In chips, four big players are each on very different paths to success in the market. Gone are the days when Intel led and everyone else struggled to follow. Now each of the major competitors is on its own unique path. It isn’t just x86 and ARM either. IBM is making a significant play to put Power back on the map and there are even some outlying technologies like HP’s Memristor on the horizon (though, to be fair, they’ve been on the horizon for the better part of a decade).
Let’s take a snapshot of the processor wars by the numbers.
AMD stopped chasing Intel a couple of years ago and has been working to differentiate itself with semi-custom efforts in the x86 and eventually the ARM space. This has allowed the company to basically own game consoles profitably, a segment that had been divided across a number of technologies before. Potentially, this allows AMD to approach opportunities that others can’t, and often don’t want to, touch.
This is actually pretty interesting because it is more skills-based than technology-based and since the value is in the skills, it is more like a service than a product. That makes it ideal for a smaller company and, assuming it can hold on to the related unique skill set (designing parts with clients), it gives AMD a far more defensible position than it previously had. Currently, much of its work is on x86 platform products but it’ll expand into ARM over time. Given that the leadership came from IBM, there is a reasonable chance that it’ll even pick up Power, giving it far more technology spread and the possibility of solutions that span all three technologies. Adding to this is AMD’s graphics capability and the only blended CPU/GPU solution, its APU, currently in market.
Each processor has its unique strengths and in automotive, manufacturing floor, or aerospace, where a blend of load types is common, it could come up with an optimized solution that others can’t touch.
AMD continues to compete in the graphics card segment and recently showcased a powerful advantage with monitors uniquely tied to AMD called FreeSync, which is far less expensive than competitive equivalents.
IBM is also taking a very interesting path. Internally, it had competition between its x86 group and its far larger and more lucrative Power-based solutions. IBM sold off its x86 business to Lenovo and then used the model similar to the one that created Linux, the OpenPOWER Foundation, to give Power momentum and to position it externally against x86 and ARM for servers. This is a very creative move, trading off control for potentially greater market share. It even has the potential, admittedly remote, of turning the other chip companies we are looking at into partners in the effort to expand Power. Though, particularly from Intel’s perspective, the words “cold day in hell” come to mind for that outcome.
This is the corporate strategic equivalent to the saying “if you love something, let it go.” The end result is that it now has momentum outside IBM on Power, it is a contender in the broad server space, and it was able to demonstrate a broad range of new products at the OpenPOWER summit this last month. This event, which was done in conjunction with NVIDIA’s GTC (GPU Technology Conference), also suggested a stronger partnership with NVIDIA over time and some rather unique solutions that are under development.
Intel is really the only company in this group still on a traditional course. Given that it still dominates most of the segments it participates in, the fondness for a model it created is well founded. It was in a war with ARM and the big battleground was tablets. ARM had made inroads into Windows with Surface RT tablets and it looked like Apple was going to move to ARM when it brought out a future version of the MacBook in a near-tablet form with similar battery life attributes.
But Windows RT has been a failure, Microsoft appears to be back on x86 for tablets, and there are rumors of a Windows phone based on x86. In addition, the new Apple tablet-like MacBook is based on the Intel Core M processor, not ARM, suggesting that on tablets the momentum is switching to x86. And while Intel’s design wins on phones are still minor, they are coming with greater efficiency, suggesting the strategy still has very strong legs. Granted, a lot of the future momentum would come from Windows 10 success with tablets and phones, but Microsoft is looking stronger (executing far better), and the Intel/Microsoft relationship seems stronger than it has been for years.
Intel has been wrapping its solutions with some rather unique capabilities, like the RealSense camera and Thunderbolt port to enhance its offerings as well. In all, Intel has probably never been stronger, but then again, its market has never been more competitive.
Qualcomm’s massive dominance on smartphones continues largely unabated, but that is not because it leads on processors or even graphics. It is because it has a massive lead on radios, which remain Qualcomm’s sustained competitive edge. It is accelerating this edge with efforts like LTE-U, which provides Wi-Fi-like capabilities that better align with carriers, MU-MIMO, which creates Wi-Fi performance in line with gigabit wired Ethernet, and its continued drive to the next generation of LTE. It has also developed a biometrics technology, Snapdragon Sense, based on sonic waves, which tests as superior to anything else currently in market.
In short, Qualcomm’s strategy and competitive advantage is similar to Intel’s, but with radios. When the two companies bump heads, if the key differentiator is performance, Intel wins. If it is connectivity, Qualcomm wins, which is why Intel has held Qualcomm off on PCs and Qualcomm has held Intel off on smartphones, with tablets being the current battlefield.
Wrapping Up: A Very Diverse Battlefield That NVIDIA May Have Left
AMD is largely taking itself out of the fight with Intel over time as it focuses on semi-custom and no one else seems all that interested, yet, in this space. IBM is going to war with Intel in servers but focusing on getting allies because it knows it can’t go it alone. Intel is fighting the most battles but from the position of dominance in its traditional markets and it is now making progress in mobile. And Qualcomm is using a similar strategy, using its dominance in radios to fight successfully; where radios are most important, it dominates.
So what about NVIDIA? Well, with the Shield effort, NVIDIA is slowly becoming more like Apple and Samsung, making a move as a vertically integrated vendor with far more unique technology experience down to a chip level than any other product company has. It is becoming far less a chip company and far more a solutions company. Increasingly, we’ll be comparing it to a completely different class of company. On the server side, its Grid effort largely stands alone and, coupled with clients running NVIDIA technology, could eventually provide the biggest end-to-end disruption this market has ever seen.
This all makes for an incredibly dynamic market unlike any we have ever seen before.
Rob Enderle is President and Principal Analyst of the Enderle Group, a forward-looking emerging technology advisory firm. With over 30 years’ experience in emerging technologies, he has provided regional and global companies with guidance in how to better target customer needs; create new business opportunities; anticipate technology changes; select vendors and products; and present their products in the best possible light. Rob covers the technology industry broadly. Before founding the Enderle Group, Rob was the Senior Research Fellow for Forrester Research and the Giga Information Group, and held senior positions at IBM and ROLM. Follow Rob on Twitter @enderle, on Facebook and on Google+