[gomp] Challenges in Implementing OpenMP

[Biagio Lucini]

On Thursday 21 October 2004 16.08, Robert Dewar wrote:
>
> Being opposed to abstraction on the grounds that it stands
> in the way of last nanosecond efficiency is an attitude
> that was more common in 1960 than now :-)

Not that so: I realise that HPC is a niche in the market, but if you spend 
some figure with 5 or more zeros for a machine, you really want to make the 
most of it (otherwise you needed not buy it right from the beginning). For 
instance, we still keep writing matrix multiplications in assembly :-(

Biagio

-- 
=========================================================

Biagio Lucini				      
Institut Fuer Theoretische Physik
ETH Hoenggerberg      
CH-8093 Zuerich - Switzerland           
Tel. +41 (0)1 6332562  
 
=========================================================