Energy-Efficient HPC A Tools Perspective February 2nd, 2016 - - PowerPoint PPT Presentation
Mitglied der Helmholtz-Gemeinschaft Energy-Efficient HPC A Tools Perspective February 2nd, 2016 Michael Knobloch Short Intro to Performance Tools Trigger performance data Sampling Code instrumentation Recording performance data Profiling
Mitglied der Helmholtz-Gemeinschaft
February 2nd, 2016 Michael Knobloch
Trigger performance data
Sampling Code instrumentation
Recording performance data
Profiling Tracing
February 2nd, 2016 Michael Knobloch Slide 2
February 2nd, 2016 Michael Knobloch Slide 3
February 2nd, 2016 Michael Knobloch Slide 4
February 2nd, 2016 Michael Knobloch Slide 5
February 2nd, 2016 Michael Knobloch Slide 6
February 2nd, 2016 Michael Knobloch Slide 7
February 2nd, 2016 Michael Knobloch Slide 8
February 2nd, 2016 Michael Knobloch Slide 9
February 2nd, 2016 Michael Knobloch Slide 10
February 2nd, 2016 Michael Knobloch Slide 11
eeClust - Energy-Efficient Cluster Computing
Project partners: Uni Hamburg, TU Dresden, JSC, ParTec www.eeclust.de Goals: Identify phases of low resource utilization and turn hardware to lower power-states in such phases Integral point: Extension of performance analysis tools to analyse application power consumption and hardware utilization
February 2nd, 2016 Michael Knobloch Slide 12
MPI Busy-Waiting
Power consumption in phases of busy-waiting is very high due to constant CPU activity.
February 2nd, 2016 Michael Knobloch Slide 13
Instr. target application Measurement library Distributed trace Parallel wait-state search Wait-state report Report browser February 2nd, 2016 Michael Knobloch Slide 14
time processes A B C D Send Recv Send Recv
Wait State
February 2nd, 2016 Michael Knobloch Slide 15
Idle-Waiting
ESP = max
p∈PS((tw ∗ Ap1) − (tw − tTp,p1) ∗ Ip + ETp,p1)
Busy-Waiting
ESP BW = max
p∈PS((tw ∗ Ap1) − (tw − tTp,p1) ∗ Ap + ETp,p1)
PS – Set of power states tw – Waiting time Ap – Active energy in P-State p tTp1,p2 – Transition time Ip – Idle energy in P-State p ETp,p1 – Transition energy
February 2nd, 2016 Michael Knobloch Slide 16
February 2nd, 2016 Michael Knobloch Slide 17
February 2nd, 2016 Michael Knobloch Slide 18
EIC - Exascale Innovation Center
Project partners: IBM Germany R&D and JSC Goal: Co-Design for next-gen of Supercomputers One work-package on energy-efficiency Investigation of power consumption on Blue Gene Fine-grained power measurements on POWER7 Now in transition to PADC - POWER Acceleration and Design Center
February 2nd, 2016 Michael Knobloch Slide 19
Amester
IBM Automated Measurement of Systems for Temperature and Energy Reporting software. Results were published at EnA-HPC 2013.
Sensor name Units Time scale Description PWR1MS W Instantaneous Node power consumption PWR1MSP0 W Instantaneous Processor power consumption PWR1MSMEM0 W Instantaneous Memory power consumption PWR32MS W
Node power consumption PWR32MSP0 W
Pocessor power consumption PWR32MSMEM0 W
Memory power consumption IPS32MS Mips Every 32 ms Instructions per second rate
February 2nd, 2016 Michael Knobloch Slide 20
February 2nd, 2016 Michael Knobloch Slide 21
February 2nd, 2016 Michael Knobloch Slide 22
20 40 60 80 100
Time (s)
50 100 150 200
Power (W)
Disk IO Memory GPU Fan CPU
February 2nd, 2016 Michael Knobloch Slide 23
Score-E
Main Tools Partners: JSC, TU Dresden, TU Munich Successor of SILC and LMAC Extension of Score-P measurement system (www.score-p.org)
Common measurement system for Scalasca, Vampir, and Periscope
Power and Energy measurements from different sources, e.g. RAPL, Xeon Phi/GPU power sensors, etc. Energy modelling from power consumption data Enable auto-tuning for energy efficiency New visualization based on application geometries
February 2nd, 2016 Michael Knobloch Slide 24
Metric discussion
Still on the quest for the right metric Power vs. Energy
Might require different analyses
User motivation (still) hard
Tools need
Sensors that provide relevant information
Power, energy, temperature, etc.
At all relevant system levels Scalable APIs
February 2nd, 2016 Michael Knobloch Slide 25
Challenges for tools
How reliable is performance data from dynamic applications? How reliable is performance data on dynamic hardware? How to analyse dynamic applications?
Requirements on tools
?
February 2nd, 2016 Michael Knobloch Slide 26