May 10, 2012
Memory manufacturers are gearing up DDR4, the industry’s the next-generation DRAM standard. While new technology typically draws an air of excitement and possibility, the next iteration of DDR memory has received a mixed reaction from Ars Technica.
The Joint Electronic Devices Engineering Council (JEDEC) creates memory standards and is close to finalizing the DDR4 standard. Currently, DDR4 is a draft specification. This means any modules manufactured between now and the time DDR4 becomes an official spec, might not be compatible with their intended devices. But a number of DRAM vendors, including Micron, Samsung, and Hynix, are looking to ship the technology this year.
While manufacturing to an unofficial spec is not a typical practice, it’s not the first time vendors have chosen this path. Prior to the IEEE 802.11n standard, home networking companies were selling draft-n home routers and network adapters to end-users. However in this case, there are no devices that can utilize draft DDR4 memory modules.
While industry adoption of the new memory technology is probably a couple of years away, KitGuru suggested that manufacturers are pushing for DDR4 because of marketplace conditions, writing, “The memory market is still in an unstable state due to the near collapse of one of the biggest manufactures Elpida, and it is hoped that the early release of DDR4 memory will again [stabilize] the market.”
Last August, JEDEC announced a number of elements in the draft DDR4 spec. The new memory is set to double transfer rates while consuming 20 percent less energy than its DDR3 predecessor. Specifically, it will perform 3.2 billion transfers per second while consuming a maximum of 1.2 volts.
A main feature of DDR4 memory is its use of point-to-point connections between memory modules and the CPU controller. This is an advantage over shared memory channels that exist between individual modules and the CPU’s memory controller.
While the architecture helps memory performance, the process of manufacturing becomes more complex as more memory modules result in more point-to-point connections.
Earlier this week, Micron announced that it has developed a DDR4 chip ready for testing. Unfortunately, it falls short of the JEDEC’s draft spec, able to handle 2.4 billion transfers per second (so 8 million short). Micron and other vendors will have time to improve their process, as Intel has no plans to incorporate the new memory technology into its roadmap until 2014.
Takeaway
DDR4 carries the promise of better power efficiency and higher bandwidth than the current DDR3 technology. Memory vendors appear to view their practice of early manufacturing as a means to gain market dominance and sell higher-margin parts. The strategy is risky though and could result in incompatible hardware or rejection in the marketplace.
Full story at Ars Technica
In quieter times, sounding the bell of funding big science with big systems tends to resonate further than when ears are already burning with sour economic and national security news. For exascale's future, however, the time could be ripe to instill some sense of urgency....
Read more...
In a recent solicitation, the NSF laid out needs for furthering its scientific and engineering infrastructure with new tools to go beyond top performance, Having already delivered systems like Stampede and Blue Waters, they're turning an eye to solving data-intensive challenges. We spoke with the agency's Irene Qualters and Barry Schneider about..
Read more...
Large-scale, worldwide scientific initiatives rely on some cloud-based system to both coordinate efforts and manage computational efforts at peak times that cannot be contained within the combined in-house HPC resources. Last week at Google I/O, Brookhaven National Lab’s Sergey Panitkin discussed the role of the Google Compute Engine in providing computational support to ATLAS, a detector of high-energy particles at the Large Hadron Collider (LHC).
Read more...
05/10/2013 | Cleversafe, Cray, DDN, NetApp, & Panasas | From Wall Street to Hollywood, drug discovery to homeland security, companies and organizations of all sizes and stripes are coming face to face with the challenges – and opportunities – afforded by Big Data. Before anyone can utilize these extraordinary data repositories, however, they must first harness and manage their data stores, and do so utilizing technologies that underscore affordability, security, and scalability.
04/15/2013 | Bull | “50% of HPC users say their largest jobs scale to 120 cores or less.” How about yours? Are your codes ready to take advantage of today’s and tomorrow’s ultra-parallel HPC systems? Download this White Paper by Analysts Intersect360 Research to see what Bull and Intel’s Center for Excellence in Parallel Programming can do for your codes.
In this demonstration of SGI DMF ZeroWatt disk solution, Dr. Eng Lim Goh, SGI CTO, discusses a function of SGI DMF software to reduce costs and power consumption in an exascale (Big Data) storage datacenter.
The Cray CS300-AC cluster supercomputer offers energy efficient, air-cooled design based on modular, industry-standard platforms featuring the latest processor and network technologies and a wide range of datacenter cooling requirements.