old hypervisor email

from long ago and far away.

To: wheeler

On our previous subject . . . yes, I think you should contact Dr. Bill Worley. He has 801, new interactive software and some low end processors. He may be a bit circumspect because of security, but he's probably the right person to start with.

Thanks alot for the description of Amdahl Corp's new machine. If you receive any other information I would like to receive a copy.

Amdahl's idea of "macrocode", at least the way you describe it, is a very good one. I say it's a good one because we've been pushing to try it for years. The idea is to recognize that S-370 buttmebler language could be subset, logically, and allow a performance improvement. FOr example, the machine could buttume the software had no bugs in it. Or the machine could buttume only "simple" instructions would be used (e.g. no overlapped moves, no MVCLs or CLCLs, no TRTs, etc.). Doing that would create another state of machine end. It also admits to another "implementation technology" in the machine (circuitry, microcode and BALcode).

An example would be the hypervisor function of CP-370. That should be "in the machine" but there's obviously no reason to rewrite the whole program into 4341 microcode. That one has been a dream of mine for quite some time.

... snip ...

obviously, email was before Worley left for hplabs, misc refs

there was project to replace the wide variety of corporate microprocessors with 801s, including the microcoded engines used in the low-end and mid-range 370s. misc. collected posts mentioning 801

misc. past posts mentioning macrocode:

eventually PR-SM was produced for the 3090 and has since evolved into the current LPARs (logical parbreastions). post related to SIE, sort of leading up to pr-sm

lots of past posts mentioning PR-SM and-or LPARs