Economist.com | Articles by Subject | Computing
1 sur 3
http://www.economist.com/research/articlesBySubject/PrinterFriendly...
Articles by subject: Topics: Computer technology SCIENCE & TECHNOLOGY
Computing
The cell of a new machine Feb 10th 2005 From The Economist print edition
Is the new Cell chip really as revolutionary as its proponents claim? ANALOGIES are often drawn between the fields of computer science and biology. The information-processing abilities of DNA are a form of natural molecular computing, and computer viruses leap from machine to machine in ways reminiscent of their disease-causing namesakes. At the International Solid-State Circuits Conference in San Francisco this week, a trio of mighty information-technology firms—Sony, Toshiba and IBM—pushed the analogy a little further. They unveiled a much anticipated new computer chip, four years in the making, the very name of which is a biological metaphor: the Cell. As its name suggests, the Cell chip is designed to be used in large numbers to do things that today's computers, most of which are primitive machines akin to unicellular life-forms, cannot. Each Cell has as its “nucleus” a microprocessor based on IBM's POWER architecture. This is the family of chips found inside Apple's Power Mac G5 computers and IBM's powerful business machines. The Cell's “cytoplasm” consists of eight “synergistic processing elements”. These are independent processors that have a deliberately minimalist design in order, paradoxically, to maximise their performance. A program running on a Cell consists of small chunks, each of which contains both programming instructions and associated data. These chunks can be assigned by the nucleus to particular synergistic processors inside its own Cell or, if it is deemed faster to do so, sent to another Cell instead. Software chunks running on one Cell can talk to chunks running on other Cells, and all have access to a shared main memory. Since chunks of software are able to roam around looking for the best place to be processed, the performance of a Cell-based machine can be increased by adding more Cells, or by connecting several Cell-based machines 14/02/2005 19:09
Economist.com | Articles by Subject | Computing
2 sur 3
http://www.economist.com/research/articlesBySubject/PrinterFriendly...
together. All of this means that programs designed to run on Cell-based architecture should be able to fly along at blistering speeds—and will run ever faster as more Cells are made available. The prototype Cell being discussed this week runs at 256 gigaflops (a flop—one “floating-point” operation per second—is a measure of how fast a processor can perform the individual operations of digital arithmetic that all computing ultimately boils down to). A speed of 256 gigaflops is around ten times the performance of the chips found in the fastest desktop PCs today; the Cell is thus widely referred to as a “supercomputer on a chip”, which is an exaggeration, but not much of one. On the top500.org list of the world's fastest computers, the bottom-ranked machine has a performance of 851 gigaflops. A machine based on only four Cell chips would easily outrank this.
Hard cell If you believe the hype, all this has other chipmakers—notably Intel, whose Pentium series is the market leader—quaking in their boots. No doubt they are nervous. After all Andy Grove, one of Intel's founders, called his most famous book “Only the Paranoid Survive”. But it is not yet clear that Cell will sweep all before it. The reason is that existing programs have not been designed in the chunky way required if they are to run on Cell-based machines—and rewriting them would be a monumental task. For the moment, that will not worry Cell's designers because the kinds of things Cell chips are intended to be used for require specially designed software anyway. Cell chips are well suited to processing streams of video and sound, and for modelling the complex three-dimensional worlds of video games, so Cell's debut will be in Sony's next-generation games console, the PlayStation 3, which is expected to contain four of the beasts. Cell chips will also be ideal for use inside consumer-electronics devices such as digital video-recorders and high-definition televisions. Both Sony and Toshiba plan to use Cell chips in such products. For its part, IBM is talking up the Cell's potential to power supercomputers, the fastest of which, IBM's Blue Gene/L, consists of thousands of special chips that are, in many ways, more primitive versions of Cell. Using Cell chips instead would not, therefore, be a big stretch. And supercomputer programmers, like video-game designers, do not mind learning to program an entirely new machine provided it delivers the performance they are seeking. If Cell did eventually break out of these specialist applications and into general-purpose computers, Intel would have every right to be paranoid. But Kevin Krewell, the editor of Microprocessor Report, an industry journal, sounds a note of caution. The Cell is too power-hungry for handheld devices, and it would need to have its mathematical functions tweaked to be really suitable for use in supercomputers. The Cell is impressive, but, in Mr Krewell's view, “it is no panacea for all those market segments”. Similar claims to those now being made for Cell were made in the past about the Sony/Toshiba chip called the Emotion Engine, which lies at the heart of the PlayStation 2. This was also supposed to be suitable for non-gaming uses. Yet the idea went nowhere, and the company set up by Toshiba to promote other uses of the Emotion Engine was closed down. Extravagant claims were also made about the RISC, POWER and Transmeta architectures, notes Dean McCarron, a chip analyst at Mercury Research in Scottsdale, Arizona. These once-novel methods of chip design have done respectably in specialist applications, but have not dethroned Intel as it was suggested they might when they were launched. Yet both Mr McCarron and Mr Krewell acknowledge that things could be different this time. As Mr McCarron puts it, there are “more ingredients for success present than on previous attempts”. Intel was able to see off earlier pretenders to its throne by increasing the performance of its Pentium chips, and by exploiting its economies of scale as market leader. But in this case, the performance gap looks insuperable, and Sony, Toshiba and IBM plan to exploit economies of scale of their own. Quite how revolutionary the Cell chip will turn out to be, then, remains to be seen. And though it may not be an Intel killer, it could prevent that firm from extending its dominance of the
14/02/2005 19:09
Economist.com | Articles by Subject | Computing
3 sur 3
http://www.economist.com/research/articlesBySubject/PrinterFriendly...
desktop into the living room. Consumer-electronics devices, unlike desktop PCs, do not have to be compatible with existing software. In that sense, the Cell does pose a threat to Intel, which regards the “digital home” as a promising area for future growth. Stand by, therefore, for another round of creative destruction in the field of information technology. And no matter what the Cell does to the broader computer-industry landscape, the virtual game-vistas it will conjure up are certain to look fantastic.
Copyright © 2005 The Economist Newspaper and The Economist Group. All rights reserved.
14/02/2005 19:09
