Saman Amarasinghe: Difference between revisions
Created new entry for impactful computer scientist Tag: use of deprecated (unreliable) source |
(No difference)
|
Revision as of 02:02, 24 April 2024
Saman Amarasinghe is a Professor in the Department of Electrical Engineering and Computer Science at the Massachusetts Institute of Technology (MIT), conducting research as part of the Computer Science and Artificial Intelligence Laboratory (CSAIL) and leading the lab’s Commit compiler research group[1].
His work has focused on computer architectures, programming languages and compilers that maximize application performance, including helping create multiple influential domain-specific languages (DSLs) such as Halide for image processing.
In 2019 he was named a fellow of the Association for Computing Machinery for “contributions to high performance computing on modern hardware platforms, domain-specific languages, and compilation techniques”[2].
Amarasinghe earned his bachelor of science degree in Electrical Engineering and Computer Science from Cornell University in 1988, before getting a PhD at Stanford University in 1997. He then joined the faculty at MIT, where he has served a term as Associate Department Head for Electrical Engineering and Computer Science[3].
He has published more than 120 articles in international conferences and journals, accruing more than 26,000 citations and an h-index of 72[4]. His work has included contributions to high-performance computing on modern hardware platforms, the development of several important DSLs, and key compilation techniques.
Among the DSLs he has helped develop are Halide, TACO[5], and GraphIt[6].
Halide was designed with the goal of making it easier to write and maintain high-performance image processing or array processing code, taking advantage of multicore CPUs and GPUs to separate algorithms from their execution schedules[7].
TACO (“tensor algebra compiler”) was designed for sparse data operations. Research from Amarasinghe’s group has shown that its performance can be several times faster than MATLAB[8] and competitive with high-performance libraries like Eigen [9] and Intel oneMKL[10].
GraphIt was created to support faster and more efficient graph processing[11].
Amarasinghe has also helped launch multiple companies. He founded Determina Corporation, which was later acquired by VMware[12]. He also co-founded Venti Technologies and DataCebo[13].
- ^ mit.edu http://groups.csail.mit.edu/commit/.
{{cite web}}: Missing or empty|title=(help) - ^ acm.org https://awards.acm.org/award-recipients/amarasinghe_4130225.
{{cite web}}: Missing or empty|title=(help) - ^ thetech.com https://thetech.com/2018/04/26/6-11-joint-major.
{{cite web}}: Missing or empty|title=(help) - ^ scholar.google.com https://scholar.google.com/citations?user=cF6i_goAAAAJ&hl=en&oi=ao.
{{cite web}}: Missing or empty|title=(help) - ^ phys.org https://phys.org/news/2017-10-tensor-algebra-fold-speedups-previous.html.
{{cite web}}: Missing or empty|title=(help) - ^ datanami.com https://www.datanami.com/2018/12/10/graphit-promises-big-speedup-in-graph-processing/.
{{cite web}}: Missing or empty|title=(help) - ^ betterprogramming.pub https://betterprogramming.pub/write-fast-and-maintainable-code-with-halide-part-1-6a5c3a519250/.
{{cite web}}: Missing or empty|title=(help) - ^ mathworks.com https://www.mathworks.com/help/simulink/libraries.html.
{{cite web}}: Missing or empty|title=(help) - ^ tuxfamily.org https://eigen.tuxfamily.org/.
{{cite web}}: Missing or empty|title=(help) - ^ intel.com https://www.intel.com/content/www/us/en/develop/documentation/oneapi-programming-guide/top/api-based-programming/intel-oneapi-math-kernel-library-onemkl.html.
{{cite web}}: Missing or empty|title=(help) - ^ datanami.com https://www.datanami.com/2018/12/10/graphit-promises-big-speedup-in-graph-processing/.
{{cite web}}: Missing or empty|title=(help) - ^ zdnet.com https://www.zdnet.com/article/vmware-buys-determina/.
{{cite web}}: Missing or empty|title=(help) - ^ crunchbase.com https://www.crunchbase.com/person/saman-amarasinghe.
{{cite web}}: Missing or empty|title=(help)