AMD CEO talks of long-term turnaround

[ShintaiDK]

High performance ZEN CPUs for Server and Desktop.
High performance APUs for Mobile (Laptops) and Desktop
High performance GPUs for HPC, professionals and Desktop Gaming.

Nothing VIA related in those products.

1. Zen doesnt exist yet.
2. Doesnt sell. Bleeding badly.
3. Losing marketshare like crazy.

You clearly dont understand the compare to VIA.

 

[dark zero]

I actually agree with that sentiment. I do believe AMD is shrinking to become a smaller player like Via. I don't see AMD going anywhere unless someone buys them. They will probably just muddle along like they've done for most of their existence. VIA wasn't particularly profitable when they had all that marketshare a decade ago -- so they shrunk down their operations and focused on improving margins. I don't understand why anyone would view that as a bad thing. Better to be profitable and agile -- than bloated and bleeding.

If that happen, Intel would easily kill their own industry to move to another one... anyways, they have money.

 

[frozentundra123456]

If that happen, Intel would easily kill their own industry to move to another one... anyways, they have money.

nvm

 

[myocardia]

Really? So what market segment would you put them in?

The slow one. The one where AMD's fastest, 220 watt factory overclocked CPU regularly gets beaten by their competitor's 54 watt dual-core.

Not making the fastest option in a segment does not mean that the products do not belong to that segment.

It does as far as I'm concerned. Well, it does if your products get trounced by the percentages that AMD's products do. And when I say 'trounced', I mean using the software(s) that people actually want to use, not that one that you're busily looking for a graph of, where the 220 watt 9590 can actually beat a 54 watt i3. Seriously, 13 people on Earth care how fast a processor that they own can zip or unzip files, and all 13 of them already own either i7-5960s or Xeons.

 

[myocardia]

Edit. What I believe Lisa meant is that they don't want to develop for open PC's. It does not mean that they will stop developing GPUs and CPUs. The market is going in the direction of closed architectures. Consoles, laptops, smartphones, tablets, all-in-ones. You will not buy one part, but whole computer. That way of design gives every producer more revenue income every year, with designed obsolesce. Is it hard to see it?

So, you're saying that they are going to abandon the only people left who actually buy their CPUs/APUs, besides the console companies, of course. Does AMD even have a 1% market share in 2015, in the pre-made computers/laptops/tablet market that you mentioned? I would swear that they do not.

 

[Fjodor2001]

The slow one. The one where AMD's fastest, 220 watt factory overclocked CPU regularly gets beaten by their competitor's 54 watt dual-core.

I guess you're trying to be funny, but you're only making a fool of yourself.

It does as far as I'm concerned. Well, it does if your products get trounced by the percentages that AMD's products do. And when I say 'trounced', I mean using the software(s) that people actually want to use, not that one that you're busily looking for a graph of, where the 220 watt 9590 can actually beat a 54 watt i3. Seriously, 13 people on Earth care how fast a processor that they own can zip or unzip files, and all 13 of them already own either i7-5960s or Xeons.

Ok, so Ferrari is not in the luxury car segment, because there are other cars that are faster and more expensive.

And Western Digital is not in the enterprise hard drive market, because Seagate's biggest HDD can store more than WD's.

 

[dark zero]

nvm

If that sound weird, I must agree, sounds extreme, but I am seeing not from the death of Intel... but of x86. Actually the architecture has their limits, so... do you think that Intel is living only on x86?
Definately not. Maybe they are planning to make their succesor.
Anyways... PC industry is dying in favor for Tablet and phone one. And Intel is not scaling well on x86... so why not create another one and according to the current era?
They could only ditch x86 in favor on their succesor... they ended Alpha with Itanium.. why not taking down x86?

 

[aigomorla]

did any of u guys see this and can tell me if there is any truth in it.

http://www.zdnet.com/article/why-microsoft-buying-amd-would-be-a-bad-idea/

???

 

[ShintaiDK]

Unlikely. Just like all the previous "buyers".

It would be nothing more than a money drainer.

 

[mrmt]

did any of u guys see this and can tell me if there is any truth in it.

http://www.zdnet.com/article/why-microsoft-buying-amd-would-be-a-bad-idea/

???

Instead of paying $100 to AMD they would end up paying $85 to TSMC, assume all the AMD debt, deal with a failing business and inherit the WSA with Globalfoundries until 2024. Not a winning proposition in my book, makes the Nokia deal look a bargain.

 

[dark zero]

If even MS is ditching Nokia... Why buying AMD? Is easier to them to merge with Intel in order to ditch Linux and Apple at the same time.

 

[pablo87]

did any of u guys see this and can tell me if there is any truth in it.

http://www.zdnet.com/article/why-microsoft-buying-amd-would-be-a-bad-idea/

???

I doubt it would be good for MS to swallow the whole thing. However, if GPU and CPU were to split, then for MS to strategically invest in GPU would make a ton of sense. Ditto Sony, Samsung. Edit: WAG, 20% would cost them $400-600MM. A whole lot less than Nokia with a lot less competition and probably more upside.

What we do know is that AMD as currently structured has been an abject failure, the biggest beneficiaries being Intel and Nvidia.

With the HBM inflection point, x86's need for R&D funding and Oil producers being in a world of pain, the timing could not be more ideal.

Unfortunately, when AMD's Board of Directors arrive at the fork in the road, they usually don't follow Yogi's advice.

 

[jpiniero]

Samsung, Apple, et al would be interested in the GPU IP but it wouldn't be anywhere near enough to pay down the debt and the CPU business is basically worth nothing.

 

[dark zero]

Samsung, Apple, et al would be interested in the GPU IP but it wouldn't be anywhere near enough to pay down the debt and the CPU business is basically worth nothing.

Apple would LOVE AMD CPU since that would finally give them independance from Intel. Also... as long AMD is still alive, Intel would make improvements, so if CPU department dies, x86 would start to be stuck and dies eventually due the lack of improvements... it would end into another Power PC.

 

[pablo87]

Give me an independent ATI/GPU today at 1/6th the value of Nvidia and any analyst worth his salt would have a buy recommendation.

How AMD GPU managed to trump Nvidia with HBM after focusing on APUs and Consoles for the past 5 years speaks volumes about the opportunity an unshackled GPU would be. And it isn't exactly a ringing endorsement for Nvidia's strategic vision.

 

[Phynaz]

Apple would LOVE AMD CPU since that would finally give them independance from Intel. Also... as long AMD is still alive, Intel would make improvements, so if CPU department dies, x86 would start to be stuck and dies eventually due the lack of improvements... it would end into another Power PC.

Why would Apple want independence from Intel? They could have it today by buying CPUs from AMD.

 

[ClockHound]

Why would Apple want independence from Intel? They could have it today by buying CPUs from AMD.

Or Apple could just buy AMD and Intel and clear the Windows rabble off the internet, bringing peace and stability to the walled garden. A shiny, no-user-replaceable parts utopia. Doesn't that sound appealing?



/sarcasm

 

[mrmt]

How AMD GPU managed to trump Nvidia with HBM after focusing on APUs and Consoles for the past 5 years speaks volumes about the opportunity an unshackled GPU would be. And it isn't exactly a ringing endorsement for Nvidia's strategic vision.

Interesting Pablo, I see AMD adoption of HBM as yet another failure for the company. They basically adopted HBM in a time where HBM is unsuitable for professional products, bottom market products and APU products. Basically HBM products are only suitable for a very small niche. Nvidia OTOH built a very solid architectural foundation, and by the time HBM2 arrives they will have a solid solution on the wings for discrete cards.

With HBM AMD basically misread the market and shot themselves in the foot, being unable to compete on the mobile market and only being able to field HBM in a very small range of products. Was it worth to sacrifice the mobile market and the rest of the product stack in order to field, what, three or four HBM SKUs, that costs more to make than comparable Nvidia products? I doubt it.

 

[MongGrel]

I guess you're trying to be funny, but you're only making a fool of yourself.

And Western Digital is not in the enterprise hard drive market, because Seagate's biggest HDD can store more than WD's.

Um, yeah right.

:whiste:

 

[turtile]

Interesting Pablo, I see AMD adoption of HBM as yet another failure for the company. They basically adopted HBM in a time where HBM is unsuitable for professional products, bottom market products and APU products. Basically HBM products are only suitable for a very small niche. Nvidia OTOH built a very solid architectural foundation, and by the time HBM2 arrives they will have a solid solution on the wings for discrete cards.

With HBM AMD basically misread the market and shot themselves in the foot, being unable to compete on the mobile market and only being able to field HBM in a very small range of products. Was it worth to sacrifice the mobile market and the rest of the product stack in order to field, what, three or four HBM SKUs, that costs more to make than comparable Nvidia products? I doubt it.

Fury's power consumption would be huge without HBM. AMD also spent years and years of work on HBM with the ability to use it before anyone else. Yes, they didn't have the money to develop a new architecture with it but that doesn't make it a failure when they started development so many years ago.

 

[CHADBOGA]

Fury's power consumption would be huge without HBM. AMD also spent years and years of work on HBM with the ability to use it before anyone else. Yes, they didn't have the money to develop a new architecture with it but that doesn't make it a failure when they started development so many years ago.

If they can't turn it into meaningful marketshare gains, it will undoubtedly be a failure.

 

[myocardia]

I guess you're trying to be funny, but you're only making a fool of yourself.

Actually, I don't find it funny in the least that AMD's 220 watt, factory-clocked CPU regularly gets embarassed by their competitor's 54 watt dual-core. Does that strike you as funny?

Ok, so Ferrari is not in the luxury car segment, because there are other cars that are faster and more expensive.

No, Ferrari isn't in the luxury car segment for an entire list of other reasons, that have nothing at all to do with either speed or cost. Also, since when was it a necessity for luxury cars to be fast? You might want to notify the most widely-accepted-as-luxury car manufacturer on Earth, Rolls Royce, about that. BTW, where exactly did the guy you quoted mention Ferraris, or even automobiles at all?

And Western Digital is not in the enterprise hard drive market, because Seagate's biggest HDD can store more than WD's.

Again, exactly where did I make mention of hard drives, enterprise or other? You know, if you have no ability to make arguments that support your claim(s), it's really better if you don't reply. It makes you look like much less of a moron.

 

[pablo87]

Interesting Pablo, I see AMD adoption of HBM as yet another failure for the company. They basically adopted HBM in a time where HBM is unsuitable for professional products, bottom market products and APU products. Basically HBM products are only suitable for a very small niche. Nvidia OTOH built a very solid architectural foundation, and by the time HBM2 arrives they will have a solid solution on the wings for discrete cards.

With HBM AMD basically misread the market and shot themselves in the foot, being unable to compete on the mobile market and only being able to field HBM in a very small range of products. Was it worth to sacrifice the mobile market and the rest of the product stack in order to field, what, three or four HBM SKUs, that costs more to make than comparable Nvidia products? I doubt it.

HBM is an inflection point for the video card business (and maybe the entire PC industry) and AMD is leading the charge. Its a feather in their cap. Whether they made the right decision short term is moot; what matters is the long term and there I think they hit a homerun.

I'm shocked tbh how much AMD'S competitive position improved with HBM. Big morale boost for the Red Team. I think even if the BoD said no (they probably haven't a clue what is HBM other than they're cutting some big cheques to Hynix), they'd still go ahead with HBM top to bottom somehow - its the soul of a new machine.

The best is yet to come. I think mobile gaming and desktop cards under $150 will be rejuvenated by HBM. Does the industry ever need it.

 

[turtile]

If they can't turn it into meaningful marketshare gains, it will undoubtedly be a failure.

You mean like the fact that it's enabling them to pull money out of a very old architecture and keeping it competitive. Marketshare doesn't matter to a company that needs to make profits.

 

[ShintaiDK]

HBM is an inflection point for the video card business (and maybe the entire PC industry) and AMD is leading the charge. Its a feather in their cap. Whether they made the right decision short term is moot; what matters is the long term and there I think they hit a homerun.

I'm shocked tbh how much AMD'S competitive position improved with HBM. Big morale boost for the Red Team. I think even if the BoD said no (they probably haven't a clue what is HBM other than they're cutting some big cheques to Hynix), they'd still go ahead with HBM top to bottom somehow - its the soul of a new machine.

The best is yet to come. I think mobile gaming and desktop cards under $150 will be rejuvenated by HBM. Does the industry ever need it.

You act like nVidia cant do the same. But they can and will when the time is right.

The problem still exist and it shows, AMD jumped on the HBM wagon way too soon. At the end of the day, HBM was nothing more than an expensive bandaid for the lack of a new efficient GCN uarch. And when nVidia does the same, we are back to before. Not just in the high end, but in the entire range.

No mobile, no professional segment, just a tiny highend section where AMD may sell how many cards again?

Just look at the steam survey for reference.

What AMD did was to make a dejavou of the native quadcore. Look look, we are first....

AMD just released a 32GB Hawaii card for the same reason.