March 2, 1998

Workstation war

I enjoyed reading the Hands-On evaluation "Has Wintel won the workstation war?" (EDN, Nov 6, 1997, pg 65). It was a timely look at the high-end hardware for Windows-based EDA computing. The article, however, seems to sidestep a key issue regarding EDA in the Wintel environment.

Wintel PCs do not scale like Unix workstations can. Windows 95 supports a maximum of one processor, and Windows NT supports a maximum of two processors. CPU processing power will max out with a pair of DEC's Alpha chips. Compare these limitations to Unix options. Unix workstations, running essentially the same software, can either use a simple architecture with one RISC processor or use arrays of 16, 32, or more processors on sophisticated backplanes. The ability of Unix workstations to scale over a wide performance range distinguishes PCs from workstations in a more significant way than the type of SCSI adapter used, the regulation of airflow through the chassis, or even the thickness of the motherboard.

It is true that many EDA applications are not multithreaded and thus cannot take full advantage of a multiprocessor environment. On the other hand, most engineering computing environments are networked. A workstation usually deals with TCP/IP traffic that can easily keep a processor busy. Further, many engineering tasks require multiple applications running simultaneously. Wright's article shows routing benchmarks of 11 to 30 minutes. What is the engineer supposed to do while this program is running, other than, of course, read EDN?

Andrew Mitz, PhD
Biomedical Engineer
National Institutes of Health Animal Center
Poolesville, MD

Corrections and updates

The Design Idea "Program converts temperature from RTD sensor" (EDN, June 5, 1997, pg 116) features a C program to convert platinum RTD resistance into temperature. It is incorrect. You should calculate RTD temperatures using the Callendar-Van Dusen equation. Over a limited temperature range, such as 0 to 100°C, you can use a linear approximation. In either case, the coefficients depend on the purity of the platinum used in the sensing element. National Instruments' Application Note 046 (available on www.natinst.com) is a tutorial on measuring temperature with RTDs. Also see Handbook of Transducers by Harry Norton, Prentice Hall, 1989.

Debora Grosse
EDN Contributing Editor

Sound off

Send your letters to Signals and Noise Editor, EDN, 275 Washington St, Newton, MA 02158 or
e-mail us at bmorrison@edn.cahners.com. Our fax is 1-617-558-4470. EDN reserves the right to edit letters for clarity and length.

| EDN Access | Feedback | Table of Contents |