Top Banner
This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme [FP7/2007-2013] under grant agreement n° 288777. http://www.montblanc-project.eu Exploiting mobile phone technology to build energy efficient supercomputers: the Mont Blanc project Simon McIntosh-Smith, University of Bristol, UK Disclaimer: Speaking for myself ... All references to unavailable products are speculative, taken from web sources. There is no commitment from ARM, Samsung, TI, Nvidia, Bull, or others, implied.
17

Www.montblanc-project.eu This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme.

Dec 15, 2015

Download

Documents

Aleah Hubbart
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Www.montblanc-project.eu This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme.

This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme [FP7/2007-2013] under grant agreement n° 288777.

http://www.montblanc-project.eu

Exploiting mobile phone technology to build energy efficient supercomputers:

the Mont Blanc project

Simon McIntosh-Smith, University of Bristol, UK

Disclaimer: Speaking for myself ... All references to unavailable products are speculative, taken from web sources.There is no commitment from ARM, Samsung, TI, Nvidia, Bull, or others, implied.

Page 2: Www.montblanc-project.eu This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme.

Motivation: from MegaFLOPS to ExaFLOPS

• High Performance Computing (HPC) systems are now a common part of large-scale research infrastructure

• These systems have been growing larger and larger, and now themselves can consume considerable energy• Today we have systems capable of performing O(1015) floating

point operations per second, and we're working towards O(1018)

• Power consumption has grown by two orders of magnitude in the last two decades, from hundreds of kW in the '90s to tens of MWs today

• This is starting to cause PROBLEMS!

Page 3: Www.montblanc-project.eu This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme.

Power Consumption growing in the Top500

2008 2009 2010 2011 2012 20130

1

2

3

4

5

6

7

8

Po

wer

[M

W]

TOP10

TOP50

TOP5003.25 x in 5 y

3.13 x in 5 y

5.04 x in 5 y

Graph courtesy of Erich Strohmaier, Lawrence Berkeley National Laboratory

http://top500.org

Page 4: Www.montblanc-project.eu This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme.

Providing power is not the problem

• We can provide tens of MWs if they're really needed…

Page 5: Www.montblanc-project.eu This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme.

kWh cost for Medium size industries (€, 2011)

• The HPC power wall isn’t due to power availability• We could provide 500 MW to an HPC facility

• It’s the electricity bill at the end of the month! Also CO2…• 1 MWatt ~ 1 M EUR / year

Page 6: Www.montblanc-project.eu This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme.

The problem … and the opportunity

• Europe represents ~40% of the HPC market• Yet, it does not have HPC technology of its own

• Nobody yet knows how to build a sustainable EFLOPS supercomputer• Power defines performance• Consensus about it being revolutionary, not evolutionary

• Europe is very strong in embedded computing• The most energy-efficient computing technology today

Page 7: Www.montblanc-project.eu This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme.

Mont-Blanc project goals

• To develop a European Exascale approach • Leverage commodity and embedded power-efficient

technology

• Supported by two EU projects so far, 25.5M€ total:

• FP7 Objective ICT-2011.9.13 Exascale computing, software and simulation:• 3-year IP Project (October 2011 - September 2014)• Total budget: 14.5 M€ (8.1 M€ EC contribution),

• FP7 Objective ICT-2013.12.1 Exascale computing platforms, software and applications:• 3-year IP Project (October 2013 - September 2016)• Total budget: 11.5 M€ (8.0 M€ EC contribution),

Page 8: Www.montblanc-project.eu This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme.

Mont-Blanc 2 consortium

Page 9: Www.montblanc-project.eu This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme.

First, vector processors dominated HPC

• 1st Top500 list (June 1993) dominated by traditional HPC (vector) architectures• Cray vector,41%• MasPar SIMD, 11%• Convex/HP vector, 5%

• Fujitsu Wind Tunnel is #1 1993-1996, with 170 GFLOPS

Page 10: Www.montblanc-project.eu This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme.

Then, commodity displaced special purpose

• ASCI Red, Sandia• 1997, 1 TFLOPS• 9,298 cores @ 200 MHz• Intel Pentium Pro

• Upgraded to Pentium II Xeon, 1999, 3.1 TFLOPS

• ASCI White, LLNL• 2001, 7.3 TFLOPS• 8,192 proc. @ 375 MHz, • IBM Power 3

Transition from Vector parallelism to Message-Passing Programming Models

Page 11: Www.montblanc-project.eu This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme.

11

The killer microprocessors

• Microprocessors killed the Vector supercomputers• They were not faster ...• ... but they were significantly cheaper (and more energy efficient)

• Needed 10 microprocessors to achieve the performance of 1 Vector CPU – but this was still a win!

Cray-1, Cray-C90

NEC SX4, SX5

Alpha AV4, EV5

Intel Pentium

IBM P2SC

HP PA8200

1974 1979 1984 1989 1994 199910

100

1000

10.000

MF

LO

PS

Page 12: Www.montblanc-project.eu This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme.

The killer mobile processorsTM

• Microprocessors killed the Vector supercomputers• They were not faster ...• ... but they were significantly

cheaper and greener

• History may be about to repeat itself …• Mobile processors are not

faster …• … but they are significantly

cheaper (and greener?)

Alpha

Intel

AMD

NVIDIA Tegra

Samsung Exynos

4-core ARMv8 1.5 GHz

1990 1995 2000 2005 2010100

1.000

10.000

100.000

MF

LO

PS

2015

1.000.000

Page 13: Www.montblanc-project.eu This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme.

Mont-Blanc roadmap to Exascale

• The goal is for Europe to become a serious force in the HPC Exascale race• Mont-Blanc NG system by 2017, 200 PFLOPS on 10 Mwatt• Mont-Blanc EX system by ~2020, 1 EFLOPS on 20 MWatt

2011 2012 2013 2014 2015 2016

Integrated ARM+ Accelerator

200 PFLOPS10 MWatt

ARM Multicore+ Discrete GPU

2017 2018 2019 2020

ARMMulticore

1 EFLOPS20 MWatt

1018 FLOPS

GF

LOP

S / W

att

Page 14: Www.montblanc-project.eu This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme.

Tibidabo: The first ARM HPC multicore cluster

• Proof of concept• It is possible to deploy a cluster of smartphone processors

• Enable software stack development

Q7 carrier board2 x Cortex-A92 GFLOPS1 GbE + 100 MbE7 Watts0.3 GFLOPS / W

Q7 Tegra 22 x Cortex-A9 @ 1GHz2 GFLOPS5 Watts (?)0.4 GFLOPS / W

1U Rackable blade8 nodes16 GFLOPS65 Watts0.25 GFLOPS / W

2 Racks32 blade containers 256 nodes 512 cores

9x 48-port 1GbE switch

512 GFLOPS3.4 Kwatt0.15 GFLOPS / W

Page 15: Www.montblanc-project.eu This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme.

Samsung-based Mont Blanc Daughter Card

• Compute card based on the Samsung Exynos 5 Dual mobile phone chip

• Each daughter card is a full HPC compute node• (CPU + GPU) + DRAM

+ NAND storage + Ethernet NIC

Page 16: Www.montblanc-project.eu This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme.

: Prototype architecture

Carrier blade15 x Compute cards485 GFLOPS1 GbE to 10 GbE200 Watts (?)2.4 GFLOPS / W

Exynos 5 Compute card1x Samsung Exynos 5 Dual 2 x Cortex-A15 @ 1.7GHz 1 x Mali T604 GPU6.8 + 25.5 GFLOPS (peak)~10 Watts3.2 GFLOPS / W (peak)

7U blade chassis9 x Carrier blade 135 x Compute cards4.3 TFLOPS2 KWatt2.2 GFLOPS / W

1 Rack4 x blade cabinets 36 blades 540 compute cards

2x 36-port 10GbE switch 8-port 40GbE uplink

17.2 TFLOPS (peak)8.2 KWatt2.1 GFLOPS / W (peak)

80 Gb/s

Page 17: Www.montblanc-project.eu This project and the research leading to these results has received funding from the European Community's Seventh Framework Programme.

Conclusions

• The bigger the science, the larger the supporting infrastructure!

• Our HPC systems have grown to tens of MWs to process all the data we are generating

• The processors developed for the mobile device space might yield a significant jump in energy efficiency for computation

• The Mont Blanc project is exploring Europe's strengths in mobile technology to exploit the opportunity arising in energy-dominated HPC

http://www.montblanc-project.eu