[Nets-seminars] TODAY, 2 PM: UCL CS Distinguished Lecture: Prof Babak Falsafi (EPFL), MPEB 6.12

[Brad Karp]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Just a reminder about today's Distinguished Lecture by Prof Babak
Falsafi of EPFL (formerly of Carnegie Mellon), renowned computer
architect, who will be speaking on fundamental scaling limitations
facing modern CPUs, and how to continue building faster CPUs despite
these limitations.

2 PM in MPEB 6.12.

Original talk announcement with title, abstract, and bio below.

This is a rare chance to hear about the future of server CPUs from a
leader in the field; his talk will target a general CS audience. All
very welcome to attend!

- -Brad, bkarp at cs.ucl.ac.uk

- -------- Original Message --------
Subject: UCL CS Distinguished Lecture: Prof Babak Falsafi (EPFL), Tue
4 Feb, 2          PM, MPEB 6.12
Date: Sun, 2 Feb 2014 02:47:47 +0000
From: Brad Karp <B.Karp at cs.ucl.ac.uk>
To: research at cs.ucl.ac.uk, nets at cs.ucl.ac.uk, nets-seminars at cs.ucl.ac.uk
CC: Brad Karp <B.Karp at cs.ucl.ac.uk>

Greetings, everyone.

It's my pleasure to announce a talk by Prof Babak Falsafi of EPFL
(formerly of CMU), renowned computer architect. Babak will speak in
MPEB 6.12 at 2 PM on Tuesday, the 4th of February.

Let me emphasize that no matter the area of CS you work in, this is a
talk you should find relevant and interesting.

Computer architecture is hugely important to the work of practically
every computer scientist, as it dictates what the CPUs we all use look
like (and thus what algorithms need to look like). For decades,
computer architects devised ingenious techniques for increasing CPU
clock speeds and extracting ever greater performance from the same
sequential programs. They then determined that clock speed increases
could not be sustained any longer because of fundamental limits on
heat dissipation. Thus began the era of multi-core CPUs, which offer
increased parallelism, but not clock-rate increases. That shift has
required rethinking algorithms and data structures, compilers, OS
kernels, memory consistency models, and even debugging and testing,
among other areas.

Babak will speak on the next round of scaling limitations that future
server-class CPUs will encounter, and how those limitations will
change the way CPUs look, including his own creative work on how to
ensure that future CPUs continue to offer performance increases
despite these limitations.

He will target his talk to a general CS audience--he will not be
assuming specialist knowledge of computer architecture.

All very strongly encouraged to attend. Again, this a topic that
undergirds what all of us do!

Title, abstract, and bio follow.

- -Brad, bkarp at cs.ucl.ac.uk

- ---------

UCL CS Distinguished Lecture

Speaker:

	Babak Falsafi
	Professor of Computer and Communication Sciences
	Director, EcoCloud Research Center
	EPFL School of Computer and Communication Sciences

Title:

	Big Data and Dark Silicon: Taming Two IT Inflection Points on a
Collision Course

Time and place:

	4th February 2014, 2 PM, MPEB 6.12

Abstract:

Information technology is now an indispensable pillar of a modern-day
society, thanks to the proliferation of digital platforms in the past
several decades. We are now witnessing two inflection points, however,
that are about to change IT as we know it. First, we are entering the
Big Data era where demand for robust and economical data processing,
communication and storage is growing faster than technology can
sustain. Second, while forecasts indicate that chip density scaling
will continue for another decade, the diminishing returns in supply
voltage scaling and the impending "energy wall" are leading server
designers towards energy-centric solutions, and ultimately Dark
Silicon. In this talk, I will motivate these two IT trends and present
promising research avenues for server chip design including
specialized scale-out processors, on-chip networks and die-stacked
DRAM cache hierarchies.

Bio:

Babak is Professor in the School of Computer and Communication
Sciences and the founding director of the EcoCloud research center
pioneering future energy-efficient and environmentally-friendly cloud
technologies at EPFL. He has made numerous contributions to computer
system design and evaluation including a scalable multiprocessor
architecture which was prototyped by Sun Microsystems (now Oracle),
snoop filters and temporal stream prefetchers that are incorporated
into IBM BlueGene/P and BlueGene/Q, and computer system simulation
sampling methodologies that have been in use by AMD and HP for
research and product development. His most notable contribution has
been to be first to show that contrary to conventional wisdom,
multiprocessor memory programming models -- known as memory
consistency models -- prevalent in all modern systems are neither
necessary nor sufficient to achieve high performance. He is a
recipient of an NSF CAREER award, IBM Faculty Partnership Awards, and
an Alfred P. Sloan Research Fellowship. He is a fellow of IEEE.



-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.22 (FreeBSD)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iEYEARECAAYFAlLwwE0ACgkQNz6hPDTA3IEmNACfcJeh6G1urdjFafbL5n7Q8A4J
ovUAn1vphaZHEfR6vsG/aBFN7eTzehbz
=j2Ah
-----END PGP SIGNATURE-----