[Colloq] Invited talk by Dr. Norman Rubin, NVIDIA
David Kaeli
kaeli at ece.neu.edu
Mon May 19 21:06:03 EDT 2014
All,
You are invited to the following invited talk:
*"Parallel Javascript: Dynamic Dispatch with High Performance"*
*Presenter: Dr. Norman Rubin, Principal Research Scientist, NVIDIA *
*May 21, 2014, 3:00PM*
*442 Dana Research Center*
Abstract:
JavaScript is often considered a low performance scripting language but
lately JavaScript performance has improved enough to allow building
substantial performance critical applications within the browser.
This talk will describe a research prototype that allows JavaScript
programmers to access the power of a modern GPU without needing to use
specialized languages.
The prototype implementation of parallel JavaScript (PJS) adds a new data
type (the accelerated array) and small set of new higher level library
functions such as map and reduce. These operations can execute on either
a GPU or CPU, the PJS runtime automatically dispatches work over
heterogeneous processors without user intervention. Provided with a high
level functional model named PJS, we allow developers to write code using
the natural JavaScript idioms while still providing lots of performance.
Since the operations express the intent of the program, PJS has the
necessary information to generate code that is often performance portable
– that is code that is competitive with CUDA code on the GPU and pure
loops on the CPU. At the same time high level operations allow
programmers to write very compact code. On a low-end GPUs, PJS is up to 20
times faster than stock JavaScript . The talk will include some demos.
Bio:
Norm Rubin has many years of experience delivering commercial compilers
for processors ranging from embedded (ARM), desktop (HP, ALPHA) and
supercomputer (KSR), and is a recognized expert in th field. He was the
architect and lead implementer for the widely used graphics compiler for
AMD/ATI. That compiler is currently shipping on millions of machines
including cell phones, consoles, and PCs. Norm was part of the AMD
architecture team that designed GCN (Graphics core next). He was the lead
designer of HSAIL, the virtual machine used in the HSA system
architecture. Around a year ago he moved to NVIDA Research where he is
working in algorithms and future programming models. Lately Norm has been
looking at extending JavaScript to use GPUS and heterogeneous devices. Norm
is also a visiting scholar at Northeastern University.
Dr. Rubin holds a PhD from the Courant Institute of NYU. Besides his work
in compilers and architecture, he is well known for his work in GPU
systems, compiler related parts of the tool chain, binary translators and
dynamic optimizers.
Host: David Kaeli
*________________________________________*
*Prof. David Kaeli*
*Dept. of Electrical and Computer Engineering*
*Northeastern University*
*333 Dana Research Center*
*Boston, MA 02115*
*(617)-373-5413*
*email: kaeli at ece.neu.edu <kaeli at ece.neu.edu>*
*________________________________________*
More information about the Colloq
mailing list