Joe Fjelstad, President of Verdant Electronics, delivered the above message February 28, 2008, in a keynote presentation that helped open the Microelectronics Packaging and Test Engineering Council's (MEPTEC's), 4^th Annual Thermal Management Symposium at the Wyndham Hotel in San Jose, CA.

In 1943, the Electronic Numerical Integrator and Computer (ENIAC), with its 19-thousand vacuum tubes, dissipated a mere 200 kilowatts (kW). This was a beginning for the electronic computer industry that eventually proved the need for solid-state electronics. By the time the ENIAC ceased operations in 1955, the transistor had been invented.  In 1997, the Tsubame supercomputer--located at the Tokyo Institute of Technology--was based on advanced micro devices (AMD) and complementary metal oxide semiconductor (CMOS) microprocessor power. This system dissipated approximately 1.2 kW and performed 80 trillion floating-point operations per second. This period, with all its gains in computing power and reductions in thermal power, turned out to be an evolutionary succession of temporary solutions.  CMOS was the answer to bipolar heat dissipation in 1985. By 2000, CMOS was pushed to its thermal limits with 100-watt chips, resurrecting Bipolar cooling requirements.

Fjelstad suggested several suggested concerning thermal technology solutions for advanced products in his keynote address which included:

• Considering incipient thermal problems first, which requires the use of concurrent design;
• Multi-level successive phase overheat protection culminating in auto-shutdown, and;
• Optimizing material TC, shortening paths, cleaning up critical paths and eliminating parasites.

Cooling tools, such as liquids, gases and sprays, heat pipes and thermoelectric systems must also be considered. Thermal modeling identifies problems and measures solution affectivity--a preview of Rajit Chandra's presentation, which was next on the agenda. Finally, embedded devices are a unique, new contribution to thermal management. Such devices increase the ratio of package area to volume and shorten paths. To balance the program, MEPTEC should have highlighted a speaker to expand on the many benefits of embedded components.

Fjelstad was very circumspect in his keynote role--not even mentioning his Occam Process invention, which could possibly be the next major paradigm shift in thermal solution packaging.

Look for the continuation of this article which will address the many distortions and misconceptions surrounding the Occam Process being peddled by some--I'll provide links to their devious articles.