[CPlatypus]

X86 is not going away, I agree, but Intel can hardly exercise the kind of dominance they've enjoyed for the last several years when they're facing serious threats at both the low and high end. At the low end, ARM simply beats x86 for anything with a battery. Intel has already lost the phone and tablet markets, and laptops are highly likely to follow.

At the high end, look at the http://top500.org/. #1 is based on SPARC VIIIfx. #10 is based on the PowerXCell 8i. #2 and #4 both derive much of their power from GPUs. Even that understates the situation, because many of the most powerful computers next year - Blue Waters, Mira, Sequoia - will also be based on non-x86 architectures. Then look at what Tilera or Adapteva are doing with many-core, what Convey is doing with FPGAs, what everyone is doing with GPUs. Intel is going to be a minority in the top ten soon, and what happens in HPC tends to filter down to servers.

So Intel has already lost mobile and HPC. Even if Intel keeps all of the desktop market, what percentage of the laptop and server markets could they afford to lose before they follow AMD? Maybe it will happen, maybe it won't, but anybody who can see beyond the "Windows and its imitators" segment of the market would recognize that as a realistic possibility.

[chancho]

> what happens in HPC tends to filter down to servers

Is this conventional wisdom? How does a petaflop race affect app servers and databases? It seems like most traditional server workloads could get by without a single FPU. The only thing they have in common is IO. Are there many data centers using Infiniband? (Maybe there are I don't know.)

The Cell architecture is an evolutionary dead end. SPARC is no more of a threat to X86 now than before. GPUs may be the next big thing for HPC but its got a long way to go to get out of its niche in the server market. (That niche being... face detection for photo sharing sites? Black-Scholes? Help me out here.)

I mean, I agree with your overall point, but I think it's more likely that ARM will steal all the data center work before anything from the HPC world does. They are too focused on LINPACK.

[CPlatypus]

Are there many data centers using IB? Yes. SMP was common in HPC before it came down-market, likewise NUMA. Commodity processors have many features - vector instructions, specilative execution, SMT - first found in HPC. Power and cooling design at places liks Google and Facebook is heavily HPC-influenced as well. Certainly some things go the other way - e.g. Linux - but usually today's server design looks like last year's HPC design.

I'm not quite sure it's valid to write off SPARC as an architectural dead end when the current fastest computer in the world uses it, and the next crop of US competitors for that crown are all based on the Cell/BlueGene lineage. GPUs are also more broadly applicable than you might think. Besides video and audio processing, they can be used for many crypto-related tasks (witness their popularity for Bitcoin mining), various kinds of math relevant to data storage (e.g erasure codes or hashes for dedup), and so on. Many of their architectural features are also being copied by more general-purpose processors as core counts increase, as well.

Yes, high-end HPC is too obsessed with LINPACK. Nonetheless, it remains a good place to look when trying to predict the future of commodity servers. Even if ARM does displace x86 instead, many features besides the ISA are likely to come from HPC. Perhaps more relevantly, either outcome is still very bad for Intel.

[InclinedPlane]

Who cares what the top supercomputers are powered by, what matters is the market. I'm fairly certain that numerical simulation systems do not dominate the high-end computer market. Instead, it's still all about servers that spend most of their time shuffling data around, which intel platforms still do quite well at.

Additionally, intel still has plenty of time to get up to pace in the mobile market. The tablet market is as yet largely untapped, especially globally. I wouldn't be surprised if next gen atom processors made their way into leading edge tablets in the next few years, for example.

Generally speaking: forecasts that require intel to roll over and take a massive beating while billions upon billions of business leaks away to its competitors don't tend to pan out in reality. The only way that works is if intel goes bankrupt the instant a competitor comes on the scene, and that's just fantasy.

[ippisl]

Assuming tablets will run android , i would be surprised if intel made much money from atom processors on the tablet, considering the competition.

[InclinedPlane]

I don't see it as a given that x86 tablets would have to run android. There are some roadblocks to running iOS, for example, but none that are insurmountable.

[FooBarWidget]

> At the low end, ARM simply beats x86 for anything with a battery.

I've seen this claim often, yet I could not find any sources that could back up this claim. Can you post a link to an article or some research that compares performance/watt (as well as actual power usage) between ARM and x86? I'm genuinely interested in this.