Timeline for Power apps for S5R4 Run?

Rod
Rod
Joined: 3 Jan 06
Posts: 4,396
Credit: 811,266
RAC: 0

I would like to put my two

I would like to put my two cents worth... My computer is only powered up during the day that I might use it within a hour.. Other wise it is put to sleep... Primary purpose for my Mac is for my use, Boinc takes a secondary role and I am honored to contributed. But that it... there are limits..

I am just curious as to how many of the currently ~ 207000 users who are like me

There are some who can live without wild things and some who cannot. - Aldo Leopold

Odd-Rod
Odd-Rod
Joined: 15 Mar 05
Posts: 38
Credit: 8,094,587
RAC: 7,090

There are still a lot of P4's

There are still a lot of P4's out there - and even lesser hosts (like some of mine ;) )

While we may be beaten by the newer pc's, remember that many slow hosts together still add up to a lot of crunch power. After all, isn't that how the whole Boinc concept works? Even the best host on Boinc is slow when compared to the total crunch power of Boinc.

As long as my slowest hosts are returning valid results in time. I'll keep them going. (Assuming I can afford the electricity, of course!)

Regards
Rod

Nothing But Idle Time
Nothing But Idl...
Joined: 24 Aug 05
Posts: 158
Credit: 289,204
RAC: 0

RE: There are still a lot

Message 83474 in response to message 83473

Quote:

There are still a lot of P4's out there - and even lesser hosts (like some of mine ;) )

While we may be beaten by the newer pc's, remember that many slow hosts together still add up to a lot of crunch power. After all, isn't that how the whole Boinc concept works? Even the best host on Boinc is slow when compared to the total crunch power of Boinc.

As long as my slowest hosts are returning valid results in time. I'll keep them going. (Assuming I can afford the electricity, of course!)

Regards
Rod


There are a lot of P4s and perhaps their collective output is sizeable. But it's also slow and there is the data base and communications overhead associated with it...what is the cost versus the output? There are factors to consider other than the number of computers. I can only look at it from my perspective and my contribution is negligible until and if I can upgrade my computer. From a project perspective would you rather have 100,000 slow P4s or 50,000 fast Core 2 duos and quads?

RandyC
RandyC
Joined: 18 Jan 05
Posts: 6,146
Credit: 111,139,797
RAC: 0

RE: RE: There are still a

Message 83475 in response to message 83474

Quote:
Quote:

There are still a lot of P4's out there - and even lesser hosts (like some of mine ;) )

While we may be beaten by the newer pc's, remember that many slow hosts together still add up to a lot of crunch power. After all, isn't that how the whole Boinc concept works? Even the best host on Boinc is slow when compared to the total crunch power of Boinc.

As long as my slowest hosts are returning valid results in time. I'll keep them going. (Assuming I can afford the electricity, of course!)

Regards
Rod


There are a lot of P4s and perhaps their collective output is sizeable. But it's also slow and there is the data base and communications overhead associated with it...what is the cost versus the output? There are factors to consider other than the number of computers. I can only look at it from my perspective and my contribution is negligible until and if I can upgrade my computer. From a project perspective would you rather have 100,000 slow P4s or 50,000 fast Core 2 duos and quads?

From a Project perspective, if I needed the crunching power and they were available I would take 100,000 slow(?) P4s AND 50,000 fast Core 2 duos/quads.

Slow is a relative term.

Seti Classic Final Total: 11446 WU.

rbpeake
rbpeake
Joined: 18 Jan 05
Posts: 266
Credit: 997,689,131
RAC: 653,029

RE: From a Project

Message 83476 in response to message 83475

Quote:

From a Project perspective, if I needed the crunching power and they were available I would take 100,000 slow(?) P4s AND 50,000 fast Core 2 duos/quads.

Slow is a relative term.


The more the merrier! :)

But I too was in the position of figuring how best to use my old, slow computer to get the "biggest bang for the buck", and where I could use it most efficiently and most productively.

Eventually it got to the point where I got a new computer, much faster and more energy efficient on a per calculation basis! :)

Erik
Erik
Joined: 14 Feb 06
Posts: 2,815
Credit: 2,645,600
RAC: 0

The project is going to be

The project is going to be happy with any computing power resources contributed to it no matter how slow/fast the computer is, provided that the wu's are returned in the time frame granted. That is a given.

It is a matter of decision by the contributors on weighing the pros and cons, for whatever reason(s), that they decide whether to continue (and to what amount) or stop supporting the project.

For myself, I only have the quad hitting on two cylinders ATM for BOINC. That is all I can conscionably give right now after considerable consideration.

Odd-Rod
Odd-Rod
Joined: 15 Mar 05
Posts: 38
Credit: 8,094,587
RAC: 7,090

RE: There are a lot of P4s

Message 83478 in response to message 83474

Quote:

There are a lot of P4s and perhaps their collective output is sizeable. But it's also slow and there is the data base and communications overhead associated with it...what is the cost versus the output?


I would imagine that the amount of database and communications overhead is the same per WU. If that is so, then with the faster hosts downloading and uploading more often, they are actually making more demand on these that the slower hosts. Milkway@home, in fact, increased their crunch time to reduce the server load.

Quote:

From a project perspective would you rather have 100,000 slow P4s or 50,000 fast Core 2 duos and quads?


Ah, but it's not a question of 'OR' here, rather 'AND'. I'd rather have the fast and slow ones than only the fast ones. As RandyC also said.

Regards
Rod

Martin P.
Martin P.
Joined: 17 Feb 05
Posts: 162
Credit: 40,156,217
RAC: 0

RE: RE: I've stopped

Message 83479 in response to message 83464

Quote:
Quote:
I've stopped running S5R4 Einstein until there is a power app for Windows. I had one task that ran for 32 hours for a "standard" credit, whereas in S4R3 my run times with the power app were 8.5 to 11+ hours. Other S5R4 tasks ran at 18+ hours. That's too much of a credit per CPU time cut. I'd rather give my time to Rosetta even though their credit is also low.

I contributed to Einstein for 3 years now but I finally have to suspend it. My last task was over 31 hours while my wingman did it in half the time. My intel P4 contribution is miniscule these days and I too am throwing my meager cpu cycles to Rosetta. Credit is irrelevant. I might not discover gravity waves or pulsars but maybe I can contribute to curing health related diseases and do it in a more timely manner. If I ever upgrade my computer maybe I'll return. Sorry.


I agree. Just set E@H on my Dual-P4-3GHz machine to "no new tasks", my PowerMac (G5-Dual-2.7GHz) will follow tonight. For the time being only my MacPro will stay with E@H.

Randall  McPherson
Randall McPherson
Joined: 14 Mar 05
Posts: 6
Credit: 8,341,604
RAC: 0

I'm not sure this is the best

I'm not sure this is the best thread, but I have a question regarding S5R4 runtime on my Xeon E5440. It is quad core and the benchmarks are:

2903 Whetstone
5671 Dhrystone

for each core. My S5R4 WUs are taking 50 hours to complete. This is far longer than I would expect given my dual core E8500 cpu (with slightly better per core benchmarks) finishes the same S5R4 WUs in under 8 hours.

Does anyone know why the performance is so slow on the Xeon?

Thunder
Thunder
Joined: 18 Jan 05
Posts: 138
Credit: 46,754,541
RAC: 0

RE: I'm not sure this is

Message 83481 in response to message 83480

Quote:

I'm not sure this is the best thread, but I have a question regarding S5R4 runtime on my Xeon E5440. It is quad core and the benchmarks are:

2903 Whetstone
5671 Dhrystone

for each core. My S5R4 WUs are taking 50 hours to complete. This is far longer than I would expect given my dual core E8500 cpu (with slightly better per core benchmarks) finishes the same S5R4 WUs in under 8 hours.

Does anyone know why the performance is so slow on the Xeon?

Odd... I don't have any 54xx Xeons but a 5345 that I have is at least in the ballpark with your 8500 at 9:30-9:50 or so (and that's on a Win2003 server running exchange, sharepoint, AD, etc. etc. so it's not like E@H gets to line up it's data neatly in the caches).

Without knowing anything more about the system (OS, other software running, etc.) I could only speculate that it might be the thermal management on the 5440 kicking in frequently. I'm not completely positive, but I think they can lower themselves to 1Ghz or so with thermal throttling, but that seems pretty unlikely.

I'm sure if you can provide as much detail about the system as possible, some of the gurus around here will come up with ideas for solutions.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.