AMD Now Almost Worth A Quarter of What it Paid for ATI

[medi01]

AMD needs to make products people want to buy....

And processors that can perform well and are priced well....

The Athlon 64 was released over 10 years ago It was a Pentium 4 killer...

AMD needs another Athlon 64.... Or they could go the way of mips and SGI

AMD created Carrizo I want to buy.
No notebook manufacturer gives a damn.

Even before Carrizo, I'd rather go with AMD's APU than Intel's overpriced CPUs that suck at gaming, but alas, neither that was possible.

Both Microsoft and Sony had good reasons to go with AMDs APU in their current gen consoles.

So, no, it's not really about the product.

 

[medi01]

Indeed. A year after the acquisition, the combined company was worth half its combined (pre-acquisition) value, thanks to AMD's gross over-valuation of ATI worth, and the $2bn worth of debt AMD saddled itself with buying ATI. Having said that, AMD's BoD and execs have been continually rewarded for underperformance. The occasional Chief [insert title] Officer gets the golden parachute treatment as a sacrificial lamb for the combined assent of the BoD's signing off on whatever strategies are in vogue at any given time, but generally it's business as usual.

AMD couldn't produce APUs, and would not be used by Microsfot/Sony if they didn't buy ATI.

 

[medi01]

Sigh. There are only 2 reasons why AMD is manufacturing console hardware:

1. it gives them a steady income source so their bottom line doesn't look as horrible as it actually is
2. it gives AMD's directors something to distract shareholders with when the hard questions about profitability start coming in

Neither of these imply that the console deal is anything but barely profitable. That is why nVIDIA told the console companies to take a hike - because nVDIA knew it could make more money selling discrete graphics cards. And they have.

AMD on the other hand, were essentially forced to take the console deal to stay in business - in exactly the same way they had to sell their fabs to stay in business. Neither of those decisions were best for the business long-term, but they were required if there was to be a business at all.

Console deal is more than it looks like.
It's hard to even imagine how on earth it is in any way negative for AMD.
It brings profit.
It forces game developers to go multi-threaded (AMDs greates weeknes and Intel's greatest advantage is single thread performance)

 

[FordGT90Concept]

AMD's Jaguar CPU in the consoles is extremely weak (think Intel Atom processor). The GPU and RAM are the only powerful components of those consoles. AMD's experience with GDDR5 particularly helped with the design of PlayStation 4's hybrid memory architecture.

 

[ZeDestructor]

Also, a wonderful argument regarding sockets. LGA775 was a gong show. DDR2 and DDR3, FSB 800, 1066 and 1333 all existing on one socket made for a hell of a mess. The current socket strategy makes it much easier for less-knowledgeable consumers to get something that works.

Even for the more experienced among us, it just makes life easier, simpler and faster.

AMD created Carrizo I want to buy.
No notebook manufacturer gives a damn.

Even before Carrizo, I'd rather go with AMD's APU than Intel's overpriced CPUs that suck at gaming, but alas, neither that was possible.

Both Microsoft and Sony had good reasons to go with AMDs APU in their current gen consoles.

So, no, it's not really about the product.

AMD has been for a very long time uncompetitive in terms of absolute power consumption at low loads. This shuts them out of the bulk laptop market that doesn't give a shit about GPUs besides making youtube/netflix play smoothly.

Console deal is more than it looks like.
It's hard to even imagine how on earth it is in any way negative for AMD.
It brings profit.
It forces game developers to go multi-threaded (AMDs greates weeknes and Intel's greatest advantage is single thread performance)

As I explained earlier in this thread, it's an incredibly minor win. If Nvidia, Intel, IBM, Qcomm, etc wanted the miserly margins of consoles, they'd be on at least one of the consoles. All of them walked away instead, and that's very telling of the state of the console hardware side.

 

[EarthDog]

The current socket strategy makes it much easier for less-knowledgeable consumers to get something that works.

Less knowledgeable consumers buy OEM. Most of those users only upgrade HDDs or SSDs/Optical/and Videocards.

Intel just can't win... You have a long lived socket and people complain that there are too many options for it. They have their standard fare now lasting ~2 years and its not long enough... I dont get it.

 

[ZeDestructor]

Less knowledgeable consumers buy OEM. Most of those users only upgrade HDDs or SSDs/Optical/and Videocards.

Intel just can't win... You have a long lived socket and people complain that there are too many options for it. They have their standard fare now lasting ~2 years and its not long enough... I dont get it.

Intel doesn't care. They decided after the clusterfuck that was LGA775 to go for the current one socket per generation approach, nice and easy, and on the server and laptop side, have never cared.

 

[EarthDog]

Me either. Educated consumer FTW.

 

[Dent1]

AMD needs to make products people want to buy....

And processors that can perform well and are priced well....

The Athlon 64 was released over 10 years ago It was a Pentium 4 killer...

AMD needs another Athlon 64.... Or they could go the way of mips and SGI

Yeah and people still bought the clapped out Pentium 4. So obviously you have no clue what you're talking about.

 

[ZeDestructor]

Yeah and people still bought the clapped out Pentium 4. So obviously you have no clue what you're talking about.

Largely irrelevant: AMD sold Opterons by the tray. Server is where the real money is made.

 

[HumanSmoke]

Largely irrelevant: AMD sold Opterons by the tray. Server is where the real money is made.

Very true.
While Intel's overall growth has been pretty much flat, since AMD failed to successfully update their C32/G34 based platforms, Intel's enterprise revenue and profit lines is clear for all to see, while AMD's dithering, slow reactions, and lack of focus have ceded the ARM-based market to companies like Applied Micro and Cavium.

[Source]
Makes me wonder if AMD's enterprise division cringe every time they look back on their bold claims of the past. They'd better pull their finger out if they intend to get that 25% market share - presently they are looking down the barrel at 1.5% of x86, and well under 1% total taking into account RISC and ARM.

 

[ZeDestructor]

Very true.
While Intel's overall growth has been pretty much flat, since AMD failed to successfully update their C32/G34 based platforms, Intel's enterprise revenue and profit lines is clear for all to see, while AMD's dithering, slow reactions, and lack of focus have ceded the ARM-based market to companies like Applied Micro and Cavium.

[Source]
Makes me wonder if AMD's enterprise division cringe every time they look back on their bold claims of the past. They'd better pull their finger out if they intend to get that 25% market share - presently they are looking down the barrel at 1.5% of x86, and well under 1% total taking into account RISC and ARM.

Right now AMD is stuck on 32nm Bulldozer/Vishera/Steamroller. Intel meanwhile is sampling 14nm and has been shipping 22nm for a while, on a superior microarchitecture with more cores and hyperthreading on those increased core numbers. Zen/K12 need to succeed. Of the two, I expect K12 to be the more immediately successful one of the two (mainly because it will go up to the rather less experienced AppliedMicro/Cavium in the server space), but if Zen delivers... Man, we really need more than just comparing current Intel server power usage and IO vs new Intel server power usage and IO on the server end, given IBM's horrendous POWER8 pricing and ARM's overall laughable high-end performance. Sure, ARM is lower in absolute power, but in performance/W, Intel rules the roost, with IBM's POWER8 (barely) being the party popper to industry-wide domination.