester]

Well obviously they would need to be able to produce these with decent yields, but I'm not quite sure what you mean by scaling in the current context.

scaling, as in, in 3-4 months, they'll "release" higher clocked cpu's, which of course they can do now, but need to meet demand for lower clocks, the word 'binned' should be familiar to you

 

[Gautam]

x86 will have a short life

You say that as if it would be a bad thing.