A lot of visitors from SETI this week - they had a two day outage to move their servers to a new and better-resourced co-location facility. I imagine the vast majority of wingmates will report in in due course, but possibly not until close to the 14-day deadline.
OK -- that's plausible -- I do SETI stuff as well and shifted a bit over the weekend in anticipation of the outage.
Also, anyone doing POEM work may have come this way -- POEM went offline on Tuesday and is still unreachable.
Quote:
A lot of visitors from SETI this week - they had a two day outage to move their servers to a new and better-resourced co-location facility. I imagine the vast majority of wingmates will report in in due course, but possibly not until close to the 14-day deadline.
A lot of visitors from SETI this week - they had a two day outage to move their servers to a new and better-resourced co-location facility. I imagine the vast majority of wingmates will report in in due course, but possibly not until close to the 14-day deadline.
Wonder why the deadline is 14 days for the GPU WUs? It seems excessive. Took a look at the top hosts and some have up to over 1500 WUs queued. I would think that a shorter deadline and perhaps a WU cap would be easier on the database and work better generally.
I tried some of the Einstein GPU work units in the past -- they seemed to run relatively inefficiently compared to MW, Moowrap, Collatz, or even POEM and WorldGrid. I don't know if that has changed recently.
If it has, I might re-enable Einstein AMD GPU use as an alternate on the three workstations running World Grid GPU's particularly as the POEM outage drags on.
My sense regarding POEM is that somewhat similar to SETI they are dealing with a server relocation. The differences for POEM -- 1) the relocation was apparently not anticipated so all background and pre-planning time is done serially during the outage which of course extends the length of the outage; and 2) POEM typically is not particularly proactive about communication.
In any event,I will give Einstein another look for GPU.
Quote:
Wonder why the deadline is 14 days for the GPU WUs? It seems excessive. Took a look at the top hosts and some have up to over 1500 WUs queued. I would think that a shorter deadline and perhaps a WU cap would be easier on the database and work better generally.
A lot of visitors from SETI this week - they had a two day outage to move their servers to a new and better-resourced co-location facility. I imagine the vast majority of wingmates will report in in due course, but possibly not until close to the 14-day deadline.
Wonder why the deadline is 14 days for the GPU WUs? It seems excessive. Took a look at the top hosts and some have up to over 1500 WUs queued. I would think that a shorter deadline and perhaps a WU cap would be easier on the database and work better generally.
I tried some of the Einstein GPU work units in the past -- they seemed to run relatively inefficiently compared to MW, Moowrap, Collatz, or even POEM and WorldGrid. I don't know if that has changed recently.
Rather depends on your measure of 'efficiency'.
Round here, we tend to use "Pulsars found per Mega-WU". What did you use at those other projects?
Richard -- a couple of 'measurements' -- also, like Einstein, with their degree of arbitrary. (Clearly Einstein is FAR more effective at Pulsars found per Mega-WU than any other project). One of course is the credit system itself -- which again is pretty arbitrary and on a project by project basis. The other is the combination of how long the Einstein GPU WU's run (quite long from what I recall) along with how Einstein GPU WU's use concurrent CPU's.
Quote:
Quote:
I tried some of the Einstein GPU work units in the past -- they seemed to run relatively inefficiently compared to MW, Moowrap, Collatz, or even POEM and WorldGrid. I don't know if that has changed recently.
Rather depends on your measure of 'efficiency'.
Round here, we tend to use "Pulsars found per Mega-WU". What did you use at those other projects?
Richard -- a couple of 'measurements' -- also, like Einstein, with their degree of arbitrary. (Clearly Einstein is FAR more effective at Pulsars found per Mega-WU than any other project). One of course is the credit system itself -- which again is pretty arbitrary and on a project by project basis. The other is the combination of how long the Einstein GPU WU's run (quite long from what I recall) along with how Einstein GPU WU's use concurrent CPU's.
Credit is pretty arbitrary on every project, its what work is done...
How long GPU units run doesnt really have any bearing on its efficiency either, its the work that is done in those work units and they arent that long running either. My 670 will run up to 3 at a time in 40-48minutes depending on what else it is doing
Mike, I understand. The word 'efficiency' appears to have kicked up a certain amount of dust here. I can accept that. What is 'efficient' for me with a collection of mid range ATI GPU's might be significantly different than for others.
I got the impression (might be wrong, and glad if I am) that Richard might have been taking umbrage at my use of that word. Then I responded in a way that might be interpreted as 'umbrage at the umbrage'. In any event, I'm quite happy providing CPU support for this project (and have been for nearly 8 years).
Pendings rising??
)
nothing to worry about it's normal.
PC setup MSI-970A-G46 AMD FX-8350 8 core OC'd 4.45GHz 16GB ram PC3-10700 Geforce GTX 650Ti Windows 7 x64 Einstein@Home
A lot of visitors from SETI
)
A lot of visitors from SETI this week - they had a two day outage to move their servers to a new and better-resourced co-location facility. I imagine the vast majority of wingmates will report in in due course, but possibly not until close to the 14-day deadline.
OK -- that's plausible -- I
)
OK -- that's plausible -- I do SETI stuff as well and shifted a bit over the weekend in anticipation of the outage.
Also, anyone doing POEM work may have come this way -- POEM went offline on Tuesday and is still unreachable.
RE: A lot of visitors from
)
Wonder why the deadline is 14 days for the GPU WUs? It seems excessive. Took a look at the top hosts and some have up to over 1500 WUs queued. I would think that a shorter deadline and perhaps a WU cap would be easier on the database and work better generally.
I tried some of the Einstein
)
I tried some of the Einstein GPU work units in the past -- they seemed to run relatively inefficiently compared to MW, Moowrap, Collatz, or even POEM and WorldGrid. I don't know if that has changed recently.
If it has, I might re-enable Einstein AMD GPU use as an alternate on the three workstations running World Grid GPU's particularly as the POEM outage drags on.
My sense regarding POEM is that somewhat similar to SETI they are dealing with a server relocation. The differences for POEM -- 1) the relocation was apparently not anticipated so all background and pre-planning time is done serially during the outage which of course extends the length of the outage; and 2) POEM typically is not particularly proactive about communication.
In any event,I will give Einstein another look for GPU.
RE: RE: A lot of visitors
)
RE: I tried some of the
)
Rather depends on your measure of 'efficiency'.
Round here, we tend to use "Pulsars found per Mega-WU". What did you use at those other projects?
Richard -- a couple of
)
Richard -- a couple of 'measurements' -- also, like Einstein, with their degree of arbitrary. (Clearly Einstein is FAR more effective at Pulsars found per Mega-WU than any other project). One of course is the credit system itself -- which again is pretty arbitrary and on a project by project basis. The other is the combination of how long the Einstein GPU WU's run (quite long from what I recall) along with how Einstein GPU WU's use concurrent CPU's.
RE: Richard -- a couple of
)
Credit is pretty arbitrary on every project, its what work is done...
How long GPU units run doesnt really have any bearing on its efficiency either, its the work that is done in those work units and they arent that long running either. My 670 will run up to 3 at a time in 40-48minutes depending on what else it is doing
Mike, I understand. The word
)
Mike, I understand. The word 'efficiency' appears to have kicked up a certain amount of dust here. I can accept that. What is 'efficient' for me with a collection of mid range ATI GPU's might be significantly different than for others.
I got the impression (might be wrong, and glad if I am) that Richard might have been taking umbrage at my use of that word. Then I responded in a way that might be interpreted as 'umbrage at the umbrage'. In any event, I'm quite happy providing CPU support for this project (and have been for nearly 8 years).