There is a sign-up schedule for this event that can be found at

http://apps.cs.utexas.edu/talkschedules/cgi/list_events.cgi

Type o

f Talk: UTCS Faculty Candidate

Speaker/Affiliation: Joseph Devietti/Un

iversity of Washington, Seattle, WA

Talk Audience: UTCS Faculty, Gr

aduate Students, Undergraduate Students and Outside Interested Parties

Date/Time: Thursday, April 5, 2012, 11:00 am

Location: ACES 2.302

Host: Calvin Lin

Talk Title: No Such Thing as Luck: Improving Pa

rallel Programmability with Determinism

Talk Abstract:
Nondeterminis

m is a key complication in programming multicore systems.
It makes testin

g more difficult and less useful, since another run of
the program can p

otentially introduce new behaviors. Nondeterminism
also frustrates debugg

ing efforts by making bugs hard to reproduce.
Previous approaches to copi

ng with nondeterminism in parallel programs
have focused on recording an

execution for subsequent replay, or
required that programs be written in
restrictive languages, but have
not addressed the underlying nondetermi

nism of multicore systems in a
direct way.

In this talk, I will sho

w how to use novel hardware and software
techniques to provide determinis

tic execution for arbitrary parallel
programs written in today''s languag

es. I''ve built a series of
deterministic platforms, from new hardware a

rchitectures to compilers
and language extensions, that show how the cha

llenge of nondeterminism
can be addressed across the computing stack. Har

dware speculation,
memory consistency relaxations, and hardware-softwar

e co-design all
play key roles in improving the performance and simplicit

y of
determinism. I''ll also share my plans for future work, from levera

ging
determinism to accelerate safety and security checks, to new parall

el
computer architectures that enable a unified task+data parallelism
a

bstraction.

Bio:
I am a 5th-year PhD student at the University of Wa

shington. I work
with my advisers Luis Ceze and Dan Grossman on making mu

ltiprocessors
easier to program by leveraging changes in computer archite

cture,
compilers, runtime systems and languages. I was awarded an Intel
PhD
Fellowship for 2011-12, and I‚ve had two papers selected for the

annual "IEEE Micro Top Picks from Computer Architecture Conferences"
aw

ard.