<html><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; "><br><div><div>On Aug 21, 2008, at 8:52 PM, Dan Nagle wrote:</div><br class="Apple-interchange-newline"><blockquote type="cite"> <div> <!-- Converted from text/plain format --><p><font size="2">Hello,<br> <br> On Aug 21, 2008, at 8:38 PM, Van Snyder wrote:<br> ><br> > We're thinking about DSP/GPU coprocessors for these two subproblems, <br> > but<br> > then there's the rest....<br> ><br> > There isn't much of our program of the form "do n**3 things on n**2<br> > data." Mostly (82% or so) it's "do n things on n data." That is, <br> > it's<br> > mostly memory-bandwidth limited, not cache- or processor-bandwidth<br> > limited.<br> <br> <br> If you're bandwidth limited, there's absolutely *no way* in silicon<br> that a DSP/GPU can save you. The big problem with them is their<br> load and unload bandwidths.<br> <br></font></p></div></blockquote><div><br></div><div>_If_ one's memory footprint is sufficiently small then the entire domain can be placed on the high-speed memory within a GPGPU. Even if not, the IBM Cell tends to get such a high percentage of peak bandwidth too/from main memory that one can still beat conventional technology by a healthy factor. (Won't defend cost performance though.)</div><div><br></div><div>In any event, Moore's law is not the problem here. The transistor density is still growing on a nice exponential. Van's problem is with Amdahl's law. Plenty of memory bandwidth out there if you can distribute your data sufficiently. Might waste a few bucks on the unused CPU cores, but those are getting cheap compared to memory anyway.</div><div><br></div><div>Cheers,</div><div><br></div><div>- Tom</div><div><br></div><br><blockquote type="cite"><div><p><font size="2"> Game over. Insert 25 cents. ;-)<br> <br> It appears that Actual Thought will be required.<br> <br> --<br> Cheers!<br> <br> Dan Nagle<br> <br> <br> <br> <br> _______________________________________________<br> J3 mailing list<br> <a href="mailto:J3@j3-fortran.org">J3@j3-fortran.org</a><br> <a href="http://j3-fortran.org/mailman/listinfo/j3">http://j3-fortran.org/mailman/listinfo/j3</a><br> </font> </p> </div> </blockquote></div><br><div apple-content-edited="true"> <span class="Apple-style-span" style="border-collapse: separate; color: rgb(0, 0, 0); font-family: Helvetica; font-size: 12px; font-style: normal; font-variant: normal; font-weight: normal; letter-spacing: normal; line-height: normal; orphans: 2; text-align: auto; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-border-horizontal-spacing: 0px; -webkit-border-vertical-spacing: 0px; -webkit-text-decorations-in-effect: none; -webkit-text-size-adjust: auto; -webkit-text-stroke-width: 0; "><div style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; "><span class="Apple-style-span" style="border-collapse: separate; color: rgb(0, 0, 0); font-family: Helvetica; font-size: 12px; font-style: normal; font-variant: normal; font-weight: normal; letter-spacing: normal; line-height: normal; orphans: 2; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-border-horizontal-spacing: 0px; -webkit-border-vertical-spacing: 0px; -webkit-text-decorations-in-effect: none; -webkit-text-size-adjust: auto; -webkit-text-stroke-width: 0px; "><div><div>Thomas Clune, Ph.D.<span class="Apple-tab-span" style="white-space: pre; "> </span>301-286-4635 (W)</div><div>Advanced Software Technology Group<span class="Apple-tab-span" style="white-space: pre; "> </span>240-266-0400 (F)</div><div>Software Integration and Visualization Office<span class="Apple-tab-span" style="white-space: pre; "> </span><<a href="mailto:Thomas.L.Clune@nasa.gov">Thomas.L.Clune@nasa.gov</a>></div><div>NASA GSFC (610.3) <span class="Apple-tab-span" style="white-space: pre; "> </span><span class="Apple-converted-space"> </span><<a href="http://sivo.gsfc.nasa.gov">http://sivo.gsfc.nasa.gov</a>></div><div><br class="webkit-block-placeholder"></div></div></span><br class="Apple-interchange-newline"></div></span><br class="Apple-interchange-newline"> </div><br></body></html>