Distributed Computing Economics Jim Gray Microsoft Research gray@microsoft.com Presentation To Microsoft Venture Capital Summit 28 April 2004 Distributed Computing Economics Why is Seti@Home a great idea? Why is Napster a great deal? Why is the Computational Grid uneconomic? When does computing on demand work? What is the “right” level of abstraction? Is the Access Grid the real killer app? Based on: Distributed Computing Economics, Jim Gray, Microsoft Tech report, March 2003, MSR-TR-2003-24 http://research.microsoft.com/research/pubs/view.aspx?tr_id=655 Computing Is Free Computers cost 1k$ (if you shop right) (yes, there are 1μ$ to 1M$ computers, but..) So 1 cpu day = 1$ (computers last 3 years) If you pay the phone bill, internet bandwidth costs 50…500$/mbps/m (not including routers and management) So 1GB costs 1$ to send and 1$ to receive Caveat: All numbers rounded to nearest factor of 3. Why Is Seti@Home A Good Deal? Send 300 KB: Costs 3e-4$ User computes for ½ day: Benefit .5e-1$ ROI: 1500:1 Seti@Home The worlds most powerful computer 67 TF is sum of top 4 of Top 500 67 TF is 9x the number 2 system 67 TF more than the sum of systems 2...10 Seti@Home http://setiathome.ssl.berkeley.edu/totals.html Users Results received Total CPU time Floating Point Operations 26 April 2004 Total 5M 1.3 B Last 24 Hours 1,138 1,5 M 1.5 M years 5 E+21 flops 6 E+18 FLOPS/day 5 zeta flops 67 TeraFLOPs 1,199 years Why Was Napster A Good Deal? Send 5 MB costs 5e-3$ ½ a penny per song Both sender and receiver can afford it Same logic powers web sites (Yahoo!...) 1e-3$/page view advertising revenue 1e-5$/page view cost of serving web page 100:1 ROI Computing Equivalents 1$ buys 1 day of cpu time 4 GB (fast) ram for a day 1 GB of network bandwidth 1 GB of disk storage for 3 years 10 M database accesses 10 TB of disk access (sequential) 10 TB of LAN bandwidth (bulk) 10 KWhrs == 4 days of computer time Depreciating over 3 years, and there are about 1k days in 3 years. Some Consequences Beowulf networking is 10,000x cheaper than WAN networking factors of 105 matter The cheapest and fastest way to move Terabytes cross country is sneakernet 24 hours = 4 MB/s 50$ shipping vs 1,000$ wan cost Sending 10PB CERN data via network is silly: buy disk bricks in Geneva, fill them, ship them TeraScale SneakerNet: Using Inexpensive Disks for Backup, Archiving, and Data Exchange Jim Gray; Wyman Chong; Tom Barclay; Alex Szalay; Jan vandenBerg Microsoft Technical Report may 2002, MSR-TR-2002-54 http://research.microsoft.com/research/pubs/view.aspx?tr_id=569 Computational Grid Economics To the extent that computational grid is like Seti@Home or ZetaNet or Folding@home or…it is a great thing The extent that the computational grid is MPI or data analysis, it fails on economic grounds: move the programs to the data, not the data to the programs The Internet is not the cpu backplane An alternate reality: Nearly free networking Telcos go bankrupt and price=cost=0 Taxpayers pay your phone bill so price=0 and telcos receive a BIG government subsidy When To Export A Task IF instruction density > 100,000 instructions/byte AND remote computer is free (costs you nothing) THEN ROI > 0 ELSE ROI < 0 Computing On Demand Was called outsourcing/service bureaus in my youth. CSC and IBM did it It is not a new way of doing things: think payroll. Payroll is standard outsourced service Now Hotmail, Salesforce.com, Oracle.com,… Works for standard apps COD works for commoditized services Airlines outsource reservations. Banks outsource ATMs But Amazon, Amex, Wal-Mart, eTrade, eBay... Can’t outsource their core competence What’s The Right Abstraction Level For Internet Scale Distributed Computing? Disk block? File? Database? Application? No too low No too low No too low Yes, of course Blast search Google search Send/Get eMail Portals that federate astronomy archives (http://skyQuery.Net/) Web Services (.NET, EJB, OGSA) give this abstraction level Access Grid Q: What comes after the telephone? A: eMail? A: Instant messaging? Both seem retro: text & emotons Access Grid could revolutionize human communication But, it needs a new idea Q: What comes after the telephone? Supercomputers You Use Hotmail, Yahoo!, Google: ~10k servers Amazon, Barnes&Noble Expedia, Orbitz Dell, HP,… Service-oriented architectures Not computing on demand, but information on demand! Distributed Computing Economics Why is Seti@Home a great idea? Why is Napster a great deal? Why is the Computational Grid uneconomic When does computing on demand work? What is the “right” level of abstraction? Is the Access Grid the real killer app? Based on: Distributed Computing Economics, Jim Gray, Microsoft Tech report, March 2003, MSR-TR-2003-24 http://research.microsoft.com/research/pubs/view.aspx?tr_id=655 Poll Is there a market for Supercomputers? Yes, Google, Expedia, Hotmail,… Is Computing On Demand a highmargin business? I think not Do you know the equivalent highmargin business? Information on demand Take Aways Computing on demand is a service business; probably not high margin; questionable economics; think LoudCloud Distributed computing is coming, but it is probably via Service Oriented Architecture (SOA) Web Services is the way to do SOA Outline Overview of Microsoft Research Distribute Computing Economics Q&A The Cost Of Computing Computers are NOT free! IBM, HP, Dell make $billions Capital Cost of a TpcC system is mostly storage and storage software (database) IBM 32 cpu, 512 GB ram 2,500 disks, 43 TB TpcC Cost Components DB2/AIX http://www.tpc.org/results/individual_results/IBM /IBM p690es_05092003.pdf software 10% storage 61% (680,613 tpmC @ 11.13 $/tpmc available 11/08/03) http://www.tpc.org/results/individual_results/IBM/IBMp690es_05092003.pdf A 7.5M$ super-computer Total Data Center Cost: 40% capital & facilities 60% staff (includes app development) cpu/mem 29%