You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
There was a rumor [1][2] that Altera was unhappy with Intel's process and was going to bail. Intel did buy Altera in order to prevent it from defecting to Tsmc.
As for all the talk about using FPGA's as a compute engine - well it was possible to achieve that using a strategic partnership - no need to spends $16 billion.
No, Altera leaving Intel Custom Foundry was a total bluff. Altera cannot compete with Xilinx at the same foundry. Reference 28nm and 20nm. Here is the Intel call and slides on the Altera acquisition:
Intel Acquires Altera Conference Call
<nobr>06/01/15</nobr> -7:00 AM PT
Speaker: Brian Krzanich, Stacy Smith, CEO, CFO [table] cellpadding="1"
|-
| style="width: 16px" |
| Listen to webcast
There was a rumor [1][2] that Altera was unhappy with Intel's process and was going to bail. Intel did buy Altera in order to prevent it from defecting to Tsmc.
As for all the talk about using FPGA's as a compute engine - well it was possible to achieve that using a strategic partnership - no need to spends $16 billion.
I highly doubt that as a foundry customer Altera is worth $17B to Intel (I am not sure Altera paid a single $ to Intel yet). There must be more to this story than this (and Intel does have an official explanation for their motives which sound at least somewhat plausible).
Isn't it true that the data center business is only a tiny part of Altera's overall market?
So, if Intel are buying Altera exclusively for integration with data servers, then what do they do with all the other parts of the business?
Skeptical of the wisdom of the Atlera acquisition, BMO's Ambrish Srivastava has downgraded Intel (INTC-2%) to Market Perform, and cut his target by $7 to $33.
Srivastava: "We really do not see the opportunity/rationale that Intel is laying out for parting with ~$17 billion. Specifically, we do not concur with Intel's assumption of 7 percent CAGR for the business. Altera has grown at a negative CAGR of 2 percent in the midst of one of the more aggressive carrier CapEx builds we have seen."
He also wants more details about the deal. "[W]e do not know ultimately how much debt Intel will end up issuing to finance the deal, potentially in the range of $7 billion to $12 billion, in the medium turn, it could lead to potentially reduced capital return via share repurchase..."
Isn't it true that the data center business is only a tiny part of Altera's overall market?
So, if Intel are buying Altera exclusively for integration with data servers, then what do they do with all the other parts of the business?
It's true. It is also true that data centers is a huge part of Intel business. So the analysts are saying that Intel needs FPGAs to maintain an edge in this business. I think that from the purely computing perspective it makes perfect sense. I am not qualified to say whether it also makes sense from a business perspective. In general though, with PC business stalling (to put it mildly) it is probably inevitable that Intel should start looking at expanding into other areas (like storage, memory, mobile, HPC etc.)
It is a well known/well researched fact that only a minority (said to be about 30%) of major acquisitions add value to shareholders. Most result in losses. I have searched in vain for a concise statement of why Intel has splashed out all this money. I am unwilling to believe that Krysanich has acted for no reason at all. The server market benefits appear limited. Locking in a foundry customer seems ridiculous. Can anyone assist me?
