4GHz+ Laptop?

Winterknight
Winterknight
Joined: 4 Jun 05
Posts: 1,465
Credit: 377,805,796
RAC: 121,814

RE: I'm betting on the app,

Message 68217 in response to message 68204

Quote:

I'm betting on the app, since as you said, you'd have to be really hard core to get into the work involved with pushing a laptop platform to that kind of clock frequency.

AFAIK, there aren't any conventional desktop MB's utilizing the mobile Core 2's yet (but I really haven't researched that aspect at all), so just about everything in the cooling system would have to be custom built for the unit.

Alinator


Ref MB's for mobile Core 2 Duo, AOpen, at least, do one. idid i345

Andy

Bikeman (Heinz-Bernd Eggenstein)
Bikeman (Heinz-...
Moderator
Joined: 28 Aug 06
Posts: 3,522
Credit: 740,298,765
RAC: 1,094,271

Hi! As for "hidden" PCs:

Hi!

As for "hidden" PCs: If the PC was not hidden at some time, you will still be able to retrieve the list of all hosts belonging to a user (at that earlier time) by googling for
'Einstein "Computers belonging to xxxxx"' and following the link to the CACHE. The links shown are still valid even after "hiding"

The concept of "hiding" a PC is a bit misleading in BOINC, IMHO, as it just removes the links between the PC stats and the user stats.

CU

BRM

ohiomike
ohiomike
Joined: 4 Nov 06
Posts: 80
Credit: 6,453,639
RAC: 0

The question is- when do we

The question is- when do we get our hands on the optimized app?


DanNeely
DanNeely
Joined: 4 Sep 05
Posts: 1,364
Credit: 3,562,358,667
RAC: 0

EDIT: Sorry for the repeated

EDIT: Sorry for the repeated posts. They appeared to be failing on my end.

Brian Silvers
Brian Silvers
Joined: 26 Aug 05
Posts: 772
Credit: 282,700
RAC: 0

RE: This host is running

Message 68221 in response to message 68216

Quote:

This host is running native Linux. ;-)

cu,
Michael

[edit]
My stats say, it got about 22.3 c/h equals to 8.84 [c/(h · clock)]
The einstein app scales linear to clock frequency, so you can easily compare. ;-)
[/edit]

Well, according to AndyK's pending credit calculator, each core of that host is getting about 22.74 cr/hr, while my faster (per core) 3700+ system is only getting 14.85 cr/hr. So, it does indeed look like a 30% penalty for me.

Note: Yes, I understand there's a difference in the frequency ranges, but I think 30% is a fair estimate.

Brian

Brian Silvers
Brian Silvers
Joined: 26 Aug 05
Posts: 772
Credit: 282,700
RAC: 0

RE: The question is- when

Message 68222 in response to message 68219

Quote:
The question is- when do we get our hands on the optimized app?

I speculate not before S5R3. Since he is admitting that it is Core2 specific optimization, then perhaps the overall optimization is lower on a percentage basis. The vast majority of hosts for EAH are Windows-based, and there is a relatively large amount of AMD K8-based hosts. If both the AMD penalty and SSE optimization comes out for S5R2, S5R2 will likely be finished within 4 weeks instead of the current ~9 weeks. I doubt that's something they'll be willing to do... Even just removing the AMD penalty would probably take 2 weeks off of the completion times.

I hope I'm wrong. I hope that the statement that we would see increases in performance DURING the S5R2 run is honored...

Brian

Akos Fekete
Akos Fekete
Joined: 13 Nov 05
Posts: 561
Credit: 4,527,270
RAC: 0

RE: The question is- when

Message 68223 in response to message 68219

Quote:
The question is- when do we get our hands on the optimized app?


I can only try out some ideas and suggest them to the project programmer.
But I know he works on the client errors nowdays.
Those are more important than the speed.

Dave Burbank
Dave Burbank
Joined: 30 Jan 06
Posts: 275
Credit: 1,548,376
RAC: 0

RE: RE: The question is-

Message 68224 in response to message 68223

Quote:
Quote:
The question is- when do we get our hands on the optimized app?

I can only try out some ideas and suggest them to the project programmer.
But I know he works on the client errors nowdays.
Those are more important than the speed.

I can only take this as a good sign.

Work is being done to both optimize and correct the validation errors. The problem is that narrowing down the specific cause of the cross platform validation errors is turning out to be harder than expected.

Once the validation errors have been solved we might get to see an optimized app.

There are 10^11 stars in the galaxy. That used to be a huge number. But it's only a hundred billion. It's less than the national deficit! We used to call them astronomical numbers. Now we should call them economical numbers. - Richard Feynman

Bikeman (Heinz-Bernd Eggenstein)
Bikeman (Heinz-...
Moderator
Joined: 28 Aug 06
Posts: 3,522
Credit: 740,298,765
RAC: 1,094,271

RE: Work is being done to

Message 68225 in response to message 68224

Quote:

Work is being done to both optimize and correct the validation errors. The problem is that narrowing down the specific cause of the cross platform validation errors is turning out to be harder than expected.

I've heard that it's pretty much narrowed down, which doesn't mean it's easily fixed tho. But progress is definitely being made wrt. this issue.

CU

BRM

Dave Burbank
Dave Burbank
Joined: 30 Jan 06
Posts: 275
Credit: 1,548,376
RAC: 0

RE: I've heard that it's

Message 68226 in response to message 68225

Quote:


I've heard that it's pretty much narrowed down, which doesn't mean it's easily fixed tho. But progress is definitely being made wrt. this issue.

CU

BRM

That's good to hear. At this point I can only hope it's fixed before the end of this run. I've been pretty lucky and have had only a few WU's fail validation, the SIGABRT bug was much more of an issue for me.

And thanks for the google cache tip. It's funny that I tried earlier in the week to find the hidden computers and failed, then by chance stumbled across one and didn't realize what it was....LOL.

There are 10^11 stars in the galaxy. That used to be a huge number. But it's only a hundred billion. It's less than the national deficit! We used to call them astronomical numbers. Now we should call them economical numbers. - Richard Feynman

Comment viewing options

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