SAN FRANCISCO -- I’m betting Advanced Micro Devices has or plans to get an ARM architectural license. But the big question is whether it will survive long enough to use it.
AMD staked out the unique position of being the first company designing both x86 and ARM server chips. But it’s a dicey position to be in .
Today Intel sells server chips that cost $2,000 or more. AMD struggles to make a business selling chips that have similar performance at half the price. When we get to the ARM server SoC era in 2014 we may be talking about $30 as a high-end part. [Get a 10% discount on ARM TechCon 2012 conference passes by using promo code EDIT. Click here to learn about the show and register.]
Ouch! Ask AMD CEO Rory Read how he will ride that water slide and he talks about the move from custom designs to SoCs. Sure, you can turn out more chips faster with fewer engineers when building them out of existing intellectual property blocks. But there are still a few hitches to contend with.
IP blocks often have to be re-jiggered when you drop them into a new SoC, especially if it’s in a new process or linking to blocks other than those in the previous design. That’s especially the case if your company is fairly new to SoC design, as is AMD.
Both Read and his general manager of products, Lisa Su, were clear AMD is not getting out of custom design altogether. AMD will still design its own x86 cores—and I suspect custom ARM cores in the future.
Neither Read or Su would answer a direct question about whether AMD has taken an ARM architectural license needed to design custom ARM cores. But the indication I get from both of them is they will do custom cores, given enough time.
Both execs were clear they will retain custom engineering capabilities. Su noted AMD recently hired processor guru Jim Keller, who has worked both on past AMD Opteron custom CPUs and ARM SoCs.
Custom work is certainly not on the near-term horizon. Initially AMD is just licensing ARM’s Atlas 64-bit core to pack into its 2014 chips. But further out it could get a chance to expand its horizon.
AMD has announced that they are launching a new ARM Cortex-A57 64bit ARMv8 Processor in 2014, targetted for the servers market.
ARM is getting hot lately, and they do compete with INTEL on server-side thus, it is not a bad strategy that AMD wants to jump in and ride the waves of ARM. If they have a good execution plan, they can make it work.
Designing their own cores does make sense for them - otherwise they do not have any differentiation. I doubt the one-size-fits-all cores from ARM can address all the markets. I would suspect that ARM's own cores would likely want to retain dominance in the mobile space and are targetted to those market first. This is not to say that they cannot be used in servers but a dedicated server design would compete better.
Based on the news coming out of the ARM camp, it seems to validate this - almost all of them - AppliedMicro, Cavium, Qualcomm, nVidia are indicating they have their own core designs.
AMD does have decades of experience designing CPUs so that is their strength - the question is that of execution and focus when trying to ride two horses at once.
AMD is unique in that in can do both x86 and AMD, but as analyst Nathan Brookwood pointed out to me over lunch today, there's no great advantage to that.
Also, Nathan said he thinks there's no great advantage to designing your own custom core either given ARM is doing a pretty good and timely job covering the waterfront with its own designs.
Nvidia Tegra folks say Qualcomm, for example, went to a lot of trouble to design its own ARM cores but is not getting a huge technical advantage with them.