AMD’s Andrew Feldman Talks About ARM

Lower Development Costs Will Power AMD’s Comeback in Server Space

AMD - logoAMD’s general manager of their server business unit, Andrew Feldman, presented this morning at J.P. Morgan’s global technology, media and telecom conference. As the head of AMD’s server group Mr. Feldman talked almost exclusively about the ARM versus x86 debate, and how he thinks that the ARM server market is going to shape up over the next five years. AMD announced last year that it will be releasing an ARM based server chip in the middle of 2014, so this is a market that AMD is watching very carefully.

Right out of the gate Mr. Feldman laid down a few reasons why he believes that ARM server chips will take market share from traditional x86 offerings. According to him innovation is much faster paced in the ARM market, than in the x86 space, due to the large number of ARM licensees. In practice this means that ARM chips can adapt to new markets faster than their x86 counterparts. He also believes that the introduction of ARM based server chips will be a disruptive event for the server market. But more than that Mr. Feldman is betting on the server industry’s history of picking smaller, more mass-producible, options over larger and more complex offerings. This is a pattern that reared its head most recently when x86 chips displaced SPARC chips from the server market almost two decades ago.

Mr. Feldman expects that the first adopters of these new ARM based servers will be mega-data centers and cloud storage providers. But he believes that ARM chips will waterfall down from these applications into more traditional roles. Apparently the server market is no longer about having the highest single threaded performance, it’s now about using lots of little power efficient cores to solve simple problems, like search queries. Servers are changing to meet this new workload by adopting ARM chips, which are better suited for this role than traditional x86 servers.

The first credible 64-bit ARM based server chip will be coming from Applied Micro Circuits (NASDAQ: AMCC) near the end of this year according to Mr. Feldman. This chip will have a time-to-market advantage, but it’s being manufactured on the aging 40nm process which leaves the door open for more advanced offerings from AMD. By AMD’s count there are no credible offerings for vendors who want to use ARM chips in servers, but AMD expects the market for ARM server chips to really heat up in the middle of 2014 around the time period where it plans to introduce its first ARM based server chip.

For server workloads Mr. Feldman believes that ARM processors have an advantage over x86 chips in terms of power consumption, cost, and die space. AMD’s value proposition in the market is two-fold; first off it’s the only ARM vendor with experience in traditional servers and AMD’s off-the-shelf IP blocks will give them an advantage over the competition which may not be able to match IP portfolio that AMD has built for itself over the years. There are of course trade-offs in choosing one instruction set over the other, but AMD really sees the additive IP blocks that they can offer on their server chips as their biggest advantage over the other ARM licensees. To quote Mr. Feldman directly, “This isn’t religion, this is instruction sets.”

The discussion then turned to AMD’s prediction for the ARM based server market. AMD expects that these servers will gain acceptance and represent over twenty percent of the total server market by 2017. But AMD sees the one-size-fits-all mentality of chip building coming to an end due to the ease of ARM based chip development, “This [ARM based server chips] is the final stake in the heart of the one-size-fits-all world of just using different bins of the same chip.” With only a few exceptions the way Intel and AMD have catered to specific market segments is by binning different version of the same basic chip. This isn’t a bad solution per say but with the low development cost associated with ARM, AMD expects to be making more application specific offerings in the server market.

By Mr. Feldman’s count it takes 400 million dollars and three years to build an x86 server chip. But only 30 Million dollars and 18 months are required to make an ARM chip. As far as AMD is concerned the ARM ecosystem offers massive cost and time-to-market advantages over building traditional x86 chips.

When Mr. Feldman was questioned about competition from Intel’s Atom line-up he responded by asserting that, “Intel lost every handset on Earth because ARM was better than Atom.” He then tempered expectations by saying that Intel’s process technology does give them an advantage, but he also cautioned that Intel’s strategy of paying billions to maintain their fabrication advantage wouldn’t be enough to keep them afloat in the long-term.

Mr. Feldman was confident that AMD would lead the ARM server market because of AMD’s Freedom fabric interconnect and its unique IP blocks. He was even more confident that ARM based server chips would see market acceptance because of the interest that large OEMs and the top-tier of customers was showing in AMD’s upcoming products. According to him Dell and HP will be the two biggest ARM server vendors, but Lenovo and the rest of the OEMs will follow them into the market eventually.

AMD is investing heavily in ARM based servers. This is without a doubt a risky bet because of the unproven nature of the market. But at the same time AMD seems to be taking an aggressive approach to capture whatever pent-up demand there maybe for ARM based servers and out maneuver its more consumer oriented competitors. It’s going to take a while for us to see AMD’s efforts come to fruition, but at the moment they look to be well positioned to have a big impact on this growing market space.S|A

The following two tabs change content below.
Thomas Ryan is a freelance technology writer and photographer from Seattle, living in Austin. You can also find his work on SemiAccurate and PCWorld. He has a BA in Geography from the University of Washington with a minor in Urban Design and Planning and specializes in geospatial data science. If you have a hardware performance question or an interesting data set Thomas has you covered.