15.5. Systolic Processing
Some of the earliest work in highly parallel systems was done by HT Kung and CE Leiserson at Carnegie Mellon University. As early as 1979, Kung had become an early advocate for a parallel-processing style he dubbed “systolic processing.” Kung and Leiserson were attempting to develop a general methodology for mapping high-level algorithmic computations onto processor arrays. They envisioned the processor array as resembling an automotive assembly line, with processors in the array working on the same data at different times and the entire array working on many data blocks simultaneously. A key element of the systolic-processing concept was the idea that data—in the form of partial results—could flow among the processors ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access