From devendra.rai at tik.ee.ethz.ch Tue Apr 5 14:11:05 2011 From: devendra.rai at tik.ee.ethz.ch (Devendra Rai) Date: Tue, 5 Apr 2011 20:11:05 +0200 Subject: [Hotspot] Boundary Conditions: Need a re-check Message-ID: Hello HotSpot I am using the default EV6/alpha floorplan which ships with hotspot. I found some multi-processor platforms with the ASU's magma project, and I picked up an 8x8 EV6 floorplan. I am running power numbers (again, magma project, bzip2 load), and I find a strange issue. The cores on the corner show the highest temperature while the cores towards inside show a lower temperature. This is contrary to the perception which says that for the same power dissipation, cores on the corner will be the coolest. Attached are the matlab fig files for your review. Can you please let me know if there is any explanation for this? The hotspot "engine" is un-modified. Best, Devendra Rai -------------- next part -------------- A non-text attachment was scrubbed... Name: FPMult_Silicon_ML_Corner.fig Type: image/x-xfig Size: 28222 bytes Desc: not available Url : http://www.cs.virginia.edu/pipermail/hotspot/attachments/20110405/49c33b9a/attachment-0002.bin -------------- next part -------------- A non-text attachment was scrubbed... Name: FPMult_Silicon_ML_inner.fig Type: image/x-xfig Size: 28150 bytes Desc: not available Url : http://www.cs.virginia.edu/pipermail/hotspot/attachments/20110405/49c33b9a/attachment-0003.bin From wh6p at virginia.edu Tue Apr 5 19:55:57 2011 From: wh6p at virginia.edu (Wei Huang) Date: Tue, 5 Apr 2011 18:55:57 -0500 Subject: [Hotspot] Boundary Conditions: Need a re-check In-Reply-To: References: Message-ID: HotSpot assumes no heat transfer path on the edges of silicon. That is why it reports higher corner temperatures. -Wei On Tue, Apr 5, 2011 at 1:11 PM, Devendra Rai wrote: > Hello HotSpot > > I am using the default EV6/alpha floorplan which ships with hotspot. I > found some multi-processor platforms with the ASU's magma project, and > I picked up an 8x8 EV6 floorplan. > > I am running power numbers (again, magma project, bzip2 load), and I > find a strange issue. The cores on the corner show the highest > temperature while the cores towards inside show a lower temperature. > This is contrary to the perception which says that for the same power > dissipation, cores on the corner will be the coolest. > > Attached are the matlab fig files for your review. Can you please let > me know if there is any explanation for this? The hotspot "engine" is > un-modified. > > Best, > > > Devendra Rai > > _______________________________________________ > HotSpot mailing list > HotSpot at mail.cs.virginia.edu > http://www.cs.virginia.edu/mailman/listinfo/hotspot > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20110405/beada2b1/attachment.html