Università degli Studi di Siena Department of Information Engineering and Mathematics (DIISM) Course of High Performance Computer Architecture 2017-2018

Latest News (home)   Registration   Lessons (restricted access)   Errata slides   Tools for lab   Syllabus   Office Hours   Mid-terms   Projects   Group registration (restricted access)   Exam Rules   Feedback   RELATED LINKS   BEST PROJECTS bgcolor="#FFFFFF" width="840" height="3900" border="0" cellpadding="0" cellspacing="0" valign="top" align="left">  FIRST PART SLIDES (THEORY) IN A SINGLE FILE  SECOND PART SLIDES (THEORY) IN A SINGLE FILE COURSE SCHEDULE (TENTATIVE): THE FOLLOWING DATES ARE INDICATIVE AND RELATED TO THE HYPOTHETIC REGULAR PROGRESS OF THE LESSONS. NOTE (THIS HAS BEEN ASKED TO ME ...) THE NUMBERING OF LESSON REFERS ONLY TO THE TOPIC, IT DOES NOT IMPLY BY ANY MEANS THAT THE LESSONS WILL BE GIVEN IN A CERTAIN ORDER. NOTE: the actual schedule will be updated weekly. BIBLIOGRAPHIC REFERENCES: M. Dubois, M. Annavaram, P. Stenstrom, "Parallel Computer Organization and Design", Cambridge University Press, 2012, ISBN: 978-0-521-88675-8  LESSON #01 of 02-Oct-2017 (9:00-13:00) Introduction, Evaluating Computers Dubois Chap. 1,3.3 (Hennessy-Patterson-4 - 2.1,2.2) (Hennessy-Patterson-5 - 3.1,3.2)  LESSON #02 of 03-Oct-2017 (9:00-10:00) Dynamic Instruction Scheduling Dubois Chap. 3.4,3.4.1 (Hennessy-Patterson-4 - 2.4,2.5) (Hennessy-Patterson-5 - 3.4,3.5) An Efficient Algorithm for Exploiting Multiple Arithmetic Units tomasulo.c  PRACTICING/LAB #01 of 03-Oct-2017 (01:00-02:00) Dynamic Scheduling exercize. Exercize from a previous test (23/6/2005).  LESSON #03 of 09-Oct-2017 (9:00-11:00) Branch Prediction: speculation of branch condition and branch target, BPRED, BTB. Dubois - 3.3.4,3.4.3 (Hennessy-Patterson-4 - 2.3) (Hennessy-Patterson-5 - 3.3) Paper [Yeh, Patt - 1992] Paper [Nair - 1995] Paper [Young - 1995]  LESSON #04 of 09-Oct-2017 (02:00-04:00) Predictor types, Bimodal, BHSR, BHT, PHT, 2-level adaptive, GAg, PAg, PAs. Other predictors (gshare, gselect, tournament). Paper [McFarling - 1993]  PRACTICING/LAB #90 of 10-Oct-2017 (9:00-11:00) Introduction to Linux (first part)  LESSON #05 of 16-Oct-2017 (9:00-11:00) Introduction to Superscalar Processors: general scheme and Renaming. Dubois 3.3.3,3.4.6(Hennessy-Patterson-4 - 2.6-2.9) (Hennessy-Patterson-5 - 3.6-3.10)  LESSON #06 of 16-Oct-2017 (02:00-04:00) Superscalar execution example: Re-Order Buffer and Instruction Window. Case studies: MIPS, Alpha, AMD, Intel, ARM. Dubois 3.4.4,3.4.5,3.4.7,3.4.8,3.4.9  PRACTICING/LAB #90 of 17-Oct-2017 (9:00-10:00) Introduction to Linux (second part)  PRACTICING/LAB #02 of 17-Oct-2017 (01:00-02:00) Using the Superscalar simulator FREESS Educational Simulator FreeSs Exam of 22/06/2015 (exercize No.1) and solution  LESSON #08 of 23-Oct-2017 (9:00-11:00) Software methods to extract Instruction Level Parallelsim. Dubois 3.3.5,3.5,3.5.1-5 (see also Hennessy-Patterson-4 2.7-2.8)  PRACTICING/LAB #03 of 23-Oct-2017 (02:00-04:00) Exercizes on SUPERSCALAR AND VLIW processors. Exam of 06/11/2015 (exercize No.1) and solution Exam of 30/06/2008 and solution Exam of 07/11/2014 and solution  PRACTICING/LAB #04 of 24-Oct-2017 (9:00-11:00) Exercizes on Dynamic Scheduling and various excersizes. Exam of 07/11/2014 and solution Exam of 06/11/2015 (exercize No.1) and solution Exam of 07/07/2009 and solution  PRACTICING/LAB #05 of 30-Oct-2017 (9:00-13:00) Exercizes on branch prediction and verious excercizes. Exam of 16/01/2012 Exam of 28/06/2007 and solution SS-EX1.pdf  31-Oct-2017 - MIDTERM TEST (09:00-11:00)  LESSON #11 of 06-Nov-2017 (9:00-10:00) Introduction to multiprocessor systems, Flynn's taxonomy, UMA, NUMA, COMA systems, programming models Dubois 5.1,5.4 (see also Culler-Singh, Cap.1).  LESSON #12 of 06-Nov-2017 (01:00-04:00) Coherence Protocols: Write Update, Write Invalidate, Hybrid. Snoopy based protocols: the MESI and DRAGON protocols Dubois 5.4,5.5,7.3 (see also Culler-Singh, Cap.5).  PRACTICING/LAB #06 of 07-Nov-2017 (9:00-11:00) Introducing the framework for performance estimation through COTSon; examples. Paper on Dynamic sampling in COTSon. COTSON USER GUIDE  LESSON #14 of 13-Nov-2017 (9:00-11:00) Memory Consistency Models: Sequential Consistency and Relaxed Consistency Dubois 7.4,7.5,7.6,7.7 (v. Culler-Singh, cap. 5.2, 5.5) Litmus Tests for checking Memory Models  LESSON #60 of 13-Nov-2017 (02:00-04:00) Introduction to FPGAs  PRACTICING/LAB #07 of 14-Nov-2017 (9:00-10:00) Introduction to FPGA programming with Xilinx Vivado HLS  PRACTICING/LAB #08 of 14-Nov-2017 (01:00-02:00) COTSon metrics, tests, timers COTSON CJPEG BENCHMARKS & TOOLS Instructions for the CJPEG benchmark  LESSON #61 of 20-Nov-2017 (9:00-11:00) Introduction to a dataflow parallel programming model Support material.  LESSON #15 of 20-Nov-2017 (02:00-04:00) Introduction to Parallel Programming  PRACTICING/LAB #09 of 21-Nov-2017 (9:00-11:00) Programming examples of dataflow programming on MAXELER platform MAXELER website  LESSON #16 of 27-Nov-2017 (9:00-11:00) Parallelsim and Performance  PRACTICING/LAB #09 of 27-Nov-2017 (02:00-04:00) Exercizes on Coherency from past exams. Exercize on cache coherence 20/01/2010; spreedsheet for this excercize  PRACTICING/LAB #11 of 28-Nov-2017 (9:00-11:00) Experimenting several programming models: Pthreads, OpenMP, TBB, Cilk Source code of the examples in Pthreads, OpenMP, TBB, Cilk Using CILKTOOLS (cilkview,cilkscreen)  LESSON #21 of 04-Dec-2017 (9:00-10:00) Introduction to CUDA parallel programming model Bill Dally @ SC-2010  LESSON #22 of 04-Dec-2017 (01:00-02:00) CUDA Programming Elements  LESSON #23 of 04-Dec-2017 (02:00-03:00) Atomic Constructs and Threads in CUDA  LESSON #24 of 04-Dec-2017 (03:00-04:00) CUDA Memory Model  PRACTICING/LAB #12 of 05-Dec-2017 (9:00-11:00) CUDA Excersizes MATRIX-MULTIPLY CODE PI-ESTIMATE CODE CUDA CheatSheet  LESSON #30 of 11-Dec-2017 (9:00-10:00) Clusters  LESSON #40 of 11-Dec-2017 (01:00-02:00) Introduction to MPI Dubois 5.2.2, 5.3  LESSON #50 of 11-Dec-2017 (02:00-03:00) Xeon Phi  PRACTICING/LAB #17 of 11-Dec-2017 (03:00-04:00) Using MPI HELLOWORLD MPI CODE TESTMPI CODE Exam of 18/12/2015 (exercize No.1) and solution  PRACTICING/LAB #18 of 12-Dec-2017 (9:00-10:00) Programming examples of Intel Xeon Phi platform  PRACTICING/LAB #12 of 12-Dec-2017 (01:00-02:00) Various exercizes from previous exams. Exam of 16/01/2015 (exercize No.1) and solution  PRACTICING/LAB #19 of 18-Dec-2017 (9:00-13:00) Various exercizes from previous exams. Exam of 22/06/2015 (exercize No.1) and solution  19-Dec-2017 - FINAL TEST (09:00-11:00)  from 25-Dec-2017 to 02-Jan-2018: No Lesson VACANZA/PONTE To visualize the content of the above presentations in Acrobat format (.pdf) you can use Acrobat Reader, freely realesed by Adobe (Download Acrobat Reader)