dataflow

Chuck Guzis cclist at sydex.com
Tue Apr 11 20:46:32 CDT 2017


On 04/11/2017 06:24 PM, Jecel Assumpcao Jr. via cctalk wrote:

> 
> I was not aware that there had been any out of order implementations 
> after the IBM ACS until the second half of the 1990s. Given Cray's 
> passion for simplicity, I would not expect any of his designs to use 
> o-o-o (specially one as early as the CDC 6600).

Study the architecture sometime--instructions were *issued* in  order,
but sent to different functional units with differing execution times.
Since there were two increment units (which handled loads and stores),
the legendary "store out of order" problem had already showed its ugly
face--in 1964.

Really, nihil sub sole novum in much of this field.

--Chuck



More information about the cctalk mailing list