UTCS Faculty Candidate - Michael Ferdman/Carnegie Mellon University, "The Future of Cloud-Server Architecture: Looking at the end of multi-core and beyond," ACES 2.302

Contact Name: 
Jenna Whitney
Date: 
May 3, 2012 11:00am - 12:00pm

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: Michael Ferdman/Ca

rnegie Mellon University

Talk Audience: UTCS Faculty, Graduate Studen

ts, Undergraduate Students, and Outside Interested Parties

Date/Time

: Thursday, May 3, 2012, 11:00 am

Location: ACES 2.302

Host: E

mmett Witchel

Talk Title: The Future of Cloud-Server Architecture: Loo

king at the end of multi-core and beyond

Talk Abstract:
Cloud comput

ing has emerged as a dominant platform for delivering online services and r

esulted in a constantly-growing demand for datacenter performance. Followin

g Moore‚s Law, popular belief holds that the number of cores on chip wil

l grow at an exponential rate, leading to a commensurate increase in serve

r performance. However, while demand for cloud infrastructure continues t

o grow, the semiconductor manufacturing industry has reached physical limi

ts, unable to improve performance due to rising chip power. Continuing to
increase performance of the cloud while staying within physical constraint

s therefore mandates optimizing server efficiency. In this talk, I will s

ummarize our recent work on quantifying the inefficiencies of cloud softwar

e on modern processors and describe Proactive Instruction Fetch, a techniq

ue to address one of the dominant inefficiencies that we identified. I wil

l then conclude the talk by describing a projection of server architecture

in future technologies, where an increasing fraction of the chip is dark s

ilicon that we cannot afford to power.

Speaker Bio:
Mike Ferdman is

a Ph.D. candidate at Carnegie Mellon University. His research interests ar

e in the area of computer architecture, with emphasis on the design of ser

ver systems. Mike‚s primary research objective is to understand the inter

actions of software and processor microarchitecture to enable the design of
high-performance, power-efficient, and compact servers.