Squeak on Android

Just as a point of interest - I downloaded the app onto my HTC Droid Incredible, and it seems to work fine (for what works).

Tiny Benchmark gives:

31,250,000 bytecodes/sec

857,762 sends/sec

The Incredible has a Qualcomm 1GHz Snapdragon processor...

http://www.htc.com/us/products/droid-incredible-verizon#tech-specs

- Jon

An iPhone 3GS with a bit of careful tuning will do 

Dan Ingalls wrote on Tue, 21 Dec 2010 20:16:13 -0800
> Awesome.  For some reason this really grabbed me -- This is about the
> speed we achieved on the Dorado - the fastest of Xerox's microcoded
> D-machines when we were first taking Smalltalk public.  That machine
> could not be carried by two people, let alone one, gobbled power and
> had to live in a separate room due to the fan noise.
> A phone...  yeah

The number I have often seen associated with the Dorado is 400,000
bytecodes/sec, which is 80 times slower than Jon's phone and 100 times
slower than John's tuned iPhone VM. I just use 0.5Mbc/sec for the Dorado
to make the math a bit simpler. Since the Dorado could execute 14
million microcode instructions per second, this would mean 28 microcode
instructions per bytecode which does seem rather high. But I really
can't imagine it needing less than 10 microcode instructions per
bytecode, which would result in a peak of 1.4Mbc/sec.

-- Jecel

Eliot Miranda wrote:

> While things like temp var and inst var access may be cheap sends
> can run very high (think of a cache miss lookup or a cache miss lookup
> followed by MNU processing), as can some primitives.  I would think
> 28 microcode ops per bytecode is a believable average.

I agree! And in the case of Squeak we have added stuff like exception
handling that makes sends even slower. But the code from which the
tinyBenchmark numbers are obtained is not too typical for Smalltalk (it
is a loop with not too many message sends other than array accesses and
SmallInteger math) so I was speculating that if tinyBenchmarks were run
on the Dorado we might get higher than expected numbers.

Though I think I have already mentioned this here, one thing I found
hard to understand when looking at the microcode for Smalltalk on the
Dorado was that it spent quite a few microinstructions decoding the
bytecodes. But all the texts about it mention the special hardware that
the machine had to do this and how important this IFU was to getting
Smalltalk to run well. On possibility is that the code was there but got
skipped when the IFU was activated. It would be interesting to know
under what conditions the number that were cited for the Dorado were
obtained.

-- Jecel