This looks very interesting to me.<br><div style="text-align: left;"><br></div><span class="gmail_quote">On 02/01/2008, <b class="gmail_sendername">Muli Ben-Yehuda</b> <<a href="mailto:muli@il.ibm.com">muli@il.ibm.com</a>
> wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">Would there be interest from the esteemed list members in hearing the<br>following paper, to be presented at ASPLOS '08?
<br><br>Tapping into the Fountain of CPUs---On Operating System Support for<br>Programmable Devices, by Yaron Weinsberg, Danny Dolev, Tal Anker, Muli<br>Ben-Yehuda, Pete Wyckoff.<br><br>Abstract: The constant race for faster and more powerful CPUs is
<br>drawing to a close. No longer is it feasible to significantly increase<br>the speed of the CPU without paying a crushing penalty in power<br>consumption and production costs. Instead of increasing single thread<br>performance, the industry is turning to multiple CPU threads or cores
<br>(such as SMT and CMP) and heterogeneous CPU architectures (such as the<br>Cell Broadband Engine). While this is a step in the right direction,<br>in every modern PC there is a wealth of untapped compute<br>resources. The NIC has a CPU; the disk controller is programmable;
<br>some high-end graphics adapters are already more powerful than host<br>CPUs. Some of these CPUs can perform some functions more efficiently<br>than the host CPUs. Our operating systems and programming<br>abstractions should be expanded to let applications tap into these
<br>computational resources and make the best use of them.<br><br>Therefore, we propose the Hydra framework, which lets application<br>developers use the combined power of every compute resource in a<br>coherent way. Hydra is a programming model and a runtime support layer
<br>which enables utilization of host processors as well as various<br>programmable peripheral devices' processors. We present the framework<br>and its application for a demonstrative use-case, as well as provide a<br>
thorough evaluation of its capabilities. Using Hydra we were able to<br>cut down the development cost of a system that uses multiple<br>heterogenous compute resources significantly.<br><br>Cheers,<br>Muli<br>_______________________________________________
<br>Haifux mailing list<br><a href="mailto:Haifux@haifux.org">Haifux@haifux.org</a><br><a href="http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux">http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux</a><br></blockquote>
<div style="text-align: left;"><br></div>