Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
wiki:aca2017:lectures [2018/01/22 20:34] – [Jan. Thu 18] Andreas Moshovoswiki:aca2017:lectures [2018/03/27 11:12] (current) – [Mar 20, Tue: Google's TPU] Andreas Moshovos
Line 1: Line 1:
 ====== Lecture Material and Schedule ====== ====== Lecture Material and Schedule ======
 +
 +====== Alternate Execution Models and Architectures ======
  
 ===== Jan. Tue 9: Meltdown, Virtual Memory and Caches ===== ===== Jan. Tue 9: Meltdown, Virtual Memory and Caches =====
Line 23: Line 25:
 Rethinking Cache Replacement Policies: {{ :wiki:aca2017:cache_insertion_policies.pptx | Cache Insertion Policies}}. See Assignment #1. Rethinking Cache Replacement Policies: {{ :wiki:aca2017:cache_insertion_policies.pptx | Cache Insertion Policies}}. See Assignment #1.
  
-===== Jan. Tue 23  =====+===== Jan. Tue 23: Multiscalar and Slipstream  =====
  
 Accelerating Sequential Programs: Alternate Architectures Accelerating Sequential Programs: Alternate Architectures
Line 29: Line 31:
   * Multiscalar: {{ :wiki:aca2017:multiscalar.pptx |slides}}, {{ :wiki:aca2017:isca.multiscalar.pdf |paper}}   * Multiscalar: {{ :wiki:aca2017:multiscalar.pptx |slides}}, {{ :wiki:aca2017:isca.multiscalar.pdf |paper}}
   * Slipstream: {{ :wiki:aca2017:slipstream.pptx |slides}}, {{ :wiki:aca2017:a_study_of_slipstream_processors.pdf |paper}}   * Slipstream: {{ :wiki:aca2017:slipstream.pptx |slides}}, {{ :wiki:aca2017:a_study_of_slipstream_processors.pdf |paper}}
 +
 +===== Jan. Tue 30: Slipstream and Replay =====
 +
 +  * RePlay: {{ :wiki:aca2017:replay.pptx | slides}}, {{ :wiki:aca2017:replay.pdf | paper}}
 +
 +===== Feb 1, Tue: Criticality =====
 +
 +  * {{ :wiki:aca2017:critical_path.pptx | Slides}} {{ :wiki:aca2017:critical_path_prediction.pdf | "Focusing Processor Policies via Critical-Path Prediction', Fields, Rubin, Bodik, ISCA '01.}}
 +
 +====== Machine Learning Acceleration ======
 +===== Feb 5, Mon: DaDianNao =====
 +  * {{ :wiki:aca2017:dadiannao.pdf |DaDianNao}} [[https://youtu.be/4pSq8KQLKYY| Lecture Video]]
 +
 +===== Feb 13, Tue:  Eyeriss =====
 +  * {{ :wiki:aca2017:eyeriss-isca.pdf |Eyeriss ISCA}}{{ :wiki:aca2017:eyeriss.pptx |Slides}}
 +
 +===== Feb 15, Thu:  Efficient Inference Engine =====
 +  * {{ :wiki:aca2017:eie.pdf | EIE paper ISCA}}, {{ :wiki:aca2017:deep_compression.pdf | Deep Compression paper ICLR}}, {{ :wiki:aca2017:eie_presentation.pptx | Slides (base deck courtesy of Kevin Siu}}
 +
 +===== Week of Feb 22 ======
 +
 +Reading week, no lectures.
 +===== Feb 26, Tue:  SCNN =====
 +  * {{ :wiki:aca2017:scnn.pptx | Slides}}, [[https://dl.acm.org/citation.cfm?id=3080254| Link to ISCA paper through ACM's Library]] you should be able to access this through a UofT network connection, otherwise another version is on [[https://arxiv.org/abs/1708.04485|Arxiv]]
 +
 +===== Mar 1, Thurs:  Value-Based Deep Learning Acceleration =====
 +  * {{ :wiki:aca2017:based-deep-learning_1_.pdf | Overview paper}}
 +
 +===== Mar 20, Tue:  Google's TPU =====
 +  * {{ :wiki:aca2017:tpu.pdf | ISCA 2017 Paper (arxiv version)}} , {{ :wiki:aca2017:tpu.pptx | slides}}
 +
 +===== Mar 27, Tue:  Out-Of-Order Circuit Structures and Optimizations =====
 +
 +  * {{ :wiki:aca2017:scheduler_structures.ppt | Structures}}
 +  * {{ :wiki:aca2017:scheduler-optimizations.pptx |Optimizations}}