New AMD CEO readies strategy shift

[IlllI]

my guess is the strategy shift will be to spin off its hardware division under the moniker of '40 billion startup company' and become solely a software only company.

 

[lol123]

my guess is the strategy shift will be to spin off its hardware division under the moniker of '40 billion startup company' and become solely a software only company.

Lmao.

 

[Arkadrel]

Twice as much means half as long.

Lets say "Intel/AMD" with 86x have 12hours web surfing.
Now they used "twice" as much, so ARM has 24hours web surfing.

However the Intel/AMD can do more demanding things than the ARM tablet, at faster speeds.

Which do you think people will pick?
At some point... you have long enough use, you dont need anymore.

Lets just say that the daily rutine becomes to recharge it once daily.

With arm, you ll never run into a "awww im outta power" situation, but you ll have to live with haveing bad performance in certain things vs 86x that dont.



Do you really think if such a situation comes up, that power consumption will matter in the end to the buyer's/user's?

Once you reach the critical "long enough battery life" threshhold, the payoff in more effecient power designs drops drastically (is my assumption), and Im assumeing consumers would rather have speed.

 

[Khato]

The current x86 offerings are looked over because they use orders of magnitude more power than ARM, both at load and in standby, not because they're just marginally worse, which they obviously are not.

I'd tend to disagree that power is the primary reason for the low power Intel Atom offerings not being adopted. Why? The lowest power Atom is the single core Z500 running at 800 MHz with a max TDP of 0.65 W. Meanwhile according to ARM, their cortex-A9 power optimized hard macro (note that this doesn't include memory interface, most IO, or graphics) has a total power at the target frequency of 0.5 W (source is the performance tab of http://www.arm.com/products/processors/cortex-a/cortex-a9.php).

The real issue is the very same as what was supposed to be Intel's 'strength' - the fact that they're on the x86 instruction set. When the majority of the market they're trying to get into is coded and heavily optimized for ARM... yeah. And, of course, Intel didn't really endear itself to the community by trying to bring their own OS into the mix and due to that partnership giving preferential treatment to Nokia. I can't imagine that Google had much interest in optimizing Android for x86 while Intel was promising awesomeness from Meego... Now that Intel seems content to let Meego sit around collecting dust and is no longer strongly tied to a single phone supplier... well, I'd imagine that everyone has heard about Google's intention to support x86 on Android going forward. Once the software playing field is mostly level, then we'll really get to see how x86 compares to ARM.

 

[lol123]

I agree that the major hurdle for Intel in the ultramobile market (and conversely for ARM to gain entry the PC market) is the prevalence of the competing ISA and its software ecosystem. However with Google officially supporting Android on x86 and with Android running most of its applications in a JIT compiling VM, it will be easier for Intel to make inroads than it otherwise would be. Apple could just as easily make a transition to x86 if they wanted to, but for the time being I would consider that unlikely.

 

[Phynaz]

Idle power means nothing, operational power its what it counts, how many hours can you serf the net, watch TV, watch DVDs or even BlueRay movies, play games etc etc.

Intel X86 at 22nm could have close or the same Idle power, but when you start to do things it will consume more than ARM SOCs

We'll have to wait and see, but Intel will have a big competition in the Mobile sector, not what she is used to with AMD in the x86 market.

Idle power means pretty much everything for the typical user. Everything you listed except play games has the CPU sitting idle most of the time.

While I'm typing this my CPU is sitting idle. Watching video should be off-loaded to the GPU.

 

[nonameo]

How about a transmeta approach and have a CPU run more than one instruction set? I know transmeta ended up flopping, but... hey, who knows.

 

[AtenRa]

Idle power means pretty much everything for the typical user. Everything you listed except play games has the CPU sitting idle most of the time.

While I'm typing this my CPU is sitting idle. Watching video should be off-loaded to the GPU.

Well, i was talking about the entire x86 APU not only the CPU cores.

I do agree that most apps will use the GPU and thats the opportunity for AMD, they dont have to create the fastest CPU core but a better APU, with better iGPU than Intel and low power usage.

If they are not far behind in manufacturing process then they can do it

 

[tommo123]

Once you reach the critical "long enough battery life" threshhold, the payoff in more effecient power designs drops drastically (is my assumption), and Im assumeing consumers would rather have speed.

you're both right. cpu grunt is great and all but lets be honest here - a core2duo from 06 is enough for the vast majority of people. hell, an athlon 64 is (with a gpu for video).

"good enough" applies for the cpu grunt and the battery life. there is a chance that ARM will hit the point of providing "good enough" cpu grunt. the question is, can intel provide the "good enough" battery life?

if it can, then i'll have an intel win8 tablet please.

p.,s good enough for me is 18 hours of video playback at a decent brightness.

 

[RussianSensation]

Here is WHY the Board of Directors wants AMD to pursue smartphone and tablet markets:

Sticking to desktop and laptop CPUs meant making $0 off exponential growth in "Other Media Tablets Market":

http://www.bgr.com/2011/08/24/ihs-u...-sees-apples-ipad-share-sinking-through-2015/

And missing doubling of the Smartphone market:

http://www.isuppli.com/Mobile-and-W...-Majority-of-Cellphone-Shipments-by-2015.aspx


AMD can also leverage their prowess in GPUs for the smartphone gaming market that's about to EXPLODE:

Japan's Smartphone Gaming to grow 30-fold by 2015
http://www.penn-olson.com/2011/09/15/japan-smartphone-gaming-2015/

 

[RussianSensation]

Read was hired specifically to get into the mobile (smartphone) space. Dirk was fired primarily because he did not and would not consider getting into smartphones. The only twist here is the board twisting arms to get the strategy they want to see.

Seed Planning, a market research and consulting firm, recently released a report projecting massive growth in Japan’s smartphone gaming market. In short, the report forecasts 30-fold market growth from 2010 to 2015. Specifically, this is an increase from 8.5 billion yen (approx. US$110 million) in 2010 to an expected 255 billion yen (approx. $3.3 billion) in 2015.


That's in Japan alone......

 

[smartpatrol]

The current x86 offerings are looked over because they use orders of magnitude more power than ARM, both at load and in standby, not because they're just marginally worse, which they obviously are not. If you really think that the same rules will apply in few product generations when x86 has cut its power consumption possibly by a factor in the tens then you are seriously deluded. The only thing going for ARM at that point (although that very well might prove decisive) will be its already established position as the dominant ISA in the ultramobile market.

Actually I think the biggest thing ARM will have going for it is price. Intel will have to compete against Qualcomm, Nvidia, Samsung, TI, etc

http://arstechnica.com/business/news/2011/02/nvidia-30-and-the-riscification-of-x86.ars
(Skip to the last section of the first page if you don't care about the Nvidia-specific stuff)

 

[sm625]

They dont use intel in mobile devices because intel is toooo expensive. Way too expensive. No one wants to pay $62 for an atom to have to turn around and install 4 different support chips for a total of $85. Not when you can get it all on one chip for $45. If intel bolted 1GB of 64bit LPDDR2 to the atom, as well as their southbridge, then maybe it would get used. But it still would need a transceiver and gps, though I guess those can be cheaply had all in one package nowadays.

 

[PreferLinux]

Low poer x86 cores are still a lot more power hungry than ARM designs. And in the mobile and ultra mobile market. Battery longevity does matter.

And they are a lot more powerful, too.

 

[Fox5]

What benefit is there to AMD being an ARM licensee?

1. AMD can just be like everyone else, use off the shelf ARM cores and integrate things in. What does AMD really have to add here? Graphics? Nvidia, PowerVR, Qualcomm, and apparently now even ARM itself are very strong competitors in a market where graphics still don't matter very much.

2. AMD can design it's own ARM cores. Can AMD really afford to design a custom ARM core, plus high end server chips (Bulldozer derivatives), and competitive low power x86 mobile chips (Brazos derivatives)? If AMD builds its own custom ARM core, I'd expect them to either drop the Brazos line, or to get out of the server market entirely and focus on consumer electronics chips.

3. AMD will probably never be a real player in the phone market. However, the tablet market is an unexpectedly fortuitous development for them, as their current low power x86 chips aren't far from meeting the requirements, AND AMD happens to have the performance/watt advantage in both cpu and gpu compared to Intel and ARM at this moment. The Z-01 AMD tablet chip comes in at around 5W TDP, and soon they'll have a tablet chip that does closer to 4W, still on 40nm. Dropping to 28nm and customizing the design further for tablets should allow them to come out with an ideal tablet chip, without going ARM at all. The next gen after 28nm should then allow for the smaller (under 9" and thin form factors).

 

[BlueBlazer]

Historically, if anyone can remember Intel sold off its ARM division (XScale processor) to Marvell. In other words, Intel did not pursue this line of technology further. This begs to question, strategically what benefit would AMD have just to be another ARM producer, while Intel is trying to move x86 into ARM's territory? I think is likely AMD hasn't got the R&D capacity to design and produce very low power mobile chips on their own. Thus, adopting ARM designs would be faster than starting from scratch and allows AMD to enter the ever expanding handheld mobile market quickly. :hmm:

 

[Idontcare]

What benefit is there to AMD being an ARM licensee?

What benefit is there to Nvidia being an ARM licensee?

Every dollar that Nvidia makes with Tegra is a dollar that AMD could have just as easily earned, or at least competed to earn and probably have scored 50% market share versus the current 0%.

What benefit is there to AMD being an x86 licensee, or DX9/DX10/DX11?

 

[lol123]

What benefit is there to Nvidia being an ARM licensee?

Every dollar that Nvidia makes with Tegra is a dollar that AMD could have just as easily earned, or at least competed to earn and probably have scored 50% market share versus the current 0%.

What benefit is there to AMD being an x86 licensee, or DX9/DX10/DX11?

Nvidia tried as hard as they could during several years of legal battles to acquire an x86 license from Intel. When they couldn't, they went with ARM instead. AMD already has an x86 license. So what's your point really?

 

[Idontcare]

Nvidia tried as hard as they could during several years of legal battles to acquire an x86 license from Intel. When they couldn't, they went with ARM instead. AMD already has an x86 license. So what's your point really?

I see you haven't, or couldn't, answer a single one of the questions posed in my post that you quoted.

So what's your point really?

 

[lol123]

That's because your questions were apparently just random counters to another question that actually had some relevance and substance.

The reason Nvidia has an ARM license is that they want to compete in the CPU space as their existing GPU business is rapidly shrinking and they couldn't get the x86 license they wanted from Intel (like I explained in my previous post).

Regarding what the "benefit" is of AMD having an x86 license, that appears to be just random nonsense. It really should be obvious to you as a knowledgeable person in this field why it's important for AMD to have an x86 license. I have no idea where you're going with that one, and I even asked you what you meant, but all you could do was ask "what's your point?" back.

 

[BlueBlazer]

Nvidia tried as hard as they could during several years of legal battles to acquire an x86 license from Intel. When they couldn't, they went with ARM instead.

That's the chipset battle. That's why you don't see newer NVIDIA chipsets from LGA1366 systems and onwards. :hmm:

AMD already has an x86 license. So what's your point really?

Look how long Intel took to develop x86 based handheld mobile chips (and still not fully launched yet). Imagine the amount of money poured into the project.

 

[Idontcare]

That's because your questions were apparently just random counters to another question that actually had some relevance and substance.

The reason Nvidia has an ARM license is that they want to compete in the CPU space as their existing GPU business is rapidly shrinking and they couldn't get the x86 license they wanted from Intel (like I explained in my previous post).

Regarding what the "benefit" is of AMD having an x86 license, that appears to be just random nonsense. It really should be obvious to you as a knowledgeable person in this field why it's important for AMD to have an x86 license. I have no idea where you're going with that one, and I even asked you what you meant, but all you could do was ask "what's your point?" back.

Its called the Socratic Method. I can't tell you about "D", but I can ask questions that ought to lead you on a journey of self-discovery by which you self-realize steps A, B, and C...and then you surmise D all on your own

I can't spell out "D" for you, there would be repurcussions of me literally dragging the stubborn ass horses to water and forcing them too drink from it, but I can ask the very questions that ought to make it painfully self-evident what the situation is like "in the real world" for those who are intrigued by it.

And no, I don't know anything, I was never here, this meeting never occured, I have no idea what any of you are talking about. I just stumbled in here looking for the bathrooms ¿dónde están los baños

 

[lol123]

Well, all right. You suggested that since Nvidia has an ARM license, AMD should get one too as that will let them take profits away from Nvidia (although Fox5 presented a good case why going the ARM route would not be the wise choice for AMD). I replied that judging by the events leading up to Nvidia developing an ARM processor (there were rumors of an x86 CPU design for a long time, coinciding with the legal dispute between Nvidia and Intel), the only reason they chose that alternative was because they weren't able to get an x86 license. AMD, on the other hand, already has an x86 license. So if AMD wants to take a cue from Nvidia, it should actually be to pursue x86 and not ARM, which is exactly what they are doing at the moment.

And so the reasonable answer to the questions you posed is actually the opposite of what you tried to submit through "Socratic method" or whatever.

 

[Medu]

AMD late to the party... again. The strategy shift they need is to merge, or be bought out, by a company with cash and lots of it. They can't compete with Intel because they don't have the resources and they won't be able to compete in the SoC market either as they don't have the resources or IP. Yes they can make fast graphics chips, but they aren't efficient enough to be put on tablet/phones,(afaik most of the 9-18w TDP on Bobcat is from the power hungry GPU). They can't afford to go out and hire a SoC team, and if they try and spread themselves any thinner they will just fall farther behind on their core business. (assume that they don't just abandon it).

 

[Idontcare]