From something that fits in your pocket ... ... to, well, this. The - PowerPoint PPT Presentation

From something that fits in your pocket ...

... to, well, this.

The future? ...

Energy A look at cluster computers and datacenters Tarun Prabhu, Radha Venkatagiri

Datacenters

Datacenters Most (all?) of you probably know what they are Most (all?) of you know what they are used for

Energy usage in datacenters Used 76,000,000,000 kWH in 2010 2% of all electricity produced in the US ≈ 1.3% of all electricity produced globally † † Koomey, J. “Growth in Data Center Electricity Use 2005 to 2010”, Analytics Press, 2011

Increasing datacenter efficiency Reduce infrastructure overheads Reduce ancillary (non-computing) costs Reduce computing costs

How efficient are these? PUE(Power Energy Effectiveness): indicates how much energy is used for non-computing functions. Average PUE is 1.8 (this means that for every 1 Watt used for computing, another 0.8 Watts is used in overheads) † Company PUE Comments Facebook 1.07 Google 1.14 Individual facility goes to 1.06 Yahoo – Individual facility goes to 1.08 Amazon 1.45 Assumption by Amazon themselves Microsoft 1.25 Target for April 2013 Apple – They shall never tell ... Table : Efficiency of datacenter giants ‡ † http://www.datacenterknowledge.com/archives/2011/05/10/uptime-institute-the-average-pue-is-1-8/ ‡ http://gigaom.com/cloud/whose-data-centers-are-more-efficient-facebooks-or-googles/

Infrastructure overheads - What are they?

Infrastructure overheads - Cooling, power etc. Heat management to reduce hot spots Natural cooling Air - Buffalo(Yahoo), Lulea(Facebook) Sea-water - Hamina(Google) Evaporative cooling - Prineville(Facebook) Optimize power distribution Efficient power-supplies Minimize AC/DC conversion stages Nifty new ideas, for instance oil baths

Ancillary costs ... What Facebook did

Ancillary costs ... What Facebook did Toss everything and go back to the drawing board.

Ancillary costs ... What Facebook did Toss everything and go back to the drawing board. Literally

Open Compute Project - motherboards Facebook custom-designed ... power supplies everything † - - server chassis Kept only what was strictly - server racks necessary - battery cabinets 38% more efficient, 24% cheaper Made all specifications (CAD drawings etc.) publicly available http://www.opencompute.org † https://www.facebook.com/notes/facebook-engineering/building-efficient-data-centers-with-the-open- compute-project/10150144039563920

Reducing computing costs

Reducing computing costs Tackle under-utilization and overprovisioning

Server utilization Average server utilization in datacenters is ≈ 50%

Reasons for under-utilization Planning for traffic spikes Reliability considerations System-software maintenance is safer

Real reason Clients get cranky!

Energy-proportional computing Under-utilization is a problem because, as things stand today, power consumed is not proportional to work done Ideally, the dynamic range of energy consumption should be increased. In this, no power will be consumed when idle, little power will be consumed when doing minimal work and the consumption would increase gradually until the machine is fully loaded

Reducing compute costs - I Tackling under-utilization with operating system support Turn/off suspend hosts during low-usage periods Intelligent load-balancing Resource-aware scheduling Power-aware scheduling

Google’s warehouse computing Google’s approach to building datacenters Treat entire datacenter as one BIG computer Centralized resource management. Provides greater flexibility in decision-making to improve metrics

Reducing compute costs - II Customizing hardware to applications

An example - FAWN Fast Array of Wimpy Nodes Single-core AMD Geode processor(500 Mhz) 256 MB DDR SDRAM (400 Mhz) 4GB CompactFlash storage Intel Atom front-end

Academic research Jointly optimize computing and cooling energy (ICDCS ’12) Data-centric approaches by focusing on where to place data to minimize energy consumption (SC ’12) Improving network and interconnect efficiency by scaling network up and down based on traffic demands (USENIX ’10) Intelligent allocation of work to compute-units based on job characteristics, environmental conditions etc.

Supercomputers

Supercomputers Tens of thousands (or more) of compute elements operating together TB’s (now PB’s) of memory PB’s (nearing EB’s of storage)

Uses of supercomputers Molecular dynamics Fluid dynamics: Airframe design Modelling astrophysics phenomena Earthquake system science Simulation of spread of contagion Cosmology (formation of the first galaxies) Climate modelling and hypothesis confirmation

Uses of supercomputers Molecular dynamics Fluid dynamics: Airframe design Modelling astrophysics phenomena Earthquake system science Simulation of spread of contagion Cosmology (formation of the first galaxies) Climate modelling and hypothesis confirmation (of global warming perhaps)

Top 500 List

Top 500 List # Name Location Cores PFLOPS Power(MW) 1 Titan USA 560K* 17.59 8.21 2 Sequoia USA 1572K 16.32 7.89 3 K Computer Japan 705K 10.5 12.66 4 Mira USA 786K 8.16 3.95 5 JuQueen Germany 131K 4.14 1.97 6 SuperMUC Germany 147K 2.89 3.42 7 Stampede USA 204K* 2.66 8 Tianhe-1A China 186K 2.56 4.04 9 Fermi Italy 163K 1.72 0.82 10 DTS USA 63K 1.51 3.57 Table : Power consumption of world’s fastest computers http://www.top500.org/list/2012/06/100

How efficient is this? These machines can simulate a rat’s brain

How efficient is this? WARNING: Some math here

How efficient is this? WARNING: Some Bad math here

How efficient is this? WARNING: Some Bad math here Brain weight comparison = 1400 gms † W human W rat = 2 gms † 30 W ‡ P human ≈ † http://faculty.washington.edu/chudler/facts.html

How efficient is this? WARNING: Some Bad math here Brain weight comparison = 1400 gms † W human W rat = 2 gms † 30 W ‡ P human ≈ 2 ∴ P rat brain ≈ 1400 × 30 = 0 . 043 W † http://faculty.washington.edu/chudler/facts.html

How efficient is this? WARNING: Some Bad math here Brain weight comparison Metabolism fraction W rat = 1400 gms † W human P rat ≈ × P human W human W rat = 2 gms † 0 . 4 = 62 × 100 §‡ 30 W ‡ P human ≈ 2 = 0 . 64 W ∴ P rat brain ≈ 1400 × 30 = 0 . 043 W † http://faculty.washington.edu/chudler/facts.html ‡ http://hypertextbook.com/facts/2001/JacquelineLing.shtml

How efficient is this? WARNING: Some Bad math here Brain weight comparison Metabolism fraction W rat = 1400 gms † W human P rat ≈ × P human W human W rat = 2 gms † 0 . 4 = 62 × 100 §‡ 30 W ‡ P human ≈ 2 = 0 . 64 W ∴ P rat brain ≈ 1400 × 30 P rat brain = 0 . 05 × P rat = 0 . 043 W = 0 . 032 W † http://faculty.washington.edu/chudler/facts.html ‡ http://hypertextbook.com/facts/2001/JacquelineLing.shtml

How efficient is this? WARNING: Some Bad math here Brain weight comparison Metabolism fraction W rat = 1400 gms † W human P rat ≈ × P human W human W rat = 2 gms † 0 . 4 = 62 × 100 §‡ 30 W ‡ P human ≈ 2 = 0 . 64 W ∴ P rat brain ≈ 1400 × 30 P rat brain = 0 . 05 × P rat = 0 . 043 W = 0 . 032 W † http://faculty.washington.edu/chudler/facts.html ‡ http://hypertextbook.com/facts/2001/JacquelineLing.shtml

How efficient is this? WARNING: Some Bad math here Brain weight comparison Metabolism fraction W rat = 1400 gms † W human P rat ≈ × P human W human W rat = 2 gms † 0 . 4 = 62 × 100 §‡ 30 W ‡ P human ≈ 2 = 0 . 64 W ∴ P rat brain ≈ 1400 × 30 P rat brain = 0 . 05 × P rat = 0 . 043 W = 0 . 032 W † http://faculty.washington.edu/chudler/facts.html ‡ http://hypertextbook.com/facts/2001/JacquelineLing.shtml

How efficient is this? WARNING: Some Bad math here Brain weight comparison Metabolism fraction W rat = 1400 gms † W human P rat ≈ × P human W human W rat = 2 gms † 0 . 4 = 62 × 100 §‡ 30 W ‡ P human ≈ 2 = 0 . 64 W ∴ P rat brain ≈ 1400 × 30 P rat brain = 0 . 05 × P rat = 0 . 043 W = 0 . 032 W † http://faculty.washington.edu/chudler/facts.html ‡ http://hypertextbook.com/facts/2001/JacquelineLing.shtml § http://www.biomedcentral.com/1471-2458/12/439

One of these ...

is equivalent to ...

is equivalent to ...

Exascale?

Exascale? Enough computing power to simulate the human brain (2019?)

Exascale? Needs 700 MW or more?

Exascale? Needs 700 MW or more? http://farm5.staticflickr.com/4011/4710638282 5e226f00f6.jpg

Exascale? Needs 700 MW or more? http://farm5.staticflickr.com/4011/4710638282 5e226f00f6.jpg http://images4.wikia.nocookie.net/ cb20100331223557/simpsons/images/0/0c/Springfield Nuclear Power Plant 1.PNG

Exascale? Needs 700 MW or more? Typical nuclear power plant produces 400-1200MW http://farm5.staticflickr.com/4011/4710638282 5e226f00f6.jpg http://images4.wikia.nocookie.net/ cb20100331223557/simpsons/images/0/0c/Springfield Nuclear Power Plant 1.PNG