Chip doubles computer memory

IBM has developed technology that can double the memory of a computer and save companies up to $1 million (€1.05 million) on the cost of servers.

IBM's memory expansion technology works through a memory controller chip that acts as a link between the processor, or brain of the computer, and the memory.

The chip encodes the data so that they take up half of the normal space in the memory cache.

The encoding process, or compression, would normally slow down the speed of the computer. To avoid this the system automatically stores frequently accessed data and instructions close to the microprocessors.

Ms Ellen Yoffa, director of personal and scalable systems at IBM research, said the system was 10,000 times faster than software techniques for compression.

She said the chip improved performance of computers in tests by between 30 and 60 per cent adding that the technology was particularly applicable in space constrained environments.

The technology will be introduced in IBM's Netfinity line of servers with Intel processors for corporate networks and websites.

It compresses and retains less frequently accessed data and instructions in-memory, instead of on a disk, which increases the memory capacity of a computer or server by a factor of two or more according to IBM.

Users can cut costs by purchasing half the memory to achieve the same performance or increase performance by installing the same amount of memory to achieve twice the capacity.

The increased capacity technology could be used in everything from servers and PCs to mobile phones and hand-held computers.

Ms Yoffa said that IBM's goal was to have the technology pervasive throughout the industry within five years.

IBM said that for Internet service providers with a number of racks of servers, it would represent huge savings as they could double the memory of their systems at a saving of around $1 million a rack.