Cell Chip Architecture Wont Sell Without Tools - Page 2

Unless the need for breakthrough development tools can be met, Cells ambitions may seem as far ahead of the mainstream in the 2020s as similar visions were in the 1970s—when science fiction writers started to talk in concrete terms about worldwide distributed computing. As the canonical example of that fictional literature, I recommend Thomas Ryans "The Adolescence of P-1," whose machine-intelligence title character conveniently evolved from an adaptive worm designed to seize storage while evading detection.

Today, when we say that a software system "wasnt designed—it just grew," thats not a term of praise. Enterprise buyers arent attracted to systems that work as well as a human mind but are at least as poorly specified in terms of exactly what theyll do in any given situation. Go ahead, try to survive a HIPAA or a Sarbanes-Oxley audit with an architecture diagram of a cloud labeled, "Here the miracle occurs."

Its important, then, to see Cells quantum jumps in processing power and collaborative computing matched by equal breakthroughs in software specification, verification and quality-of-service assurance. Without that achievement, Cells differences may not make an overwhelming difference.

The still-plummeting costs of more conventional processors, combined with the installed base of skills and tools to use them, pave the way for bit-based business as usual. I look forward, therefore, to seeing the tools that will be introduced to sell Cell.

Technology Editor Peter Coffee can be reached at peter_coffee@ziffdavis.com.


Check out eWEEK.coms for the latest news, views and analysis on servers, switches and networking protocols for the enterprise and small businesses.