Remembering Ross Freeman
In retrospect, it seems it took a long time for Xilinx Inc. co-founder Ross Freeman to be honored by the National Inventors Hall of Fame. Freeman’s induction into the Hall of Fame May 2 comes 19 years and seven months after his untimely death in October 1989.
But then again, Freeman’s rationale for designing a high-density programmable logic device from an SRAM cell seemed radically counter-intuitive in the mid-1980s. Xilinx Fellow Emeritus Bill Carter, who was employee #8 and the first IC device design specialist hired by Xilinx, said that only Freeman’s innovative idea, combined with former Xilinx CEO Bernie Vonderschmitt’s insistence that large wafer fabs were a drain on startups, allowed people to think beyond the confines of traditional ways of looking at semiconductor resources.
“Ross had hired me at Zilog, and I was as interested as working for the person, in his new startup, as I was in considering his ideas,” Carter said. “But after he described his FPGA concept to me, my first thought was, ‘That’s crazy. That is one of the most inefficient uses of transistors I’ve heard of.’ What Ross recognized is that the implications of Moore’s Law meant we didn’t really need to worry about using transistors inefficiently.”
Freeman is being recognized for his Patent No. 4,870,302, a programmable silicon structure of open gates that could be reprogrammed as often as necessary. Basing an FPGA on an SRAM cell would make the design volatile, Carter said, but Freeman realized that did not matter. Any design with a flip-flop could be initialized on power-up, so nonvolatility was not required. And the fact that individual programming elements were not based on the most efficient transistor allocation did not matter in an era of shrinking geometries and outsourced foundry services.
Freeman earned a BS degree in physics from Michigan State University in 1969 and a master’s from University of Illinois in 1971. He worked in the Peace Corps for several years, then went to Teletype to design a custom PMOS circuit. Freeman joined Zilog Inc. as one of its first engineers, and Freeman hired Carter at Zilog to work on the Z8000 program.
I interviewed Freeman three or four times in the mid-1980s, and found him to be one of the few engineers in a CTO position who was open and gregarious in talking to press, analysts, and investors. Carter said Freeman was the same way in talking openly with design teams at Xilinx – “and he was never a prima donna. He just always showed a quiet confidence.”
Although Freeman was suffering from chronic illness by the late 1980s, Carter said he remained driven and active until the time of his death. Carter said he still remembers “two earthquakes for Xilinx employees – the big Loma Prieta earthquake that rocked the whole Bay Area, and Ross’s death soon after, that shook up our company just as hard.”
It’s good to see Freeman be inducted alongside fellow inventors such as Intel Chairman Emeritus Gordon Moore. He’ll be remembered as the man who re-defined programmable logic.