Differences

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

Link to this comparison view

seminars:seminar_13_10_28 [2013/10/18 18:53]
seminars:seminar_13_10_28 [2017/09/20 22:02] (current)
Line 1: Line 1:
 +====== Enabling Unified Code Base for Scalable Heterogeneous Parallel Computing ======
  
 +====  Wen-mei Hwu ====
 +{{:​seminars:​wen-mei_hwu_portrait.jpg?​300|}}{{:​seminars:​wen-mei_hwu_colorado_photo_2013.jpg?​300|}}
 +=== Professor and Sanders-AMD Endowed Chair ===
 +=== University of Illinois, Urbana-Champaign ===
 +=== CTO MulticoreWare Inc. ===
 +== Monday, October 28, 12PM-1PM (**Note special seminar DAY and TIME**) ==
 +== ISTC, CIC 4th Floor ==
 +
 +===== Abstract =====
 +The prevailing perception about programming heterogeneous parallel computing systems ​
 +today is that major types of computing devices, such as CPUs, GPUs, and DSPs, each 
 +requires a different version of source code to achieve high performance. Even with 
 +common data parallel language such as OpenCL, most developers assume that they need 
 +to write different versions of source code for different device types. Unfortunately,​ such 
 +code versioning drastically increases development,​ testing, and maintenance cost of 
 +applications. In this talk, I will present two systems, one at the OpenCL level (MxPA) ​
 +and one at functional programming level (Triolet), that enable single-source development ​
 +of data parallel code for diverse device types. MxPA is a commercial product today and 
 +Triolet is a research prototype. For MxPA, I will show the key compiler transformations ​
 +that enable a code base that is developed based on GPU performance guidelines to 
 +achieve high performance on multicore CPUs with SIMD instructions. I will also show 
 +the reasons why many GPU algorithms and CPU algorithms are converging. This brings ​
 +up an interesting question: does the traditional definition of processor architectures really ​
 +matter for application developers?
 +
 +===== Bio =====
 +Wen-mei W. Hwu is a Professor and holds the Sanders-AMD Endowed Chair in the 
 +Department of Electrical and Computer Engineering,​ University of Illinois at Urbana-
 +Champaign. He is also CTO of MulticoreWare Inc., Chief Scientist of UIUC Parallel ​
 +Computing Institute and director of the IMPACT research group (www.crhc.uiuc.edu/​
 +Impact). He directs the UIUC CUDA Center of Excellence and serves as one of the principal ​
 +investigators of the NSF Blue Waters Petascale computer project. For his contributions,​ he 
 +received the ACM SigArch Maurice Wilkes Award, the ACM Grace Murray Hopper Award, ​
 +the ISCA Influential Paper Award, and the Distinguished Alumni Award in Computer ​
 +Science of the University of California, Berkeley. He is a fellow of IEEE and ACM. Dr. Hwu
 +received his Ph.D. degree in Computer Science from the University of California, Berkeley.