I love this paragraph -- not every decision was as considered as every other one, and they aren't going to pretend otherwise. And the last sentence is hilarious.
"One of the more ad hoc decisions we made was choosing the number of registers per register file. The problem we faced was a lack of directly relevant statistics on the effect of this parameter on performance in the context of a directed-dataflow style of execution. So we plucked the decision out of thin air. Our collective intuition told us that 32 registers per register file was too few, and 128 registers looked difficult from an implementation viewpoint. Thus we settled on 64 registers per register file. Comforted by the lack of alternatives, we moved on."
The late 80s/early 90s was such an interesting time for supercomputer architectures. I don't remember ever hearing about Cydra. But there were lots of niche ones -- Kendall Square Research had the ALLCACHE architecture, there was the Connection Machine from Thinking Machines, and the Tera MTA with lots of threads to hide memory latency. There are lots of others (anybody have an interesting or favorite one I'm forgetting?)
> The late 80s/early 90s was such an interesting time for supercomputer architectures. I don't remember ever hearing about Cydra. But there were lots of niche ones -- Kendall Square Research had the ALLCACHE architecture, there was the Connection Machine from Thinking Machines, and the Tera MTA with lots of threads to hide memory latency. There are lots of others (anybody have an interesting or favorite one I'm forgetting?)
Multiflow's [0] Trace VLIWs were pretty impressive. They had three models - 7-wide, 14-wide, and 28-wide. Stardent's Titan used MIPS CPUs and their own vector processors. The book The Architecture of Supercomputers: Titan, a Case Study has an excellent description of how their systems worked.
Depending on your definition of a super...Meiko Surface, Multiflow Trace, Sequent Balance & Symmetry, CDC Cyber 200-series (and 180/990) that evolved into the ETA10, all the weird/wonderful things Cray did during this time (like the Y/MP), Intel Paragon, NEC SX (still being produced). Lots more I've forgotten.
If this hardware were still around today, NetBSD would be running on it. The m68020 can still run NetBSD quite well, and someone would've written support for the Cydra numerical processor.
The actual title is The Cydra 5 departmental supercomputer: design philosophies, decisions, and trade-offs. I had to shrink it to fit.
If you're familiar with Itanium, then the design will look familiar. HP hired some of the Cydra designers.