Program - ISMM 2019

Filter

Sun 23 Jun

09:00 - 11:00: ISMM 2019 - Scaling Up at 106A

09:00 - 09:05
Day opening

Welcome from the chairs

Harry XuUniversity of California, Los Angeles (UCLA), Jeremy SingerUniversity of Glasgow

09:05 - 09:45
Talk

Keynote 1: Relaxed memory ordering needs a better specification

Hans-J. BoehmGoogle

09:45 - 10:10
Talk

Automatic GPU Memory Management for Large Neural Models in TensorFlow

Tung D. LeIBM Research - Tokyo, Haruki ImaiIBM Research - Tokyo, Yasushi NegishiIBM Research - Tokyo, Kiyokuni KawachiyaIBM Research - Tokyo

10:10 - 10:35
Talk

Massively Parallel GPU Memory Compaction

Matthias SpringerTokyo Institute of Technology, Hidehiko MasuharaTokyo Institute of Technology

10:35 - 11:00
Talk

Scaling Up Parallel GC Work-Stealing in Many-Core Environments

Michihiro HorieIBM Research - Tokyo, Kazunori OgataIBM Research, Japan, Mikio TakeuchiIBM Research - Tokyo, Hiroshi HoriiIBM Research, Japan

11:20 - 12:35: ISMM 2019 - Exotica at 106A

11:20 - 11:45
Talk

Exploration of Memory Hybridization for RDD Caching in Spark

Md Muhib KhanFlorida State University, Muhammad Ahad Ul AlamFlorida State University, USA, Amit Kumar NathFlorida State University, USA, Weikuan YuFlorida State University, USA

11:45 - 12:10
Talk

Learning When to Garbage Collect with Random Forests

Nicholas JacekUMass Amherst, Eliot MossUniversity of Massachusetts Amherst

12:10 - 12:35
Talk

Timescale Functions for Parallel Memory Allocation

Pengcheng LiGoogle, Inc, Hao LuoUniversity of Rochester, Chen DingUniversity of Rochester

14:00 - 15:40: ISMM 2019 - Mechanics at 106A

14:00 - 14:50
Talk

Keynote 2: Why do big data and cloud systems stop (slow down)?

Shan LuUniversity of Chicago

14:50 - 15:15
Talk

Concurrent Marking of Shape-Changing Objects

Ulan DegenbaevGoogle, Michael LippautzGoogle, Hannes PayerGoogle

15:15 - 15:40
Talk

A Lock-Free Coalescing-Capable Mechanism for Memory Management

Ricardo LeiteUniversity of Porto, Ricardo RochaUniversity of Porto, Portugal

16:00 - 17:15: ISMM 2019 - Mechanics / Message Passing at 106A

16:00 - 16:25
Talk

snmalloc: A Message Passing Allocator

Paul Lietar, Theodore ButlerDrexel University, USA, Sylvan ClebschImperial College London, Sophia DrossopoulouImperial College London, Juliana FrancoMicrosoft Research, Cambridge, Matthew ParkinsonMicrosoft Research, UK, Alex ShamisMicrosoft Research / Imperial College London, Christoph M. WintersteigerMicrosoft Research, UK, David ChisnallUniversity of Cambridge

16:25 - 16:50
Talk

Design and Analysis of Field-Logging Write Barriers

Steve BlackburnAustralian National University

16:50 - 17:14
Talk

Gradual Write-Barrier Insertion into a Ruby Interpreter

Koichi SasadaCookpad, Japan

Link to publication DOI

17:14 - 17:15
Day closing

Final Remarks

Harry XuUniversity of California, Los Angeles (UCLA), Jeremy SingerUniversity of Glasgow

Sun 23 Jun 2019


Room	9:00						30						10:00						30						11:00						30						12:00						30						13:00						30						14:00						30						15:00						30						16:00						30						17:00						30
106A	ISMM Scaling Up 09:00 - 11:00																												ISMM Exotica 11:20 - 12:35																																ISMM Mechanics 14:00 - 15:40																								ISMM Mechanics / Message Passing 16:00 - 17:15

Sun 23 Jun 2019


Room	9:00			15			30			45			10:00			15			30			45			11:00			15			30			45			12:00			15			30			45			13:00			15			30			45			14:00			15			30			45			15:00			15			30			45			16:00			15			30			45			17:00			15			30			45
106A	ISMM Welcome from the chairs 09:00 - 09:05	ISMM Keynote 1: Relaxed memory ordering needs a better specification 09:05 - 09:45								ISMM Automatic GPU Memory Management for Large Neural Models in TensorFlow 09:45 - 10:10					ISMM Massively Parallel GPU Memory Compaction 10:10 - 10:35					ISMM Scaling Up Parallel GC Work-Stealing in Many-Core Environments 10:35 - 11:00									ISMM Exploration of Memory Hybridization for RDD Caching in Spark 11:20 - 11:45					ISMM Learning When to Garbage Collect with Random Forests 11:45 - 12:10					ISMM Timescale Functions for Parallel Memory Allocation 12:10 - 12:35																						ISMM Keynote 2: Why do big data and cloud systems stop (slow down)? 14:00 - 14:50										ISMM Concurrent Marking of Shape-Changing Objects 14:50 - 15:15					ISMM A Lock-Free Coalescing-Capable Mechanism for Memory Management 15:15 - 15:40									ISMM snmalloc: A Message Passing Allocator 16:00 - 16:25					ISMM Design and Analysis of Field-Logging Write Barriers 16:25 - 16:50					ISMM Gradual Write-Barrier Insertion into a Ruby Interpreter 16:50 - 17:14					ISMM Final Remarks 17:14 - 17:15

iCalendar subscription service for your personal schedule

You first need to have starred events () in order to use the iCalendar subscription service.

Currently Viewed Program

Downloads the currently viewed program (with filters applied) in iCal format

ISMM 2019 Program

Sun 23 Jun

Sun 23 Jun 2019

ISMM

ISMM

ISMM

ISMM

Sun 23 Jun 2019

Tracks