 |
 |
With manycore do you mean multicore? For multicore I see a lot of positive reason to put more work into it. The use of less memory has snowball effects not only on the amount of hardware we might buy but also on the possibility of using 64bit on more applications and 64bit applications are more efficient ie. higher HS06 values. Of course there is the problem of how to schedule the jobs but until there are more queues this will have little priority. Also I know experiments are at different stages for this. cheers alessandra On 13/09/2013 13:09, Wahid Bhimji wrote: > Hi > >> Is there any other technical activity that should be brought in to these meetings? This would include any technical activity that people would want in GridPP5. > Manycore and supercomputing ? (Andy won't be there so maybe covering in this meeting is not possible - but it is part of the technical landscape) . > > Generally 2 / 3 topics being cloud is probably too much focus on clouds. > > Wahid > > On 13 Sep 2013, at 10:51, David Colling <[log in to unmask]> wrote: > >> Hi Andrew et al, >> >> I have had a few private emails not to the list agreeing with this approach and if Andrew would find it difficult to make the meeting as well I think that clinches it. >> >> So could people prepare some summaries for next week about the different areas of technical activity - this could usefully be a few slides or even a text document with a few bullet points just to provide a record. >> >> Off the top of my head (and not very reliably as I am in a project review): the areas where I would expect a summary are: >> >> - The activity of the storage group (brief because this is large) >> - The state of the GridPP test cloud >> - Cloud activity from each of the experiments >> >> >> Is there any other technical activity that should be brought in to these meetings? This would include any technical activity that people would want in GridPP5. >> >> Best, >> david >> >> >> >> On 13/09/13 10:15, Andrew McNab wrote: >>> On 13 Sep 2013, at 07:27, David Colling <[log in to unmask]> wrote: >>>> We are due to re-start our technical meetings today. I am not available because of an EU review, so this would normally mean that Andrew would chair the meeting. However, as it is the restart meeting I wondered if it made sense to have a slightly larger activity recap meeting next week instead. This would be a precursor to the technical discussion in the GridPP meeting the following week as these discussions will be all about the future. >>>> >>>> What do others think about this? >>> I think this would be a very good idea and I'm not sure I could have made it this afternoon due to other committments. >>> >>> Cheers >>> >>> Andrew >>> > -- Facts aren't facts if they come from the wrong people. (Paul Krugman)