I was a bit surprised at the lack of interest in HP's Moonshot announcement by the big media. I suppose I understand the complexity combined with HP's less than stellar PR maneuvers over the last few years could keep some folks at bay, but this announcement has all the makings of a MASSIVE and exciting outcome.
If we start with the mainframe as the first commercial computing architecture to change the world—and the multiple trillions of dollars of value it has subsequently created—as the initial mega wave in "enterprise" processing, one can see what is possible when the stars align in this area.
Then the RISC processor and the mini-computer era—maybe not trillions created, but certainly many billions, and the stage was set for DISTRIBUTED processing—and business was changed yet again.
Fast forward to the x86 era and you witnessed it happen all over again. This time Intel and Microsoft were the primary beneficiaries, but a few thousand others did just fine as well—from Cisco to Dell and beyond. Trillions of dollars of value created. Lives changed. Business changed.
Now HP is looking to take ARM and Atom and upset the trillion dollar apple cart again. Initially taking this stuff to webscale environments, of which there are TRUCKLOADS now—but sooner or later, taking it right down main street. New players will emerge, as they always do, and old ones will be replaced or re-tooled (Intel makes the Atom chips—but unfortunately, they sell for pennies on the dollar compared to x86—and no, they can't just raise the pricing if the market takes off because Atom is how they play in mobile - and that's not something they can run from. Intel is boxed in, no matter what they say publicly.).
Atom is effectively x86 to apps—but it costs zip, consumes way less power, and has a small footprint—kind of what the world wants. ARM is even lower power, smaller, powerful, and CHEAPER. It runs Linux—so it erases your web servers, raid controllers, networking systems, etc. Anywhere there is a specialized embedded processor, HP can plop a blade theoretically, and collapse functions like never before.
So while they position for webscale—what they really mean is this packaging could be considered the ultimate converged infrastructure. Think of your iPad as a blade, that happens to have more power than NASA had globally only ten years ago—that can assume any personality you want—which plops into a filing cabinet that holds a zillion other blades—and runs on gerbil gasses (I think I saw this while touring HP labs). Oh yeah, they need no cooling either, comparatively.
So way smaller, way cheaper, way less power and cooling, far denser, morphable personalities...and they run the stuff I run today?
For x86 to take off, it needs APPS and an OS. No such requirement for Moonshot—they already exist.
For x86 to take off, it needed to convince the market that this was a good idea. No such requirement here. The fact is a ton of companies that rely on webscale for their livelihood will perish without SOMEONE delivering this type of capability—and soon. x86 is great—but was never designed for hyper-scale out that is the DNA of the service provider market, the SaaS market, and soon the EVERY market.
So SOMEONE is going to make a ton of dough doing this—because it's the inevitable and right thing to do. The market will see to it. HP seems to have a decent lead in this space, and decent brand to pull it off (funny how they didn't harp on the fact that they STILL dominate the x86 server market when it comes to density blades). This will start in bigger enterprises, but will inevitably make it to mobility led devices—certainly, your tablet and laptop if not your phone. Since HP still sells a PC or two, this would seem to bode well for them.
So what WILL be super interesting is that someone WILL pull this off—and the x86 market will evolve or devolve and the webscale processor market will supersede it. There will be tons of fallout and tons of new money made. New companies will rise from nothing to become massive success stories. It is inevitable. Because of HP? Not necessarily, but it sure can be a benefactor. It's because of MOBILITY. There are WAY more mobile devices that are driving the cost/tech/footprint/power thing than enterprise tech ever will—but just as Apple totally changed the BYOD game, Mobile WILL change the enterprise processor game. And the carnage will be as fascinating and interesting as the new generation of uber-companies who ride the wave. HP as the delivery vehicle for the enterprise market finds itself at the front of an awesome market upheaval—if it can hold on and ride out the turbulence, it seems to be in a perfect position to take advantage.