Gordon Bell wins S. Cray Award

Paul Koning wrote: Prism is what Cutler renamed the "SAFE" (Streamlined Architecture for Fast Execution) RISC machine project after stealing it from Alan Kotok's Advanced Development group in Marlborough MA. The *MAJOR* change (at first) was that they renamed all the data types to full VAX compatibility/stupidity. ie; "16" became "word", 32 became "long" and 64 became "quad". I seem to recall some other changes: One of SAFE's registers was always zero, and Prism moved it to the other end of the register file. The wikipedia entry says Prism eventually shrank to a 32-bit machine. We were aware of Titan (a Unix workstation), and were aiming at 64-bits with VAX data compatibility, running VMS. I was in the working with Al Blackington (who had a low -- two digit? badge number) on "VAST" a VAX Assembly language translator, which took output from the VAX assembler after macro expansion, and eliminated all the unused CISC side effects. I was doing an instruction scheduler (based on work by Dave Orbits?) after reading the papers from Fisher's VLIW project at Yale that became Multiflow. Kotok had a very open management style and forwarded lots of emails, so we all got to see the train comming at us... I wasn't aware that Prism ever generated hardware, but I left DEC not longer afterwards.... Phil