[erlang-questions] FPGA coming around the corner

[Ulf Wiger]

On 9 Jan 2011, at 04:23, Jesper Louis Andersen wrote:

> I have this notion that I call the computational gaps. Suppose a
> sequential C program is 10 times faster than a sequential Erlang
> program. Now, assume near-perfect speedup is possible for the program
> (it is embarrassingly parallel). The gap is then 10, as if we had 10
> cores, we could run as fast as the C program. If we had 100 cores, we
> would be 10 times faster. If we had 1000 cores, 100 times faster and
> so on.

Of course, if the problem is embarrassingly parallel, it will not be
particularly hard to use something like OpenMP to get good speedups
for the C program as well, ;-), but I understand that that wasn't your point.

The thing that often gets lost in these discussions is that for sufficiently
"interesting" applications, the things that matter most for performance 
are often the things we must do to achieve stability over time, as 
requirements evolve; to achieve robustness and debuggability; have
a fighting chance to stay on top of memory management and 
synchronization issues, etc.

I realize YMMV (greatly). The trick is to recognize what your main 
challenges will be and pick a technology and strategy that will serve
you all the way (I was going to say "to the finish line", but "through the
product lifecycle" is probably more accurate).

BR,
Ulf W

Ulf Wiger, CTO, Erlang Solutions, Ltd.
http://erlang-solutions.com