SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : Advanced Micro Devices - Moderated (AMD)
AMD 215.00+0.7%3:59 PM EST

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: muzosi who wrote (65082)12/7/2001 12:50:14 AM
From: PetzRead Replies (2) of 275872
 
muzosi, in fact, for ia64, the advantage held by a good programmmer is even greater. Especially true for scientific vector number-crunching. I personally programmed a VLIW machine (240 bits) to do a Hough transform in a 3-clock-cycle loop. A Hough transform is used to find straight lines in images. It was one of the things that Darpa used to compare supercomputer performance. The interesting thing about the 3-cycle loop is that different sections of the CPU were simulataneously calculating 5 separate iterations of the 3-cycle loop.

This, of course required that, if the number of angles or number starting points for the lines was less than 5, special code was needed. I told my boss that the machine was too complicated for an efficient compiler to ever be written for it, but they wasted a few $million trying, anyway.

I have a strong belief that really really good compilers for IA64 will never be written.

Petz
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext