On 11/07/15 06:02, Nick B wrote:
On Thursday, 20 February 2014 at 10:10:13 UTC, Nick B wrote:
Hi everyone.
John Gustafson Will be presenting a Keynote on Thursday 27th February
at 11:00 am
The abstract is here:
http://openparallel.com/multicore-world-2014/speakers/john-gustafson/
There is also a excellent background paper, (PDF - 64 pages) which can
be found here:
FYI
John Gustafson book is now out:
It can be found here:
http://www.amazon.com/End-Error-Computing-Chapman-Computational/dp/1482239868/ref=sr_1_1?s=books&ie=UTF8&qid=1436582956&sr=1-1&keywords=John+Gustafson&pebp=1436583212284&perid=093TDC82KFP9Y4S5PXPY
As far as I can tell, the author is pushing toward a hardware based
implementation (i.e. - have the CPU do the implementation). All of his
arguments are around "transistors are cheap, bus lines are expensive" theme.
A software based implementation will have much worse performance than an
FPU based floating point implementation, making these, even if generally
adopted (and I have my doubts), impractical.
If the powers that be think this is showing enough of a potential, then
it might be a good idea to shape D so that it CAN use unums. That would
require quite a few architectural changes to the way people think (which
is a large part of why I think unums will not be generally adopted).
Shachar
