[rescue] sparc10 cpu - what to do.

[Dave McGuire]

On 12/16/2016 09:13 AM, john wrote:
> I cannot remember the name of the thing, but what was interesting was
> that a Sparc2 could process the image in about 10% of the time it took
> on a 486.

  Unsurprising; nearly every RISC processor trounced nearly every CISC
processor for nearly every type of workload in that era.  It's still the
case for some types of workloads.

> I think the math co-processor was resident in the 486 chip by
> then, but if it wasn't, our test machine had one.

  The 80486 is little more than an 80386 with an 80387 on-chip.  There
are other differences, but that's the main one.

> the difference was so astonishing to us that we invested in a lot of
> dialogue with the developer who was also astonished.  we supposed bad
> compiler fit with Intel, but developer who felt his future lay with
> intel, not sun, market was really focused on making it work with intel. 
> And at the time, this was the best they could do.

  It wasn't the compiler, it was the processor architecture.

                -Dave

-- 
Dave McGuire, AK4HZ
New Kensington, PA