AMD Surprises With Revenue Warning

Page 2 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Status
Not open for further replies.
[citation][nom]alidan[/nom]think of 5 years out. intel tried, they did, to have a stand alone gpu years ago and failed. amd im betting saw the cpu+gpu working in tandem being a solution for the future.the cpu side of computing is more or less as good as it needs to be, even if a cpu doubled or trippled its power, it would do what? cut a 90 second operation that needs to be once into a 30 second operation? what programs require a top end intel cpu otherwise they feel sluggish? the hdd to sdd is one thing that makes a computer feel fastergoing from a phenom II x4 to an i7, while would be faster, wouldn't feel faster by much. however, look at cpu alone operations compared to gpu assisted operations. what a amd does in 10 minutes, intel does in 5, but any gpu does in 1.point being that the future is gpu+cpu in tandem, and intel wont be ready when it happens, amd will. sorry if my point doesnt come across well[/citation]

One of my friends tried to build a game that uses primarily CPU for folks that are stuck with Intel HD 2000s or 3000s.

A 3D game, with several eyecandies that are typically reserved for a GPU.

Oh god the lag, may it never see the market. I hope.
 
[citation][nom]A Bad Day[/nom]One of my friends tried to build a game that uses primarily CPU for folks that are stuck with Intel HD 2000s or 3000s.A 3D game, with several eyecandies that are typically reserved for a GPU.Oh god the lag, may it never see the market. I hope.[/citation]

While I'm sure it was an interesting project, I should hope he anticipated that result. As a programmer, I don't see how there could be an intersection between the level of programming ability required to undertake such a project and the lack of knowledge required to expect it to work (no offense); he must be an optimist :).
 
Well, considering their expenditures are in the millions, and their profits are in the "billions", I don't think AMD is going anywhere - anytime soon.

That being said, the current CEO (or was it COO?) of AMD needs a good kick in the a$$ out the front door! I don't like AMD / ATi, I haven't since the end of the K62-3 era, but they are literally killing their own company from the inside out. They need to stop this stupid licensed ARM CPU development BS, and get back on the ball with decent and innovative CPU designs, something bulldozer and pile driver were not. Maybe, If they totally re-did their video card drivers from scratch, and stopped using .NET to build the CCC with, I might actually buy an ATi product for once. I never had the kind of bugs and issues with their current drivers back in the days of the Rage Pro, and Rage 128. I don't know who at ATi thought it was a great idea to use .NET to build the CCC with, but.... I mean.... COME ON! Really???
 
[citation][nom]Maximus_Delta[/nom]After releasing stellar products, especially the 7970, the demise of AMD is hard to fathom.[/citation]
Although the 7970 is a good GPU, the supply problems they're having illustrate their struggles to releasing products on the market. NVIDIA is currently experiencing the same issues AMD was initially, and more complicated architechtures like Kepler result in much more difficult fabbing.

Outsourcing their fabrication is most likely what makes it very difficult for AMD to compete against Intel - they rely on another company to develop/purchase the fab tech, work on the chips, send them back to AMD to report problems, troubleshoot, fab, and then package and ship. Intel still has to deal with these issues, but they are mitigated by the fact that A) their fabs are ~5 years ahead of the rest of the market, and B) all of this is in-house, where they have much better communication and fewer roadblocks to getting a product to market.
 
[citation][nom]HotRoderx[/nom]To me this is scary news world would be a very bleak place if only Desktop/Laptop/Server Processor manufacture was Intel. The day AMD goes under is the day we all pay the price with our wallets in a very bad way. I am starting to wonder if maybe AMD should look at selling off ATI and just going back to strictly CPU manufacturing. Correct me if I am wrong but seems they where doing really really good (top of the market) before they purchased ATI.[/citation]

After purchasing Ati, they progressively gave Nvidia better and better competition. If you really think that the Ati deal hasn't paid for itself several times over, think again. Besides, the Ati deal is almost definitely unrelated to their inferior CPU work these days.
 
When I read the title, I thought their revenues was HIGHER than expected! That's sad, although I agree with others that their APUs will keep them afloat for quite a while. The future is there.
 
[citation][nom]boiler1990[/nom]their fabs are ~5 years ahead of the rest of the market[/citation]

Not sure I agree. Everybody has been talking about tri-gate transistors for a decade now and they're only really one node ahead. In addition, Intel manufacturers on CMOS, AMD on SOI, and Intel will eventually move to SOI, however they delayed the move for their 22nm node due to cost. If I'm correct, SOI also has the potential to use less power than CMOS at the same node.

GF are a couple of years behind, tops.
 
[citation][nom]silverblue[/nom]Not sure I agree. Everybody has been talking about tri-gate transistors for a decade now and they're only really one node ahead. In addition, Intel manufacturers on CMOS, AMD on SOI, and Intel will eventually move to SOI, however they delayed the move for their 22nm node due to cost. If I'm correct, SOI also has the potential to use less power than CMOS at the same node.GF are a couple of years behind, tops.[/citation]

GF is far behind. TSMC is the one that might not be that far behind.
 
Perhaps. I was expecting GF to be manufacturing at full speed on 22nm by 2014, making them a full node behind Intel, however they don't usually inspire much confidence. Still, what challenges must Intel overcome if they're to switch to SOI?
 
"operating expenses improved 8%" because they fired people. Let's see how long that trick keeps working.
 
[citation][nom]silverblue[/nom]Perhaps. I was expecting GF to be manufacturing at full speed on 22nm by 2014, making them a full node behind Intel, however they don't usually inspire much confidence. Still, what challenges must Intel overcome if they're to switch to SOI?[/citation]

I would think that they would either need a new set of factories or at least to do some major changes to one or more factories in order to change from bulk CMOS to SOI. It probably wouldn't be cheap to do, so they might do it on a die shrink in order to capitalize on a more necessary change like they did with the tri gate transistors on the 22nm process instead of a go-between on 32nm.

Basically, killing two birds with one stone in this case. R&D would still cost some more, but implementation should be cheaper than doing both a die shrink and a transition to SOI at different times. I'm also curious about how well the transition will go considering Intel's relative lack of work done with SOI compared to AMD and several other companies. The problem with even getting 22nm at GF by 2014 is that it probably wouldn't be 22nm Tri-gate, so they'd still be behind by an additional process technology instead of just the nm node.
 
It may not be as bad as it looks. If the next iteration of FX series processors exceeds expectations they might be able to get some traction back. Also their APU line of processors is really good and OEM's need to start getting on board more by offering a larger selection of APU based computers - especially tablets. Too many x86 tablets are based on Atom processors.
 
[citation][nom]kitekrazy1963[/nom]The economies are pretty bad in Europe and it has declined in China. I think other tech industries are going to take a hit off this.[/citation]

Yeah, Greece, Italy, and Spain.
 
[citation][nom]drbendanillo[/nom]IF AMD is able to create a CPU that uses the GPU's shaders to calculate an x86 coded FP instruction (Not OpenCL) then it would be a high performance CPU.They should build a shader that is fast enough to compute FP instruction like on the CPU's FPU. Piledriver is because they did not prioritize the Floating Point Units. OpenCL is great but.....it only applies to parallel compute algorithms, serial compute is not as fast as it is runned on generic way.[/citation]


Where did you learn about computer hardware????
 
[citation][nom]blazorthon[/nom]After purchasing Ati, they progressively gave Nvidia better and better competition. If you really think that the Ati deal hasn't paid for itself several times over, think again. Besides, the Ati deal is almost definitely unrelated to their inferior CPU work these days.[/citation]

Hmm, OK - AMD paid $5.4 billion for ATI, so they've made $16.2 billion in profit since 2005?

IIRC AMD lost around $8 billion total from 2007 through 2010, plus they now have to rely on GF and TSMC, switching from one to the other when the first one has fab or yield problems..
 
[citation][nom]fazers_on_stun[/nom]Hmm, OK - AMD paid $5.4 billion for ATI, so they've made $16.2 billion in profit since 2005? IIRC AMD lost around $8 billion total from 2007 through 2010, plus they now have to rely on GF and TSMC, switching from one to the other when the first one has fab or yield problems..[/citation]

AMD's CPU side has been losing money, not their GPU side. The GPU side is what has helped keep them afloat since buying it. As their CPU market share dwindled, their GPU market was improving. That they've now lost money in the past few years is purely on their CPU side and they would have been in a worse position if not for their GPU side being profitable.

Furthermore, what fabs that AMD uses for their CPUs didn't get worse because of the Ati deal... You also ignore the fact that a lot of money lost went to cutting strings and such with companies such as GF in order to not be tied down to a seemingly poor fab company. This will be recouped. The fact that AMD's currently most successful processors and most steadily climbing upwards processors in profitability are their APUs tells us just how important the Ati deal was for AMD's success. Where would AMD be without their GPU work to save the company? Undoubtedly still in business, but in a much worse position.
 
WHY THE FKC YOU NEED TO ISSUE A WARNING !!!!!!!!!!!.......

The world economy is struggling, every one knows that,AMD lost $10 a share for nothing.Stupid management stating from CEO Rory Read.

How many times a company can shoot itself in the foot ?
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom