Thanks to Will for explaining work queue model and perculation yesterday. If you have not read the papers, please do. This is an interesting execution model and might gain traction as we push from peta to exa scale.
I will be out of town on Thursday (phd proposal at U. of South Carolina). So no weekly meetings. Please reply to this message with your weekly status.
Ron
Hello All,
Here is your weekly reminder that we have the reading group at 3:30 today and the group meeting immediately following it. I will be on campus from 12-5pm today but I have leave promptly at 5.
Liu will be doing one technical paper today and then he has prepared a CUDA tutorial for us. Please send me the answer to this question via email: How does AutoPilot decide to lay out variables? (i.e. on-chip, off-chip, arrays v. scalars).
Also, please post your status as a reply to this message. Thanks!
Ron