My current thought is the best deal for MIPS is to sell its patent portfolio while retaining rights to it as IPWireless did in a deal with Nvidia and IV this week.
That gives MIPS a solid cash injection and the freedom to develop next-gen products.
Hi, Moronda. ARM may have won the war, as you say, but I wouldn't discount MIPS patents so quickly.
I wouldn't discount Sandeep so quickly either. Much of the difficulties MIPS had gone through are due to what happened even before Sandeep arrived at MIPS.
Lastly, I wouldn't want any company to go down the toilet. Every company, and every engineer on the team, deserves to do its/his best and eventually succeed.
The converse question to "Who might buy MIPS?" is "Why should MIPS want to sell?" I think the answer is probably obvious. Size matters, and MIPS feels it needs to be part of a bigger, better heeled company to survive and compete long-term in the CPU/GPU market.
And it isn't clear that ARM has won the war. The current war I see is all about power efficiency, and ARM is poised to make inroads into the server market because ARM cores use less power than Intel. It's even more vital in the mobile device market, where the scarce resource is battery charge, and you win by being able to work longer without needing to recharge.
I think a lot will hinge on just how good the newly announced MIPS designs are. If they are as power efficient as claimed, they'll get a lot of interest.
And while I don't expect it would happen, I know a company that might find MIPS a worthwhile acquisition: Intel. Intel has little penetration in the mobile device market, and the Atom processor hasn't really been able to compete with ARM. Intel might find it worthwhile to add MIPS both for the patent portfolio and the opportunity to better compete with ARM in the mobile market. They can certainly afford the price.
Intel used to license ARM IP, and had its own line of StrongARM processors, before selling it to Freescale in a reorganization. If they bought MIPS, they wouldn't be licensing someone else's technology. It would be theirs, because they bought it. And Intel has the fabs and the process technology to manufacture MIPS based chips in any form desired.
I don't expect it to occur for corporate ego reasons - it would be an implicit admission Intel has been unable to compete with X86 based designs in the mobile space, and had to acquire different technology they didn't create to do so. But it might make sense for them to do so.