AMD Buying SeaMicro for $334 Million

[Face]

Advanced Micro Devices on Wednesday announced a deal to acquire startup SeaMicro, a maker of microservers and data center fabric technology, for $334 million. SeaMicro chief executive Andrew Feldman will become general manager of a newly formed AMD business unit called Data Center Server Solutions.
"SeaMicro's leading-edge fabric technology together with AMD Opteron processors point the way to a compelling roadmap for the cloud data center market," said Lisa Su, senior vice president and general manager of AMD's Global Business Units, in a discussion with reporters at the chip maker's Sunnyvale, Calif. headquarters ahead of the announcement.

http://www.pcmag.com/article2/0,2817,2401008,00.asp

[breakfromyou]

obligatory post making fun of AMD's financial situation:

Wait, this means AMD has money?

[Hornet331]

obligatory post making fun of AMD's financial situation:

Wait, this means AMD has money?

No, you can acquire other companies trough other deals, like share exchange or take a new loan to buy it.

But in this case I think, they just used the money from the settlement with intel.

[Chrono Detector]

Bit surprised that AMD managed to buy yet another company for quite a large sum despite their recent fiascos, but its nice to see AMD acquiring a company which will benefit them.

[The Jesus]

No, you can acquire other companies trough other deals, like share exchange or take a new loan to buy it.

But in this case I think, they just used the money from the settlement with intel.

Must've missed that, which settlement?

[generics_user]

Must've missed that, which settlement?

AMD v. Intel

amd received 1.25b from intel back in 2010

[AliG]

No, you can acquire other companies trough other deals, like share exchange or take a new loan to buy it.

But in this case I think, they just used the money from the settlement with intel.

I'm guessing this was only possible due to the (slightly) improving economy and was some sort of leveraged buyout

AMD v. Intel

amd received 1.25b from intel back in 2010

I still think it's biting them in the ass that they sold out so quickly in their lawsuit, but as I recall they really had no choice but to take whatever Intel offered at the time to survive

[The Jesus]

Hmm, I must've missed that. Not bad and definitely true.

[savantu]

How does AMD benefit by competing with their major server customers like HP, Dell, IBM ? All of these have shown interest in high density, low power servers, some are even selling rebadged Seamicro units or developing systems of their own.

[The Jesus]

How does AMD benefit by competing with their major server customers like HP, Dell, IBM ? All of these have shown interest in high density, low power servers, some are even selling rebadged Seamicro units or developing systems of their own.

They probably aren't planning on competing with them, just selling them the software SeaMicro developed to mesh multiple servers together as one.

And compete with ARM.

[AbortRetryFail?]

How does AMD benefit by competing with their major server customers like HP, Dell, IBM ? All of these have shown interest in high density, low power servers, some are even selling rebadged Seamicro units or developing systems of their own.

I think you answered your own question.

They will continue to meet OEM demand for the Intel platforms (but, for how long? ).

I suspect they will integrate Seamicro IP into Opteron designs, and also develop their own low-power enterprise solutions (one of the new VPs is into this hybrid RISC/CISC design thing).

[Russian]

obligatory post making fun of AMD's financial situation:

Wait, this means AMD has money?

They have some, they've paid down their debt significantly. I'm sure they are able to secure favorable financing.

[Origin_Unknown]

"But, but, but AMD haz no monies" cries all the doom and gloom merchants, tears of Intel blue rolling down there faces and down there neck beards
"h-h-how can it be" more cries are heard

[SoF]

Nice move AMD - when you will try to buy Nvidia and end up like Porsche vs. Volkswagen?!

[kuroikenshi]

^^ yeah straight after VW also purchased the rights to manufacture Buggati and then made the Veyron, god I love that things moving at over 400Km/h (317mph)

[savantu]

They probably aren't planning on competing with them, just selling them the software SeaMicro developed to mesh multiple servers together as one.

And compete with ARM.

Software ? It runs industry standard software. The whole thing is about HW, their expertise is in the interconnect and cramming the CPUs with their node controllers on tiny boards.

How do they compete with ARM ? What ARM branded servers are in the market ? HP for one, is making ARM servers. That means they will compete with HP. Very smart move..

I think you answered your own question.

They will continue to meet OEM demand for the Intel platforms (but, for how long? ).

I suspect they will integrate Seamicro IP into Opteron designs, and also develop their own low-power enterprise solutions (one of the new VPs is into this hybrid RISC/CISC design thing).

The last thing AMD needs is to develop " their own enterprise solutions". Nothing is worse then competing with your own customers, especially when they are 10-20 times your size.

The fool at the head, spent $300m on a very doubtful acquisition which doesn't help any of AMD's major concerns :
-solutions for the fast growth smartphone/tablet market
-developing a new high performance desktop/server uarch to keep them in the game in the long term
-accelerate their transition to GPGPU enabled solutions comparable to Nvidia's stack;

[STEvil]

Software ? It runs industry standard software. The whole thing is about HW, their expertise is in the interconnect and cramming the CPUs with their node controllers on tiny boards.

How do they compete with ARM ? What ARM branded servers are in the market ? HP for one, is making ARM servers. That means they will compete with HP. Very smart move..



The last thing AMD needs is to develop " their own enterprise solutions". Nothing is worse then competing with your own customers, especially when they are 10-20 times your size.

The fool at the head, spent $300m on a very doubtful acquisition which doesn't help any of AMD's major concerns :
-solutions for the fast growth smartphone/tablet market
-developing a new high performance desktop/server uarch to keep them in the game in the long term
-accelerate their transition to GPGPU enabled solutions comparable to Nvidia's stack;

Because you know everything AMD has and is doing thats not public/investor knowledge apparently.

Your criticism is hardly ever balanced :/

[The Jesus]

Software ? It runs industry standard software. The whole thing is about HW, their expertise is in the interconnect and cramming the CPUs with their node controllers on tiny boards.

How do they compete with ARM ? What ARM branded servers are in the market ? HP for one, is making ARM servers. That means they will compete with HP. Very smart move..

One of SeaMicro's key innovations is a "fabric" that hooks up thousands of processors, memory units, and storage devices into a sensible whole for cloud computing. Rather than trying to compete with its own server-building customers, AMD may well offer them SeaMicro's platform on license and seek to recoup its $330 million investment that way.

[R101]

Software ? It runs industry standard software. The whole thing is about HW, their expertise is in the interconnect and cramming the CPUs with their node controllers on tiny boards.

How do they compete with ARM ? What ARM branded servers are in the market ? HP for one, is making ARM servers. That means they will compete with HP. Very smart move..



The last thing AMD needs is to develop " their own enterprise solutions". Nothing is worse then competing with your own customers, especially when they are 10-20 times your size.

The fool at the head, spent $300m on a very doubtful acquisition which doesn't help any of AMD's major concerns :
-solutions for the fast growth smartphone/tablet market
-developing a new high performance desktop/server uarch to keep them in the game in the long term
-accelerate their transition to GPGPU enabled solutions comparable to Nvidia's stack;

Intel makes their own server stuff.. It is reliable and relatively cheap. I have no problem with AMD doing the same.

[savantu]

Because you know everything AMD has and is doing thats not public/investor knowledge apparently.

The board ousted Meyer for failling miserably to adress two growth opportunities :
- smartphone/tablet space
- servers

What they did, doesn't improve their competitive position in either. In servers, they are competing with their own major customers and the market for this kind of servers is tiny.

Your criticism is hardly ever balanced :/

There is a thing called critical thinking, questioning what you see and what you're told... Why don't you activate it for a moment and think on the bigger scheme of things ? What are AMD's weaknesses and how exactly does this acquisition solves them ? How does this affect AMD's damaged reputation with server makers ? Their market share is the lowest ever since Opteron made it's debut in 2003. Don't you think this has something to do with the major servers vendors ? How does buying Seamicro improve the relationship with them ?

One of SeaMicro's key innovations is a "fabric" that hooks up thousands of processors, memory units, and storage devices into a sensible whole for cloud computing. Rather than trying to compete with its own server-building customers, AMD may well offer them SeaMicro's platform on license and seek to recoup its $330 million investment that way.

FYI, that's been done since Cray and SGI pioneered HPC. The difference is to do it with low-cost components in a power optimized system. Servers vendors suffer from the NIH syndrome, if any one of the major players thought Seamicro IP and design were valuable, they would have bought them a long time ago. Do you suppose either IBM, HP, Dell, Fujitsu, Cisco or Oracle did not have $300m at hand ?

If they didn't buy it/licensed it themselves, what incentive do they have to buy a license from AMD ?

Intel makes their own server stuff.. It is reliable and relatively cheap. I have no problem with AMD doing the same.

And it's being sold to whitebox makers who sell you the server and offer you a Linux or Windows CD with it. Does selling Seamicro servers seem like a similar business model to you ?

[Olivon]

We all know the AMD's biggest problem : Execution. Time to deliver to market.
If AMD don't resolve that, they will continue lagging behind blue team who got 2 nodes in advance.

[R101]

And it's being sold to whitebox makers who sell you the server and offer you a Linux or Windows CD with it. Does selling Seamicro servers seem like a similar business model to you ?

It seems to me that Seamicro is an ok partner for starting something similar.

[kuroikenshi]

now I have never played with server boards, for those like Stevil and Dave, do you think in your own opinion this acquisition will help AMD's server segment?

[EniGmA1987]

We all know the AMD's biggest problem : Execution. Time to deliver to market.
If AMD don't resolve that, they will continue lagging behind blue team who got 2 nodes in advance.

I think the execution problem stems more from lack of monetary resources compared to Intel. Just look at both companies R&D budgets and see how insignificant AMD is compared to Intel in development. AMD does some truly amazing work for the funds they spend.

[STEvil]

now I have never played with server boards, for those like Stevil and Dave, do you think in your own opinion this acquisition will help AMD's server segment?

Possibly, as well as adding a patent portfolio to avoid expensive legal litigation.

Also the fact that some of the management worked for AMD plays a large role into this acquisition.


but ya know, i've got no critical thinking..