Re: MeshSP vs. P21

[Eugen Leitl <ui22204@xxxxxxxxxxxxxxxxxxxxxxx>]

Some comments on the last post of Cristophe Lavarenne: 


the MeshSP is the very antithesis of P21/F21. There are a lot
of problems merely limited by the memory bandwidth while having
no or very low data locality, while being highly parallelizable.

In fact, most algorithms can be written in this way, scientific
and real-world as well. Only a very tiny class of problems is
intrinsically sequential. These are nonparallizable.

Hence, for most problems, even a single F21 is no slower than 
a Pentium, an Alpha AXP or whatever, if adequately programmed. 
Even better: using mid-grain (1-2 MByte/node) multinode desktop 
machine we can have 1-2 orders of magnitude the performance of 
a big machine (a workstation) at the same hardware price.

This number-crunching business: provided the integers are quantized
fine enough (about 64-128 bit, 20 bit is somewhat tight) there is no 
need of a dedicated FPU. Integer ALUs are faster, skinnier, virtually 
glitch-free. Moreover, using scaled integer maths helps to avoid subtle 
float bugs, which are horrible to debug and often go undetected.


_Especially_ for scientific number crunching, a F21 cluster is great.

Of course, an off shelf Fortran compiler won't run on it.
Some adaptation on the side of programmer will be needed. 
Alas, most scientists are nonprogrammers and _very_ 
conservative. Bad luck.

-- Eugene