ARM or Atom Based?
On HP's recent earnings conference call, CEO Meg Whitman said:
Later in the second quarter, we will be bringing the latest innovation from HP Labs and our Enterprise group to market, the first, commercialized product from our Project Moonshot. We expect this to truly revolutionize the economics of the data center with an entirely new category of server that consumes up to 89% less energy, 94% less space and 63% less costs than our traditional x86 server environment. This is exactly the technological inflection that can fuel the exponential growth of hyper scale computing. To put in that perspective, if just 10 large web services providers switched their traditional x86 servers to Moonshot, they could save a combined $120 million in energy operating expense and nearly one million metric tons of CO2 per year. The equivalent of taking over 180,000 cars off the road for a year. That is a game changer.
INTC now. ARM coming H2/13.
The first HP ProLiant Moonshot server is available with the Intel® Atom S1200 processor and supports web-hosting workloads. HP Moonshot 1500, a 4.3u server enclosure, is fully equipped with 45 Intel-based servers, one network switch and supporting components.
A growing ecosystem of industry leading technology partners
HP also announced a comprehensive roadmap of workload-optimized HP ProLiant Moonshot servers incorporating processors from a broad ecosystem of HP partners including AMD, AppliedMicro, Calxeda, Intel and Texas Instruments Incorporated.
Scheduled to be released in the second half of 2013, the new HP ProLiant Moonshot servers will support emerging web, cloud and massive scale environments, as well as analytics and telecommunications. Future servers will be delivered for big data, high-performance computing, gaming, financial services, genomics, facial recognition, video analysis and other applications.
Another #$%$ to turn power down - I noticed NVDA did not put a release date on "Volta".
Neither AMD nor Intel are part of the memory cube consortium - they probably try to move it faster
Also Intels Enterprise SSD fit right in - maybe they can bundle it.
During the final keynote of IDF, Intel's Justin Rattner demonstrated a new stacked DRAM technology called the Hybrid Memory Cube (HMC). The need is clear: if CPU performance is to continue to scale, there can't be any bottlenecks preventing that scaling from happening. Memory bandwidth has always been a bottleneck we've been worried about as an industry. Ten years ago the worry was that parallel DRAM interfaces wouldn't be able to cut it. Thankfully through tons of innovation we're able to put down 128-bit wide DRAM paths on mainstream motherboards and use some very high speed memories attached to it. What many thought couldn't be done became commonplace and affordable. The question is where do we go from there? DRAM frequencies won't scale forever and continually widening buses isn't exactly feasible.
Intel and Micron came up with an idea. Take a DRAM stack and mate it with a logic process (think CPU process, not DRAM fabs) layer for buffering and routing and you can deliver a very high bandwidth, low power DRAM. The buffer layer is actually key here because it helps solve the problem of routing pins to multiple DRAM die. By using a more advanced logic process it's likely that the problem of routing all of that data is made easier. It's this stacked DRAM + logic that's called the Hybrid Memory Cube.
The prototype these two companies developed is good for data rates of up to 1 terabit per second of bandwidth. Intel claims that the technology can deliver bandwidth at 7x the power efficiency of the most efficient DDR3 available today.
The big concern here is obviously manufacturing and by extension, cost. But as with all technologies in this industry, if there's a need, they'll find a way