From khurshid at wisc.edu  Mon Apr 22 18:04:07 2013
From: khurshid at wisc.edu (Mushfique Khurshid)
Date: Mon, 22 Apr 2013 17:04:07 -0500
Subject: [Hotspot] DTM in hotspot
Message-ID: <CAKLY5EmLKs+-2V4DFX4DtyME1jF2g7uXCgqSq=fAL2+hpqo1yQ@mail.gmail.com>

My question is what does hotspot do with regards to the option DTM_USED in
the *.config? So far as I see in the code, this option only is applied for
the INITIAL TEMPERATURE? Is that true? What about during the simulation?
And if it is during the simulation, then is there any way for us to check
how DTM affected performance/exec. time?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.cs.virginia.edu/pipermail/hotspot/attachments/20130422/b128f4be/attachment.html 

From a0028013 at nus.edu.sg  Sat Apr 27 07:42:27 2013
From: a0028013 at nus.edu.sg (Hou Junsong)
Date: Sat, 27 Apr 2013 22:42:27 +0800
Subject: [Hotspot] 3D IC floorplan
Message-ID: <338A8233F2BB084E8969F5323E947A2911F64F0A8A@MBX25.stu.nus.edu.sg>

Hi! 

I am trying to develop a 3D "hotfloorplan model" for my project. For the first step, I only need to handle a simplified case ( all the blocks are square with same dimension and die is also in square dimension). I have read through the souce code and I have some questions in below:


1. the grid model and block model actually use different calculation methods, is there any reason to use two different methods instead of using one? 

2. In fact the block model is more suitable for my case. If I follow the hotfloorplan block model and add the layer in a way which is used in the  gird 3d stacking model, is there any thing I need to pay special attention?

I have such questions because I don't understand the reason why only grid model is used for 3D stacking and if there is any constraints for the block model in 3D stacking.


Regards!
J.S. 

From wh6p at virginia.edu  Sat Apr 27 18:27:07 2013
From: wh6p at virginia.edu (Wei Huang)
Date: Sat, 27 Apr 2013 20:27:07 -0500
Subject: [Hotspot] 3D IC floorplan
In-Reply-To: <338A8233F2BB084E8969F5323E947A2911F64F0A8A@MBX25.stu.nus.edu.sg>
References: <338A8233F2BB084E8969F5323E947A2911F64F0A8A@MBX25.stu.nus.edu.sg>
Message-ID: <41CF6349-554C-4A5C-9F39-1527C33A785F@virginia.edu>

Hi, block model can't be used for 3d cases because it is difficult to deal with blocks on one layer that are not aligned with blocks on a adjacent layer. On the other hand, grid model doesn't have this problem since grid cells on different layers are perfectly aligned.

Hope this is helpful.

-Wei

On Apr 27, 2013, at 9:42 AM, Hou Junsong <a0028013 at nus.edu.sg> wrote:

> Hi! 
> 
> I am trying to develop a 3D "hotfloorplan model" for my project. For the first step, I only need to handle a simplified case ( all the blocks are square with same dimension and die is also in square dimension). I have read through the souce code and I have some questions in below:
> 
> 
> 1. the grid model and block model actually use different calculation methods, is there any reason to use two different methods instead of using one? 
> 
> 2. In fact the block model is more suitable for my case. If I follow the hotfloorplan block model and add the layer in a way which is used in the  gird 3d stacking model, is there any thing I need to pay special attention?
> 
> I have such questions because I don't understand the reason why only grid model is used for 3D stacking and if there is any constraints for the block model in 3D stacking.
> 
> 
> Regards!
> J.S. 
> _______________________________________________
> HotSpot mailing list
> HotSpot at mail.cs.virginia.edu
> http://www.cs.virginia.edu/mailman/listinfo/hotspot