Samsung's 1Gb Mobile DRAM

Today Samsung Electronics Co. announced that it has developed the industry’s first one gigabit (Gb) Mobile DRAM (dynamic random access memory) for mobile products, using 80nm process technology. The new chip is also called low-power DDR (double data rate) or synchronous DRAM. It will be more cost effective than other high density mobile solutions and can be used for a wide range of advanced handset applications as well as for digital still cameras, portable media players and portable gaming products.

The power consumption of this new chip drops to about 30% as compared to the double-die stack, 1Gb memory solution widely used today. This monolithic 1Gb Mobile DRAM chip uses the same packaging technique as the 512Mb double-die stack 1Gb package, however it introduces a new temperature-sensing feature. This new temperature-compensated, self-refresh feature maximizes the self-refresh cycle to reduce power drain in standby mode by about 30% over conventional memory chip designs.

Also, the new chip offers a more compact form factor. It is at least 20% thinner than a multi-stack package of 512Mb dies, allowing a single high-density package solution of 1.5Gb or even 2Gb Mobile DRAM memory, for which market demand is expected to grow in 2007.

Samsung Electronics said it plans to mass produce the new chip starting from the second quarter of 2007 when demand for high-density 1Gb mobile DRAM is expected to pick up.

Transistor Material Beyond Silicon

Jesus del Alamo

Each of us has several billion transistors working on our behalf every day in our phone, laptop, iPod, car, kitchen and more. Estimates that float around in the semiconductor industry circle state that within the next 10 to 15 years we will reach the limit, in terms of size and performance, of the silicon transistors key to the industry. As a result, both academic and industrial laboratories around the world are working on new materials and technologies that may be able to reach beyond the limits of silicon. Scientists are looking at new semiconductor materials for transistors that will continue to improve in performance, while devices get smaller and smaller.

Here steps in Jesus del Alamo, an MIT professor of electrical engineering and computer science and member of MIT's Microsystems Technology Laboratories (MTL). One such material del Alamo and his students at the MTL are investigating is a family of semiconductors known as III-V compound semiconductors. Unlike silicon, these are composite materials. A particularly hot prospect is indium gallium arsenide (InGaAs), a material in which electrons travel many times faster than in silicon. As a result, it should be possible to make very small transistors that can switch and process information very quickly.

Del Alamo's group recently demonstrated this by fabricating InGaAs transistors that can carry 2.5 times more current than state-of-the-art silicon devices. More current is the key to faster operation. In addition, each InGaAs transistor is only 60 nanometers (billionths of a meter), long. That's similar to the most advanced 65-nanometer silicon technology available in the world today.

Del Alamo notes, however, that InGaAs transistor technology is still in its infancy. Some of the challenges include manufacturing transistors in large quantities, because InGaAs is more prone to breakage than silicon. But del Alamo expects prototype InGaAs microdevices at the required dimensions to be developed over the next two years and the technology to take off in a decade or so.

The work was presented at the IEEE International Electron Devices Meeting Dec. 11-13 by Dae-Hyun Kim, a postdoctoral associate in the laboratory of Jesus del Alamo.

New Record for Network Data Transfer

An international team of Physicists, computer scientists, and network engineers joined forces to set new records for sustained data transfer between storage systems during the SuperComputing 2006 (SC06) Bandwidth Challenge (BWC).

The high-energy physics (HEP) team is led by the California Institute of Technology, CERN, and the University of Michigan and partners at the University of Florida and Vanderbilt, as well as participants from Brazil (Rio de Janeiro State University, UERJ, and the State Universities of São Paulo, USP and UNESP) and Korea (Kyungpook National University, KISTI). The HEP team's demonstration of "High Speed Data Gathering, Distribution and Analysis for Physics Discoveries at the Large Hadron Collider" achieved a peak throughput of 17.77 gigabits per second (Gbps) between clusters of servers at the show floor and at Caltech. Following the rules set for the SC06 Bandwidth Challenge, the team used a single 10-Gbps link provided by National Lambda Rail ( www.nlr.net ) that carried data in both directions.

Overall, this year's demonstration, following the team's record memory-to-memory transfer rate of 151 Gbps using 22 10-Gbps links last year at SuperComputing 2005 (read our past posting), represents a major milestone in providing practical, widely deployable applications. These applications exploit advances in state-of-the-art TCP-based data transport, servers (Intel Woodcrest-based systems) and the Linux kernel over the last 12 months. FDT also represents a clear advance in basic data transport capability over wide-area networks compared to last year, in that 20 Gbps could be sustained in a few streams memory-to-memory over long distances very stably for many hours, using a single 10-Gigabit Ethernet link very close to full capacity in both directions.

Further information about the demonstration may be found at:
http://monalisa.caltech.edu
http://monalisa.cern.ch/FDT
http://supercomputing.caltech.edu/
http://pcbunn.cacr.caltech.edu/sc2006/
https://scinet.supercomp.org/2006/bwc/

'Phase Change' Memory

The test set-up for 'phase-change' memory (photo courtsey: IBM)

Scientists from IBM, Macronix, and Qimonda achieved a major breakthrough in developing a new type of computer memory that could well be the successor to flash memory. Their joint work at IBM Research Labs succeeded in designing, building, and demonstrating a prototype "Phase-change" memory device that is able to switch over 500 times faster than 'flash' memory, while using less than half the power required to write data to a cell. Since no electrical power is required to maintain either phase of the material, "Phase-change" memory is non-volatile by nature.

The new memory material is a complex semiconductor alloy with germanium antimony semiconductor base doped with other, unspecified elements -- designed with the help of mathematical simulations specifically used in phase-change memory cells. The prototype memory device measures a miniscule 3 by 20 nanometers, which is far smaller than the size of any flash memory on the market today.

Many in industry expect flash memory to encounter significant scaling limitations in the near future (Also, read our past posting on Samsung's 40nm flash memory). Thus phase-change memories have the clear potential to play an important role in future memory systems.

The technical details of the joint research will be presented on Dec 13 at the Institute of Electronics and Electrical Engineers (IEEE) 2006 and International Electron Devices Meeting (IEDM) in San Francisco.