Date   : Wed, 14 Dec 2011 18:59:57 +0100
From   : rick@... (Rick Murray)
Subject: 32016 + 32082

On 13/12/2011 21:38, Simon Marchese wrote:

> http://acorn.chriswhy.co.uk/docs/Mags/PC/PC_Oct85_32016.pdf

Apart from one deplorable result, I find it interesting that, on the 
whole, the 6502 co-processor holds its own fairly well.

         32016   6502

BM1       0.43    0.42   <-- it's actually *faster*!?
BM2       2.32    2.10   <-- ditto :-)
BM3       4.14    5.49
BM4       4.82    5.98
BM5       5.96    6.96
BM6       9.80   10.30
BM7      15.07   15.54
BM8      12.51   35.57   <-- O.M.G.
Average   6.88   10.27

If we recalc a lazy average by adding then div by seven (i.e without the 
terrible result), it becomes:
Lazy av. 29.62   33.47

Which isn't bad when you consider that the expensive whoo-hoo 16 bit 
chip with 32 bit stuff inside and MMU/FPU capabilities and a 6MHz clock 
speed...
...is fighting a bog-standard 6502 clocking 3MHz. If we were to halve 
all of the 6502 results (for an imaginary 6MHz part), it would wipe the 
floor with this 32016. The only drawback to the 6502 is its 64K 
addressing space (as obviously two-regs-plus-acc versus heaps-o-regs is 
making less of a difference than you might expect).

This is, of course, assuming the listed benchmarks are actually useful. 
Unfortunately the one explaining the benchmarks hasn't been scanned in.


Best wishes,

Rick.

-- 
Rick Murray, eeePC901 & ADSL WiFI'd into it, all ETLAs!
BBC B: DNFS, 2 x 5.25" floppies, EPROM prog, Acorn TTX
E01S FileStore, A3000/A5000/RiscPC/various PCs/blahblah...