From johann.knechtel at ifte.de Tue May 17 08:40:51 2011 From: johann.knechtel at ifte.de (Johann Knechtel) Date: Tue, 17 May 2011 14:40:51 +0200 Subject: [Hotspot] TSV impact modelling? Message-ID: <4DD26CD3.9090305@ifte.de> In your HOWTOs, you mention that modeling TSV or thermal-via impact is possible by adapting the source code. I was wondering whether this allows flexible simulation for different layouts or means to hardcode TSV locations? Also, has this been already done and proven? Thanks Johann -- http://www.ifte.de/english/staff/knechtel.html TU Dresden, IFTE, BAR II/47, Phone: +49 351 463 39612 From wh6p at virginia.edu Wed May 18 15:09:52 2011 From: wh6p at virginia.edu (Wei Huang) Date: Wed, 18 May 2011 14:09:52 -0500 Subject: [Hotspot] TSV impact modelling? In-Reply-To: <4DD26CD3.9090305@ifte.de> References: <4DD26CD3.9090305@ifte.de> Message-ID: Hello Johann, Sorry for the late reply. HotSpot doesn't support adding and modeling individual TSVs as of now. What it can do is to model averaged thermal impact from a TSV array. A user can simulate this by changing the effective thermal properties for the layer that includes TSV and surrounding materials. However, there have been efforts to model more than one types of materials in each layer (TSVs and their surrounding dielectrics in your case), for example, Prof. Ayse Coskun's group at Boston University. -Wei On Tue, May 17, 2011 at 7:40 AM, Johann Knechtel wrote: > In your HOWTOs, you mention that modeling TSV or thermal-via impact is > possible by adapting the source code. I was wondering whether this > allows flexible simulation for different layouts or means to hardcode > TSV locations? Also, has this been already done and proven? > > Thanks > Johann > -- > http://www.ifte.de/english/staff/knechtel.html > TU Dresden, IFTE, BAR II/47, Phone: +49 351 463 39612 > _______________________________________________ > 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/20110518/f5f7c6f8/attachment.html From johann.knechtel at ifte.de Wed May 25 08:19:01 2011 From: johann.knechtel at ifte.de (Johann Knechtel) Date: Wed, 25 May 2011 14:19:01 +0200 Subject: [Hotspot] TSV impact modelling? In-Reply-To: References: <4DD26CD3.9090305@ifte.de> Message-ID: <4DDCF3B5.90605@ifte.de> Thanks Wei. I contacted Ayse Coskun, and she told me that they are working on an extension for HotSpot to handle single / grouped TSVs. As for now, I will stick to Ansys for my simulation. Thanks again, Johann On 18.05.2011 21:09, Wei Huang wrote: > Hello Johann, > > Sorry for the late reply. HotSpot doesn't support adding and modeling > individual TSVs as of now. What it can do is to model averaged thermal > impact from a TSV array. A user can simulate this by changing the > effective thermal properties for the layer that includes TSV and > surrounding materials. > > However, there have been efforts to model more than one types of > materials in each layer (TSVs and their surrounding dielectrics in your > case), for example, Prof. Ayse Coskun's group at Boston University. > > -Wei > > On Tue, May 17, 2011 at 7:40 AM, Johann Knechtel > > wrote: > > In your HOWTOs, you mention that modeling TSV or thermal-via impact is > possible by adapting the source code. I was wondering whether this > allows flexible simulation for different layouts or means to hardcode > TSV locations? Also, has this been already done and proven? > > Thanks > Johann > -- > http://www.ifte.de/english/staff/knechtel.html > TU Dresden, IFTE, BAR II/47, Phone: +49 351 463 39612 > _______________________________________________ > HotSpot mailing list > HotSpot at mail.cs.virginia.edu > http://www.cs.virginia.edu/mailman/listinfo/hotspot > > -- http://www.ifte.de/english/staff/knechtel.html TU Dresden, IFTE, BAR II/47, Phone: +49 351 463 39612 From wh6p at virginia.edu Wed May 25 13:57:25 2011 From: wh6p at virginia.edu (Wei Huang) Date: Wed, 25 May 2011 12:57:25 -0500 Subject: [Hotspot] TSV impact modelling? In-Reply-To: <4DDCF3B5.90605@ifte.de> References: <4DD26CD3.9090305@ifte.de> <4DDCF3B5.90605@ifte.de> Message-ID: Hi Johann, Thanks for the update. It would be great to cross-check ANSYS and Ayse's HotSpot extension once it is available. Please keep us informed and good luck with your thermal simulations. -Wei On Wed, May 25, 2011 at 7:19 AM, Johann Knechtel wrote: > Thanks Wei. I contacted Ayse Coskun, and she told me that they are > working on an extension for HotSpot to handle single / grouped TSVs. As > for now, I will stick to Ansys for my simulation. > > Thanks again, > Johann > > On 18.05.2011 21:09, Wei Huang wrote: > > Hello Johann, > > > > Sorry for the late reply. HotSpot doesn't support adding and modeling > > individual TSVs as of now. What it can do is to model averaged thermal > > impact from a TSV array. A user can simulate this by changing the > > effective thermal properties for the layer that includes TSV and > > surrounding materials. > > > > However, there have been efforts to model more than one types of > > materials in each layer (TSVs and their surrounding dielectrics in your > > case), for example, Prof. Ayse Coskun's group at Boston University. > > > > -Wei > > > > On Tue, May 17, 2011 at 7:40 AM, Johann Knechtel > > > wrote: > > > > In your HOWTOs, you mention that modeling TSV or thermal-via impact > is > > possible by adapting the source code. I was wondering whether this > > allows flexible simulation for different layouts or means to hardcode > > TSV locations? Also, has this been already done and proven? > > > > Thanks > > Johann > > -- > > http://www.ifte.de/english/staff/knechtel.html > > TU Dresden, IFTE, BAR II/47, Phone: +49 351 463 39612 > > _______________________________________________ > > HotSpot mailing list > > HotSpot at mail.cs.virginia.edu > > http://www.cs.virginia.edu/mailman/listinfo/hotspot > > > > > > > -- > http://www.ifte.de/english/staff/knechtel.html > TU Dresden, IFTE, BAR II/47, Phone: +49 351 463 39612 > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20110525/dd583395/attachment.html From acoskun at cs.ucsd.edu Fri May 27 22:20:55 2011 From: acoskun at cs.ucsd.edu (Ayse Kivilcim Coskun) Date: Fri, 27 May 2011 22:20:55 -0400 Subject: [Hotspot] TSV impact modelling? In-Reply-To: References: Message-ID: <4DE05C07.9020305@cs.ucsd.edu> Hi Johann, Wei, In case others are interested in the topic -- We have modeled single / grouped TSVs before in our papers (e.g., Date'10, VLSI-SoC'09). I can provide some guidelines for how to do this if anyone wants to implement it (contact me at acoskun at bu.edu). At the moment we are working on generalizing the approach -as an extension to HotSpot- to model arbitrary TSV distributions across the chip with a more flexible user-interface. We are hoping to finalize this extension by the end of the summer. -Ayse From wowlyy at hotmail.com Mon May 30 14:40:11 2011 From: wowlyy at hotmail.com (Yangyang Li) Date: Mon, 30 May 2011 12:40:11 -0600 Subject: [Hotspot] temperature changing Message-ID: Hi HotSpot, In our current program ,we need to considering the circuit temperature and temprature changing. For a gird, its temperature should be in an range, and change slowly if the transistors keep switching. For now, we just assume it will change linearly between lower and upper bound. But we know it's not the real case. Can you give us any suggestion about the approximate distribution function for a grid temperature,or what we should do to get it? Thanks in advance. Yangyang -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20110530/28c6ce3d/attachment.html