XBitlabs: Advanced Micro Devices Set to Unveil New Strategy Next Week

[RussianSensation]

doesn't make sense. They can't do that with x86, and if they try to do it with ARM, yeah right, they'll be a year late and $1m short of 'whatever their target was'-- there are already enough ARM competitors with powerful chips coming out in 2012 let alone 2014.

I agree with you. It's no wonder none of the previous executives was OK with this new strategy the Board wanted them to pursue. Also, Read was not the Board's #1 pick for the CEO spot. At least 3-4 leading executives at other firms turned down the offer. It's very possible they simply thought the Board was out to lunch with what Read is about to announce next week. I don't think AMD has enough $ or resources to go head-to-head against Samsung, Qualcomm, TI. NV now has a #5 position in smartphone market share, but of course NV is a far more profitable business (i.e., they can afford to invest a lot more $ into Tegra).

Just less than a week to find out what Read has cooked up.

Samsung had cash and equivalents of 21.75 trillion won ($19.4 billion) in the third quarter. :hmm:

 

[cbn]

Possible translation: Why worry about competing with NV/Intel at high-end (which costs so much $$ to design a good GPU/CPU and entails a lot of risk), when we can just regain whatever lost market share we have by selling more APUs in emerging markets and markets that require low-powered devices?

I.e., Is he hinting that the R&D budget for high-end discrete graphics/desktop CPUs is going to be diverted towards mobile/tablet/smartphone chips?

.

I understand that is about the Consumer APUs.

But what about the APUs and HPC?

Have there been any breakthroughs in OPEN CL since the last time I frequented these forums (Feb. 2011)?

I saw a Penguin Computing Press Release the other day, but what is happening with OPEN CL and the discrete Cards?

 

[piesquared]

Possible translation: Why worry about competing with NV/Intel at high-end (which costs so much $$ to design a good GPU/CPU and entails a lot of risk), when we can just regain whatever lost market share we have by selling more APUs in emerging markets and markets that require low-powered devices?

I.e., Is he hinting that the R&D budget for high-end discrete graphics/desktop CPUs is going to be diverted towards mobile/tablet/smartphone chips?

.

Why not. The ROI for high end desktop CPU's is pretty damn low. Even though it's the same die as a server chip, there's a lot of resources tied up in trying to sell them to a small group of people. Especially when the cards are stacked so heavily in favor of intel, the main one being the cash on hand. It's utter stupidity to pursue it any further when they have a volume chip knocking on the door that will likely gain desktop share. And besides as we've seen across so many forums, nobody wants AMD's chips for the high end anyway.

As for the GPU, the workstation market where they gained share, should drive further GPU development.

 

[BD231]

Sh t ...., lets hope Intel doesn't go back to it's tearanical ways.

 

[soccerballtux]

I agree with you. It's no wonder none of the previous executives was OK with this new strategy the Board wanted them to pursue. Also, Read was not the Board's #1 pick for the CEO spot. At least 3-4 leading executives at other firms turned down the offer. It's very possible they simply thought the Board was out to lunch with what Read is about to announce next week. I don't think AMD has enough $ or resources to go head-to-head against Samsung, Qualcomm, TI. NV now has a #5 position in smartphone market share, but of course NV is a far more profitable business (i.e., they can afford to invest a lot more $ into Tegra).

Just less than a week to find out what Read has cooked up.

Samsung had cash and equivalents of 21.75 trillion won ($19.4 billion) in the third quarter. :hmm:

if they had thought of this about 3 years ago maybe, they would have had Global Foundry's 45nm and 32nm nodes to play with. Could have worked. But if they're going after it now sounds like the board is greedy and is just seeing a golden pot at the end of a rainbow.

 

[RussianSensation]

The ROI for high end desktop CPU's is pretty damn low.

Ok but keep in mind that:

Currently it's too expensive to develop 2 specific CPU architectures: 1 for mobile and 1 for desktop markets. AMD's Fusion still uses Athlon II X4 desktop CPU architecture, while Sandy Bridge is used for Intel's laptop chips. What you need is a well-rounded architecture that performs well in both markets. Intel was able to do that. The problem is someone at the very top decided to go all out on cores for Bulldozer and neglected all other important aspects of this design decision: overall power consumption, IPC and performance/watt. The design decision of going with MOAR CORES while not putting any resources towards IPC pretty much bets the CPU on heavily multi-threaded software (we aren't there yet), OR on having a very high frequency ceiling that can overcome IPC deficit (but this often requires higher voltages and higher frequencies contribute to worse power consumption).

So the return on investment for desktop CPU architectural design overall is actually not that low. Intel has outstanding gross profit margins.

It's pretty obvious you can make a lot of $ selling desktop and laptop chips based on a robust CPU architecture (see Conroe in C2D, Nehalem/Lynnfiled in i3/i5/i7s, Sandy Bridge, etc.), Problem is AMD wasn't able to do it.

Especially when the cards are stacked so heavily in favor of intel, the main one being the cash on hand. It's utter stupidity to pursue it any further when they have a volume chip knocking on the door that will likely gain desktop share. And besides as we've seen across so many forums, nobody wants AMD's chips for the high end anyway.

Ya, that's probably true. I agree with you that in the long-term, a lot more consumers will be happy with decent performance at low price and low power consumption. But I think your 2nd comment about no one wanting AMD chips for high-end has more to do with AMD's lower performance, not necessarily with the AMD brand. If AMD produced a good desktop CPU, a lot of us would want such a processor for our systems

But ya, generally, I probably agree with you that in the short-term, it would be too risky to pour even more $$ into the next high-end CPU architecture that might come out in 4-5 years. AMD can't afford another Bulldozer II flop in 2016 that might actually send it into bankruptcy. Bobcat, Llano and Brazos have a higher potential to be more successful in the next 5 years if even more resources are allocated to them. Problem is Llano (or Trinity) would still be strapped with poor performing Bulldozer CPU foundation....and in 5 years if they don't invest into a new CPU architecture, then AMD's mid-range APU will become uncompetitive for CPU related tasks.

As for the GPU, the workstation market where they gained share, should drive further GPU development.

You can't just focus on workstation GPU market and abandon discrete GPUs for laptops and desktops. The R&D cost to make good drivers for workstations and complex enough GPUs that can be competitive is simply too high for a workstation GPU business to survive on its own. I have my doubts that NV would be able to sustain its Quadro market if it wasn't able to spread its R&D and other expenditures across discrete GPU (desktop and mobile) and its Tesla markets!

 

[Idontcare]

I agree with you. It's no wonder none of the previous executives was OK with this new strategy the Board wanted them to pursue. Also, Read was not the Board's #1 pick for the CEO spot. At least 3-4 leading executives at other firms turned down the offer. It's very possible they simply thought the Board was out to lunch with what Read is about to announce next week. I don't think AMD has enough $ or resources to go head-to-head against Samsung, Qualcomm, TI. NV now has a #5 position in smartphone market share, but of course NV is a far more profitable business (i.e., they can afford to invest a lot more $ into Tegra).

Just less than a week to find out what Read has cooked up.

Samsung had cash and equivalents of 21.75 trillion won ($19.4 billion) in the third quarter. :hmm:

Considering that Dirk thought they were out to lunch to such an extent that he would rather walk away from millions in compensation rather than lead the charge that the BOD wanted him to enact...and all the subsequent top-picks for the CEO position said "uh, no, not even for the millions in compensation that you are offering", and the then acting CEO Seifert said "no, hell no, find someone else"...Rory is either purely in it for the money or he is as equally drinking from the koolaid as the BOD's.

Either way I think we can all agree the situation at AMD with its BOD is largely unprecedented. I can't think of a comparable situation in which so many walked away from the opportunity to personally enrich themselves by so much simply for the sake of being in violent disagreement with the directives being issued by the BOD.

Usually CEO's are removed by the BOD because of their incompetence in enacting the directives of the BOD, not for their outright refusal to attempt to implement them.

Everything about AMD this past year has been very troubling, very telling, and rather predictable in hindsight. History is painting a consistent story here of failure in management from the Boardroom on down.

With the same boardroom still intact, I am not hopeful that the change AMD needs is the change that we are about to told is coming IMO.

 

[cbn]

Speaking of emerging markets, I wonder how Nvidia is planning on positioning "Grey"

Will this be a baseband integrated version of Tegra using the single core ARM11?

Can AMD field any competition to this? Maybe by buying a very small Chinese SOC design house?<---Or is even something like this too much money at the moment for AMD?

 

[beginner99]

How big is ARM actually? I guess too big to be bought by AMD? What about Intel? Couldn't intel just buy ARM?

I mean it would be easier to just buy IP and make money of it instead creating your own design.

Emerging markets make sense.Company I work for has this strategy as well and actually it's the only reason why it is growing. The western world is more or less equal, low 1 digit growth (which makes sense in that market).

 

[Tsavo]

How big is ARM actually? I guess too big to be bought by AMD? What about Intel? Couldn't intel just buy ARM?

I mean it would be easier to just buy IP and make money of it instead creating your own design.

Emerging markets make sense.Company I work for has this strategy as well and actually it's the only reason why it is growing. The western world is more or less equal, low 1 digit growth (which makes sense in that market).

Intel could buy ARM Holdings with their lunch money, but I doubt anyone would want them to.

 

[drizek]

Intel could buy ARM, AMD and nVidia and run them all into the ground if anyone would let them.

 

[Nemesis 1]

Looks like we will finally find out what it was that the BOD wanted Dirk to do so badly, and what it was that Dirk found to be so absurd in his mind that he was willing to step down and forgo millions in CEO pay rather than attempt to implement and pursue at the BOD's behest.

AMD at this point in time needs to pull an apple . Do the whole product . With their new CEO and his connections its do able

 

[Nemesis 1]

How big is ARM actually? I guess too big to be bought by AMD? What about Intel? Couldn't intel just buy ARM?

I mean it would be easier to just buy IP and make money of it instead creating your own design.

Emerging markets make sense.Company I work for has this strategy as well and actually it's the only reason why it is growing. The western world is more or less equal, low 1 digit growth (which makes sense in that market).

Arm being bought out . China has most to gain

 

[cbn]

Arm being bought out . China has most to gain

Why do you say this?

 

[Nemesis 1]

Because their the biggest growing market in the world. It would be fun to watch the 2 awakened sleeping giants do battle

 

[cbn]

Because their the biggest growing market in the world. It would be fun to watch the 2 awakened sleeping giants do battle

I am asking because the Chinese Government has partial ownership of the Loongson CPU.

http://en.wikipedia.org/wiki/Loongson

 

[Nemesis 1]

Ya the dragon has been a let down for me I figured by now it would have put AMD out of business.

 

[piesquared]

Ya the dragon has been a let down for me I figured by now it would have put AMD out of business.

Just in case anyone doubted your bias? lol

 

[cbn]

AMD has a few options. Predictions[dons wizard hat]

1) Moves BD to Server platform only. No more Mainstream BD CPU's in Retail. Future development of BD will pass some enhancements to the Fusion line, but only those that make sense.

2) Fusion variants(Llano/Zacate/future variants based around BD)will be the only Processors available in the Mainstream/Retail market.

3) AMD will get involved with ARM in some fashion. Perhaps not a complete package, but bolt on parts for others to use.

Speaking of #3, I just wonder if the June 2012 delay of 28nm Bobcat has anything to do with some strategy AMD might be using to put that part on the low leakage "Smartphone/Smartphone based Tablet" process. Currently it is on the higher power 40nm process like the desktop and Laptop chips

If the Bobcat GPU really does get the low power treatment, I wonder if this would make it easier to sell it as a "Bolt-on" part to others?

Or maybe it would make the Bobcat GPU easier to use with some type of ARM core AMD is considering?

 

[sandorski]

Speaking of #3, I just wonder if the June 2012 delay of 28nm Bobcat has anything to do with some strategy AMD might be using to put that part on the low leakage "Smartphone/Smartphone based Tablet" process. Currently it is on the higher power 40nm process like the desktop and Laptop chips

If the Bobcat GPU really does get the low power treatment, I wonder if this would make it easier to sell it as a "Bolt-on" part to others?

Or maybe it would make the Bobcat GPU easier to use with some type of ARM core AMD is considering?

Good question. No idea though, I just pulled all that out of my ***.

 

[cbn]

Considering that Dirk thought they were out to lunch to such an extent that he would rather walk away from millions in compensation rather than lead the charge that the BOD wanted him to enact...and all the subsequent top-picks for the CEO position said "uh, no, not even for the millions in compensation that you are offering", and the then acting CEO Seifert said "no, hell no, find someone else"...Rory is either purely in it for the money or he is as equally drinking from the koolaid as the BOD's.

Either way I think we can all agree the situation at AMD with its BOD is largely unprecedented. I can't think of a comparable situation in which so many walked away from the opportunity to personally enrich themselves by so much simply for the sake of being in violent disagreement with the directives being issued by the BOD.

Usually CEO's are removed by the BOD because of their incompetence in enacting the directives of the BOD, not for their outright refusal to attempt to implement them.

Everything about AMD this past year has been very troubling, very telling, and rather predictable in hindsight. History is painting a consistent story here of failure in management from the Boardroom on down.

With the same boardroom still intact, I am not hopeful that the change AMD needs is the change that we are about to told is coming IMO.

No sure what is going to happen, but AMD is probably going to get crushed from above with Haswell Laptop SOC and these *supposed* $599 Ultra Books (with cheap display, no doubt).

This is probably one strong reason why AMD BOD wants to push "emerging markets" and "low power", even with x86!

Of course, AMD has a major problem. They are in the casual market (not the business market with Intel). This means they have to compete against all the upcoming casual market ARM machines and the hordes of software programmers marching behind them.

Maybe the one saving grace here is the emergence of keyboard enabled Tablets like the ASUS transformer. That gives x86 in the Tablet format some extra value. (re: Keyboards and x86 legacy apps play nicely together).

With that being said, I am still skeptical how far AMD can go with ultra low power x86 Windows 8 Tablet. How interested are emerging markets for expensive x86 legacy apps? How much will a Windows 8 license add to the price of one of these ultra low power x86 Bobcat Tablets compared to ARM A15 with Android?

 

[Idontcare]

Just in case anyone doubted your bias? lol

That brought a smile to my face

 

[Tuna-Fish]

How big is ARM actually? I guess too big to be bought by AMD? What about Intel? Couldn't intel just buy ARM?

Intel could buy ARM Holdings with their lunch money, but I doubt anyone would want them to.

It's actually not really possible for a company to purchase ARM.

It's not really that expensive (It is still worth more than 3 times as much as AMD presently ), but it is largely owned by it's customers that have bought it as a strategic investment to ensure supply. Even if you show up with *much* more cash than it's market value, I find it doubtful that Samsung, Apple, nVidia, Nokia, TI, and many others would be willing to sell their stakes unless they felt that their supply would be guaranteed. Basically, if you wanted to buy out ARM you'd have to buy out majority of the tech sector that depends on it. No-one has that much money.

 

[eternalone]

No-one has that much money.

You sure?

 

[cbn]

It's going to get awfully competitive in smartphones/tablets soon. If AMD licenses ARM, how are they going to differentiate their chip from countless other competitors?

Why not "Bin" Stock design ARM Cores?

Right now, AMD financial guys might look at the ARM market and say, "I don't like the fact we have to pay a royalty per chip resulting in a thin margin". <----Couldn't "Binned" SOC change that?

As big caveat might be this would only work for Tablets. As I see things the contract smartphones are controlled by only a few device makers scattered over four major networks. I'd imagine a company would have to be pretty large in order to have enough "high end bin" left over to fill the spec for a special contract phone.

My Conclusion (FWIW): AMD should skip the Contract phone market (where there are too many competitors anyway) and instead focus any ARM based efforts on the non contract devices (Tablets and entry level Smart phones).