Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
World Community Grid Forums
Category: Community Forum: Chat Room Thread: Google to Donate 1 Billion Core Hours to Research |
No member browsing this thread |
Thread Status: Active Total posts in this thread: 28
|
Author |
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I see this as more of an alternative to WCG than a sponsor of WCG, but only WCG could say. I don't see why a company as big as IBM would need or want help from a bloated search engine, other than with the thread search facility An alternative, a.k.a. a (potential) competitor to WCG in the distributed computing space. 1-billion core-hours is approx equal to: - 1-billion cores running for 1-hour; - 416.7-million cores for 24hrs (1-day); - 59.5-million cores for 7-days (1-week); - 14.8-million cores for 28-days (4-weeks); - 2.3-million cores for 180-days (~6-months); - 1.15-million cores for 360-days (~1-year); Assuming: number of devices = number of cores; globalStats lastUpdated 2011.04.23Sa.1205UTC: number of devices = 1,758,291; we have WCG's output at roughly: - 1.7-million cores for 245days to match 1-billion core-hours. If Google throws in 2.3-million cores, WCG would be left behind by about (245days-180days) or 65days. However, lots of variables here. For one, how much performance per core is Google talking about? And WCG may have more cores than devices and/or more powerful cores. ; ------------ ;Edit1: I pressed that submit button accidentally !?#@ ;Edit2: Added some math/calculations ; [Edit 2 times, last edit by Former Member at Apr 23, 2011 9:19:43 PM] |
||
|
sk..
Master Cruncher http://s17.rimg.info/ccb5d62bd3e856cc0d1df9b0ee2f7f6a.gif Joined: Mar 22, 2007 Post Count: 2324 Status: Offline Project Badges: |
alternative
≠competitor |
||
|
GIBA
Ace Cruncher Joined: Apr 25, 2005 Post Count: 5374 Status: Offline |
I see this as more of an alternative to WCG than a sponsor of WCG, but only WCG could say. I don't see why a company as big as IBM would need or want help from a bloated search engine, other than with the thread search facility An alternative, a.k.a. a (potential) competitor to WCG in the distributed computing space. 1-billion core-hours is approx equal to: - 1-billion cores running for 1-hour; - 416.7-million cores for 24hrs (1-day); - 59.5-million cores for 7-days (1-week); - 14.8-million cores for 28-days (4-weeks); - 2.3-million cores for 180-days (~6-months); - 1.15-million cores for 360-days (~1-year); Assuming: number of devices = number of cores; globalStats lastUpdated 2011.04.23Sa.1205UTC: number of devices = 1,758,291; we have WCG's output at roughly: - 1.7-million cores for 245days to match 1-billion core-hours. If Google throws in 2.3-million cores, WCG would be left behind by about (245days-180days) or 65days. However, lots of variables here. For one, how much performance per core is Google talking about? And WCG may have more cores than devices and/or more powerful cores. ; ------------ ;Edit1: I pressed that submit button accidentally !?#@ ;Edit2: Added some math/calculations ; I bet that the numbers provided above by The Aspens, based on WCG average historical statistics, are more close to express the real WCG production, that is 12.8 % more powerful than Google initiative (if we consider that are comparable cores/performances on the Google side, what I doubt due the challenges, the learning curve of any project and the real aspects involved to put it on the right trail at full speed !) take a look at: Hmmm - how does this compare to what we all are contributing to research through World Community Grid right now: Average of 353 years/day = 3,092,280 hours/day 365 days/year * 3,092,280 hours/day = 1,128,682,200 1.1 Billion Core Hours/year!!!!!!!! Its good to see more people get involved.
Cheers ! GIB@
Join BRASIL - BRAZIL@GRID team and be very happy ! http://www.worldcommunitygrid.org/team/viewTeamInfo.do?teamId=DF99KT5DN1 |
||
|
GIBA
Ace Cruncher Joined: Apr 25, 2005 Post Count: 5374 Status: Offline |
alternative ≠competitor Agree.
Cheers ! GIB@
Join BRASIL - BRAZIL@GRID team and be very happy ! http://www.worldcommunitygrid.org/team/viewTeamInfo.do?teamId=DF99KT5DN1 |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Competition is good for society, by the way, and so are alternatives. But my home team is WCG. However, I can see some weaknesses in the way WCG plays the distributed computing game. Without competition or the threat of one, these weaknesses does not matter at all else we can afford to be in denial and next label those weaknesses as strengths. With competition, these weaknesses are things that winners successfully rectified while the competition did not -- and from which losers get pounded. Let's take a close look at where Google is trying to steer its ship. What better view than the look closest to the metal:
The best projects will have a very high number of independent work units, a high CPU to I/O ratio, and no inter-process communication (commonly described as Embarrassingly or Pleasantly Parallel). The higher the CPU to I/O rate, the better the match with the system. Programs must be developed in C/C++ and compiled via Native Client. Awardees will be able to consult an on-site engineering team. WCG has one project that crashes right into one area where Google declared it is aiming at. The WCG project CleanEnergyProject-Phase2 (CEP2) is an I/O-intensive project that smacks right into whatever Google may have been thinking in coming up with the formulation, "The higher the CPU to I/O rate, the better the match with the system." Either Google is misguided about what it sees as the criteria for projects to be best suited for crunching (a.k.a gridComputing, parallelComputing, distributedComputing), or WCG should have known better than accept that CEP2 project. ; |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I dont know if we need any improvement from Google--they tend to take over or control projects--ediit--i really don't trust their motives Point 1. When you depend on volunteers and someone volunteers, make sure you turn down their offer.Point 2. Every one except the ostrich knows that IBM has far superior project management resources and vastly more experience than Google. Point 3. When considering the largest donation of computer time ever proposed by a single organization, be sure to take advice from people who have never worked for IBM, WCG or Google. Point 4. WCG is an IBM project and IBM sets the rules, not Google. |
||
|
TimAndHedy
Senior Cruncher Joined: Jan 27, 2009 Post Count: 267 Status: Offline Project Badges: |
At completion of the project, we recommend, but do not require, that all the researcher's data be made freely available to the academic community. They may get a few projects that would not fly here. I for one would not work on any project that is not freely available. They do however have the staff to vet any work on their systems. |
||
|
TimAndHedy
Senior Cruncher Joined: Jan 27, 2009 Post Count: 267 Status: Offline Project Badges: |
Point 3. When considering the largest donation of computer time ever proposed by a single organization, Lets hope this inspires IBM to beat them. |
||
|
|