Rambus expands cryogenic memory research with Microsoft
[illustration; Rambus website]The technologies being developed by the companies will improve energy efficiency for DRAM and logic operation at cryogenic temperatures, defined by the U.S. National Institute of Standards and Technology as below −180C or 93.15 K and appropriate for high-performance supercomputers and quantum computers. It will enable high-speed SerDes links to operate efficiently in cryogenic and superconducting domains and allow new memory systems to function at these temperatures. Following the initial collaboration announced in December 2015, this new agreement extends joint efforts to enhance memory capabilities, reduce energy consumption and improve overall system performance.
“With the increasing challenges in conventional approaches to improving memory capacity and power efficiency, our early research indicates that a significant change in the operating temperature of DRAM using cryogenic techniques may become essential in future memory systems,” said Dr. Gary Bronner, vice president of Rambus Labs. “Our strategic partnership with Microsoft has enabled us to identify new architectural models as we strive to develop systems utilising cryogenic memory.”
Rambus has added further information in an on-line post, giving more insight into its cryogenic memory collaboration with Microsoft;
The memory systems being researched are candidates for applications in, for example, future generation data centres. Following Moore’s Law, memory systems have achieved exponential improvements in energy efficiency, density and per-bit cost for decades. These gains have facilitated the rapid growth in centralized, or cloud, computing. In recent years, the scaling of such metrics through conventional techniques has slowed. Concurrently, the demand for larger, faster data systems have increased due to the proliferation of ‘Big Data’ applications such as data analytics and machine learning.
These developments have prompted the industry to seek step function changes in performance and cryogenic computing and/or quantum computing – potential breakthrough solutions that could lead to a new era in computing. The potential improvements in cycle time, power consumption and higher density of computation per unit volume are all requirements for the most demanding applications and are all potential outcomes of this cryogenic research.
Overcoming scaling limits with cryogenic technology
The motivation for Rambus’ collaboration with Microsoft is to improve the energy efficiency and cost of ownership (COE) of memory systems in the data centre, particularly by operating them at very low temperatures. This exploratory work seeks to determine if there are sufficient energy saving opportunities or other advantages for memory systems.
While energy consumption is a primary area of emphasis, bit density scaling, performance, cost per bit and manufacturability may also benefit from reduced temperature and are being investigated. This would create an environment for potential computation speed increases at reduced power consumption.
Currently, there are multiple public and private sector research projects around cryogenic computing as well as quantum computing. These efforts show high speed processes capable of manipulating large amounts of data, which creates a multiple order of magnitude gap in the speed at which data can be sent or received from that process.
There is also a temperature gap between room temperature operation of current supercomputers (approximately 300K) and the operating temperature of a cryogenic core (4K). Rambus is seeking to close these gaps by designing and developing optimized memory sub-system solutions, capable of operating at 77K and interfacing to computers operating at liquid helium temperatures (4K).
There has been interest in low temperature computing for several decades, with early experimentation and results performed in the 1990s by Rambus’ own Gary Bronner when he was working at IBM (“A 4-Mb Low-Temperature DRAM”, published in the IEEE Journal of Solid-State Circuits, Vol 26, Issue 11, Nov 1991).
This paper concluded that, “The low-temperature performance leverage demonstrated for this DRAM is similar to the leverage possible in low-temperature logic; hence there would be no memory bottleneck in a fully low-temperature system.”
The original interest in low temperature computing was based on concerns that the technology and manufacturing techniques of the time would not be able to keep up with future performance requirements of the industry. While such work was ahead of its time in the 1990s, the end of traditional Moore’s Law scaling is now in sight and the time may be right for cryogenic computers and memory systems.
Developments in data centre computing including data analytics, deep learning, self-driving vehicles and other emerging technologies continue to push the boundaries of development and Moore’s Law [Rambus comments] s simply not keeping pace. These challenges are creating opportunities to meet technical needs in non-traditional ways and the industry is seeking a disruptive opportunity to surpass the slowing of Moore’s Law. As Doug Carmean of Microsoft discussed at the 2016 ISCA conference, quantum computing and other cryogenic computing concepts allow for a blank slate redesign opportunity.
Rambus; www.rambus.com/emerging-solutions/cryogenic-memory
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :