A company creates a true made-by-monkeys debacle when it creates demand for a new product it can’t ship and at the same time kills supply and demand for a product it can ship.
By Charles Glorioso, Contributing Writer
While it is always satisfying to solve the engineering mystery, the solution doesn’t always result in a successful outcome. The time of this adventure is the transition from ISA bus to the faster EISA bus in PCs, 1988-1990.
At the time, I worked for a company that was one of the largest suppliers of graphics cards and related driver software for the PC. All our products were based on graphics chips designed by others.
The company was one of the leaders because we were able to ship PC cards in volume within 3 months after getting first details of a new graphics chip, and because we were able to focus all our marketing effort on the new product.
This combination of skills allowed us to be first to market with new capabilities, and to make those capabilities very visible to prospective customers. Since the average life of a PC graphics card was often 6 months, a month or two lead in ramping sales volume created a significant sales advantage.
The short product development times and short product life times were in conflict with normal manufacturing lead times for part procurement and volume ramp up. This resulted in compression of the normal timeline until things were happening in parallel.
The typical product ramp up looked something like this:
With the release of the EISA based graphics product steps 7 through 9 didn’t go as planned.
The headaches began at step 7. Engineering completed its prototype hardware and drivers and began testing. We soon discovered a problem. In some subtle way certain operations caused the PC to hang. The failure looked like software, so we looked hard at the drivers with every tool available.
We first identified the actual command that was causing the hang. Then we simplified the driver until it was obvious that the problem was not in the driver.
Then we used oscilloscopes and logic analyzers to study the EISA bus and the internals of the graphics PCBA circuitry. Finally, after many long nights we had worked our way back to the graphics chip as the likely source of the failure.
All the while we were doing our sleuthing, we were talking with the chip vendor about the problem we were having. During the first month of our efforts, the chip vendor assured us that the chip was fine.
During the second month, they acknowledged a flaw in the chip and promised a software work around. Finally, in the third month, they acknowledged that the flaw was fatal and it would take three months to “spin” the chip before we would have corrected parts.
Mystery solved, the chip implementation was defective. However, while we were struggling with this mystery, manufacturing had ramped down the previous product (step 5) and marketing had switched all advertising and effort to the new design (step 6).
The result was that the company had created demand for a product we couldn’t ship, and had killed both supply and demand for a product we had been shipping.
This turned out to be a fatal misstep for the company. The lack of sales put a deep hole in the company’s cash reserves and in its reputation.
The company never really recovered from either; suffering a loss of sales. After a series of layoffs, (which included me), the company was eventually absorbed by a competitor.
Contributing Writer Charles Glorioso has a BSEE from Purdue and an MSEE from Illinois Institute of Technology. He has over 40 years experience in electronics design and management for industrial and consumer products. Employers have included Teletype Corporation, Cadence, The Exploratorium, at least 6 companies which no longer exist. Charles retired earlier this year after six years as Director of Engineering at Davis Instruments. He is now working there part time on special projects.