home .. forth .. misc mail list archive ..

Re: SPECint, Video, Marketing


I think it was Eugin who said...
>I think you are terribly overoptimistic here. This guys will be seeing
>bad float values and... Hold it, you just said SPECint?
Yes, maybe I was a little opptimistic here.  I _was_ focusing on integer
benchmarks because (1) F21 would be trounced in a floating point 
comparison, and (2) it seems that many applications could be written
using integer math, based on many comments here and elsewhere.  Of course,
most scientists and engineers only know or want to use floats.  I have 
not figured out how the F21 compares to Alpha on integer tasks.  Maybe Alpha
would kill it.  I am hoping that a ring of F21s has a chance.

>Scientific/engineering market is too small/conservative to be of
>any value and fun markets is big bucks/major players and the cake
>is already cut (and partly eaten), already. Anything I forgot?
I don't have any $$ figures on size of the scientific/engineering market,
but it must be pretty high profit.  Why are there so many vendors who
want to sell me-too Unix boxes?  

> [Video] _is_ a big market. If Jeff were not UltraTech but NEC or 
> Nintendo and had his product out by now, there wouldn't be any problems. 
> If he can persuade big players into buying the technology, we will see 
> F21s cropping >up everywhere.
Yes, this would be a huge market.  Unfortunately, I think it will get 
very bloody in the next few years.  It seems like *EVERY* major 
technology company, whether hardware or OS or software, wants to be in 
the set-top box market.  There will be lots of technologies offered.  
Most of them will fail in the competition of the market.  Maybe someone 
will appreciate MISC and sneak the F21 into a good product.  But it will 
face some really big opponants.  And I suppose that MISC seems very risky, 
especially to conservative Japanese companies.  If there was a standard 
for Interactive TV, for example, someone could try to implement it on F21 
and show how inexpensive it could be done with MISC.  I don't think Taos 
qualifies yet.  Maybe it will.  MISC video needs more color depth for 
many of these applications -- at least 10 bit pixels instead of 5.  OTOH,
Jeff plans to use 3 F21s to generate 18 bit color.  That should be 
interesting. :)

>> ...show them someone who is already doing such work.  
>Even if they would see it happen, it won't sell if they have to learn
>a new language/enviroment and switch to parallel programming first.
>No way.
Maybe so. :(  I hope SOME of you try it!  Here's a crazy idea...  What 
if someone with really good and cheap technology offered a private 
supercomputing center with programming and opperations services on the 
side to the scientific and engineering community.  It wouldn't matter if 
the technology was really weird (maspar MISC, MISC+DSP, etc.) because the
center would accept requirements on the algorithm level and code it for 
their weird machines.  Then they would run it for the scientist/engineer 
customer and trasmit the results back.  I suppose many scientists want 
to code themselves but I can imagine that just as many would rather not 
deal with the code if the center's data crunching was really cost 
effective compared to traditional approaches and the company knew how 
to work with scientists and engineers.  

I really hope there is some type of commercial success soon.  I'd like to 
see Chuck be really well financed by a company with product engineers and 
a good marketing department.  Chuck could focus on the research (getting 
accurate silicon models for deep-submicron, implementing on-chip SRAM, 
etc.), and the company could fully develop a product and sell it in the 
market.  A company like UltraTech could fit in the middle there somewhere, 
by porting an OS or development environment, for example.  Or do you have 
your sights set on a bigger target, Jeff?  :)
--
Michael A. Losh			Standard disclaimers apply.