no one seems to be considering K12, maybe they dont care about x86? there is lots of other stuff AMD offers that would take significant time to develop and implement, like an entire NUMA architecture and interconnects and server grade platforms.
I dont think they will by AMD, but i haven't seen a good reason in this thread from the naysayers as to why not.
Look at the what the involved costs might be to "revive" a long time money starved AMD. First the buyout. That includes buying out AMD's debts. Then there is the monstrous R&D deficit to feed money too. When all is said and done, and I'm no financial expert by any stretch, it would cost Samsung so much and take so long to get that money back if ever that AMD is just not an appealing buy. I'm thinking 10 to 15 billion INITIAL investment for Samsung. That is no chump change even for them.
Your making things up to suit your agenda. the only things you can really assume as a base is the cost to buy and AMD's debt levels. everything else is your rather negative connotation.
by the end of 2016 AMD should have a "high performance" ( need to be seen how well it goes) ARM 14nm server class CPU/SOC. Im going to assume amd will have HT for 2 sockets as a minimum + amd typical MCM . Assuming that Samsungs desires are server class ARM chips could they even get there in 18 months regardless of money by themselves, i doubt it.
I could just as easily say the same about you as you pretend the costs I mentioned won't exist.
Every engine needs maintenance. If you neglect the engine, the longer you wait the more potential for damage is created. When something inevitably breaks it costs many times more than what the proper maintenance would have cost. AMD R&D has been painfully cut back over the last half decade. Guess what. That isn't an option in the semiconductor industry. Years of cutting back means that every year AMD has less and less of a chance of competing with the likes of Intel, Nvidia, Qualcomm, ARM, Samsung, ad infinitum. Samsung buys AMD it starts off woefully behind in R&D. If they plan to make money they need to turn out competitive products and that takes crushing amounts of costly R&D normally. Now imagine the costs to get AMDs R&D to where it can successfully compete.
And to be honest, your remarks about my agenda were unfounded and out of place especially when my argument is fairly logical and reasonable.
Your thinking is rather narrow and short term. AMD's entire IP and engineering expertise can see benefits to Samsung's entire product lineup and more. Yet here you're making it sound like they're trying to make a quick buck of the purchase of AMD.
Your thinking is rather narrow and short term. AMD's entire IP and engineering expertise can see benefits to Samsung's entire product lineup and more. Yet here you're making it sound like they're trying to make a quick buck of the purchase of AMD.
I could just as easily say the same about you as you pretend the costs I mentioned won't exist. Every engine needs maintenance. If you neglect the engine, the longer you wait the more potential for damage is created. When something inevitably breaks it costs many times more than what the proper maintenance would have cost. AMD R&D has been painfully cut back over the last half decade. Guess what. That isn't an option in the semiconductor industry. Years of cutting back means that every year AMD has less and less of a chance of competing with the likes of Intel, Nvidia, Qualcomm, ARM, Samsung, ad infinitum. Samsung buys AMD it starts off woefully behind in R&D. If they plan to make money they need to turn out competitive products and that takes crushing amounts of costly R&D normally. Now imagine the costs to get AMDs R&D to where it can successfully compete. And to be honest, your remarks about my agenda were unfounded and out of place especially when my argument is fairly logical and reasonable.
You are assuming that what AMD are developing with Zen and K12 is not competitive :thumbsdown: By your logic AMD should not compete with Nvidia as they are not spending the same amount of money on R&D. So if AMD R9 390X matches or beats Titan-X whats your answer then ? AMD is very judicious on their R&D spending but that does not mean they cannot develop competitive products.
Ok.Ok. I can see you guys are plainly too excited at the prospect of Samsung saving AMD to think clearly. I'll not bother further.
And you are assuming that the R&D starved AMD automatically will develop 2 new CPU uarchs that is competitive with everything. not to mention GPU uarchs and what else. All with a tiny fraction of the R&D budget compared to the competition in all 3 segments.
Now tell me, who is the one with unrealistic expectations?
Someone in this thread said AMD IP isn't worth much.
Apparently their two recent wins for semi-custom SOCs in their financials indicate those alone are worth $1B over 3 years.
Do you have proof that AMD's spending on R&D is not enough to develop a competitive Zen and K12. If you don't I suggest you wait it out for reality to play itself out instead of being an armchair speculator with no clue of CPU design who makes bold claims that AMD will not succeed.
Look on the competitors R&D budgets. And tell me what magic AMD can do that everyone else cant. Something AMD beats the rest of the industry with 10 fold R&D wise in everything.
Its entire products. And 1B$ over 3 years. Do you think thats much? Thats 83M$ per quarter in revenue. Now take a guess on the margins.
You realize that AMD need not outright beat Intel in performance and still remain competitive. Bulldozer was and is the the root of all of AMD's problems. Zen does not need to beat Skylake. Zen needs to be competitive with Skylake and not fall far behind like Bulldozer did. Thats surely possible.
btw you think Jim Keller has no ideas of how to develop a good CPU core. He has worked on some of the best CPU architectures - DEC Alpha, Hammer and Apple custom ARM cores. Surely he has an idea of how to develop a competitive high performance CPU core.
You must have missed the announcement. MediaTek just paid for AMD Graphics IP for their Arm SOC.
AMD isn't selling them products with "margins" associated due to production costs.
He had a hell of a lot more money at Apple for example. And he only had to create 1 product.
Zen needs to compete with Intel. K12 needs to compete with MediaTek, Qualcomm, Samsung, Applied Micro, Broadcom and who else. And it needs to compete with nVidia in graphics. All with a tiny R&D budget.
So your conclusion is, Jim Keller fixes everything here? I assume Apple lost everything now since this famous person that singlehanded turns everythign around left.
Samsung SoCs are not rubbish and their performance is far from horrible.
http://bgr.com/2015/03/02/galaxy-s6-vs-htc-one-m9-benchmark/
You know, I guess I wouldn't necessarily mind or care, but dang, I really wish the American multinationals could stay, well, American. It was sort of a huge blow that the American pioneer in communications equipment in just about every field, Motorola, struggled terribly and, when all was said and done, wound up in the hands of a Chinese multinational corporation (Lenovo).
It's great that Intel and Nvidia are still raking in cash, but it would be great if AMD could turn around and remain a fully American property.
Note: not to beat the American war drums or anything for those of you that are not American, but it's a piece of our pioneering history and pride, especially factoring in that what was once the norm that the pioneering corporations were American, most of them have fallen by the wayside, or otherwise the pioneering has been done elsewhere.
I'm missing something... why you think that Samsung spending money on AMD RnD is a waste of money? Why would it be any worse than spending on their own RnD teams that are way behind in GPU IP for example?
AMD is all over multiplatform ecosystem with their low level approach. They develop new multi-arch platform. Their portfolio is ranging from mobile to server-class SOCs.
IMHO, cash starved amd innovating in many areas is the best long term investment for someone big like SS