1. درس تجارت الکترونیک دوره کارشناسی ارشد مجازی http://javan.edumoot.com Volunteer Computing

2. Where's the power? ● 2010: 1 billion Internet-connected PCs, 55% privately owned ● If 100M people participate: – 100 PetaFLOPs, 1 Exabyte (10^18) storage ● Consumer products drive technology – GPUs (NVIDIA, Sony Cell) your computers academic business home PCs

3. Volunteer computing history 95 96 97 98 99 00 01 02 03 04 05 GIMPS, distributed.net SETI@home, folding@home fight*@home climateprediction.net volunteer computing public [resource] computing Internet computing screensaver computing global computing @home computing peer-to-peer computing Grid computing BOINC Einstein@home

4. 4 محاسبات داوطلبانه ≠ محاسبات توري محاسبات توريمحاسبات داوطلبانه آريخيرمديريت منابع آريخيراتصال دائمي فراوانگرانقيمت و کميابپهناي باند شبکه دو راه (pull, push) يک راه (pull) اتصال به شبکه خيرآريمزاحمت ممکن استآريسيستم اعتبار بخشي پيچيده شيوه در اختيار گرفتن منابع

5. Projects and participants SETIphysics Climate biomedical Joe Alice Jens diversity, autonomy heterogeneity allocation, trust

6. Encourage participation in >1 project ● Better long-term resource utilization – project A works while project B thinks ● Better short-term resource utilization – communicate/compute in parallel – match applications to resources project computing needs think work think work time

7. Structure of a BOINC project Scheduling server (C++) BOINC DB (MySQL) Work generation data server (HTTP) Web interfaces (PHP) Retry generation Result validation Result processing Garbage collection Ongoing tasks: - monitor server correctness - monitor server performance - develop and maintain applications

8. Redundant computing ● Addresses hardware errors, hackers ● Issue 2 or more copies of each task – don't send to same host or user – timed retry up to a limit

9. Participating ● Select project(s) ● Create account(s) ● Download/install BOINC client software ● Interact via web: – preferences – leaderboards – profile – teams – message boards, dynamic FAQ

11. 11 شرايط داوطلب اجازة اجرا در زمان استفاده از باتري (مخصوص laptop) محدودة زماني مجاز تعداد پدازنده هاي قابل دسترسي درصد استفاده از ديسک حداکثر نرخ ارسال و دريافت

12. Client structure App Core client screensaver BOINC Manager servers

13. SETI@home ● Analysis of radio telescope data from Arecibo – SETI: search for narrowband signals – Astropulse: search for short broadband signals ● 0.3 MB in, ~4 CPU hours, 10 KB out ● Enhancements under BOINC: – data archival on clients – direct data distribution from observatory

14. 10/3/201514 SETI@home: >500,000 CPUs

16. Outline ● Volunteer computing ● BOINC: an OS for volunteer computing ● Applications ● Challenges and research directions

17. Climateprediction.net ● Climate change study (Oxford University) – Met Office model (FORTRAN, 1M lines) ● Input: ~10MB executable, 1MB data ● Output per workunit: – 10 MB summary (always upload) – 1 GB detail file (archive on client, may upload) ● CPU time: 2-3 months (can't migrate)

19. Biology projects ● Protein folding – Predictor@home (Scripps Institute) – Folding@home (Stanford) ● Virtual drug discovery – fightAIDS@home ● Gene sequence analysis – NTT projects – Lattice (U. Maryland)

20. Einstein@home ● Gravitational wave detection; LIGO ● UW Milwaukee/CalTech/Max Planck Inst. ● 30,000 40 MB data sets ● Each data set is analyzed w/ 40,000 different parameter sets; each takes ~6 hrs CPU ● Locality scheduling – minimize data transfer, client disk usage – minimize credit-granting delay

22. CERN projects ● LHC@home – accelerator simulation (Sixtrack) ● HEP@home – collision data analysis ● Africa@home

23. Porting of MalariaControl.net to BOINC platform ● Project team involves 3 students from Geneva, Bamako and Yaoundé ● Funded by Geneva International Academic Network, hosted at CERN ● Port takes 3 months, beta-test February 2006, open to public July 2006 AFRICA@home: first project

24. Volunteers: 7000 total, 4500 active ● Sign up rate: up to 400 new users per day ● Currently 50-60 per day Host PCs: 20,000 total, 15,000 active, ● 80% Windows, 20% Linux AFRICA@home: first results for MalariaControl.net CPU power: 3.0 Teraflops equivalent to 1,000 CPU years/yr (midrange PCs) delivered to date 1,800 CPU years (Apr 07) Simulations per day: 35,000 …+ huge public/press interest!

25. Others ● UCB Internet measurement – Map/measure the Internet and home PCs ● BURP (big ugly rendering project) – ray-tracing ● PlanetQuest – image analysis for planetary transit detection

26. Challenges and questions ● Get 100 million participants – simplified account management ● Get more projects ● Use GPUs and other resources ● Integrate with Grid (Lattice, CERN) ● …

27. Volunteer computing ● A new high-performance computing paradigm ● Benefits to projects: – enables otherwise infeasible computational research – economic advantage even for small projects ● Benefits to participants: – increase public scientific knowledge/interest – catalyze virtual communities – democratize resource allocation

28. Volunteer Computing as a Service BayanihanComputing.NET Volunteer computing

29. Idea #1: Volunteer Computing Many computers work together Client Workers Pool Service

30. Idea #2: Computation Service Simple interface to volunteer computing network Computation Client Workers Pool Service Computation Service Render(…) Bitmap

31. Idea #2: Computation Service Simple interface to volunteer computing network Computation Client Workers Pool Service Computation Service Render(…) Bitmap

32. Idea #2: Computation Service Simple interface to volunteer computing network Computation Client Virtual “Supercomputer”

33. Idea #2: Computation Service Computation Client Workers Pool Service Computation Service Web app Simple interface to volunteer computing network

34. New Computing Models

35. How It Works Monthly payment = $2.60 * # avg. simultaneous users * CPU % Step 1: The user starts playing your game and the game initiates Plura Step 2: The user's computer requests, computes, and returns WUs Step 3: Plura pays the developer for the computation

36. Integrate Plura into your Website

37. Integrate with Plura in Minutes...

38. Traditional Distributed Computing vs. Plura Processing

39. Potential applications ● Oil & gas exploration algorithms ● Financial modeling ● Bioinformatics ● Web crawling and Internet analysis ● Fluid dynamics ● And many others

40. Customers ● Our analysis requires making millions of comparisons for each pass of the data. Plura allows us to complete each pass within minutes, rather than the days or months it would take under traditional computing methods. ● Dr. Lin Ji, Quant R&D

41. Affiliates ● The parents at our school are very excited about Plura. Plura is a fantastic way for the school’s supporters to do something simple every day which generates funds to help us achieve our mission. ● David Steakley, The Rainard School

42. Plura vs Cloud vs Clusters