Folding Forum

Note my wink Bruce!
All the same Bruce, it's confusing how some units work faster than others -and points are the easy way of checking outfit progress week to week. I know unit points can be weighted due to performance issues.
Some units really mess around [though current unit just seems slow, whatever the official bench says] but your comparison on the P4 seems to put it in context against my kit.
Some units survive pC crashes [as for different clients, too] and others don't.
All in the mysterious workings of Stanford but reading posts, it does seem there are 'horses for courses' * in the 'unit stakes'.
*PS: My horse in the Aintree National [y/d] lost its rider, btw !!

new08 wrote:Some units survive pC crashes [as for different clients, too] and others don't.

There's a known bug in the current version of the protomol core. Those projects rarely survive a stop/start, let alone a crash. Other than that one, I have not associated the crash-recovery with particular projects, but I'm not sure I have a representative sample.

In my experience, there are three kind of people:
* Those who have computers that don't crash
* Those who use file-systems that survive a crash and who rarely loose work (note: I didn't say "never")
* Those who use rotten file-systems who gripe about frequent loss of work.

Cody, That rings a bell.
I had a disk crash on me a few weeks back (I suspected folding activity may have contributed) and quite a few twitchy moments sorting out XP reload and other ongoing disk problems and reboots. I noticed that some units over a week or two survived crashes with data intact and at other times careful shutdown of F@H still resulted in lost data and unit restarts from scratch. This was flagged as bad units shutdown when in fact I had stopped them in advance.
Maybe after a few weeks of this, I was hoping to see some good work kick in, and thus noticed 1799 speed/points issues rather more than normal.
This must be a common story for many.

PS: before being asked - why suspect folding?
Well, I suspect the boot sector got corrupted in a write state during update on the other main disk due to setting on caches, as later there were faults on SMART data- but still taking data, now it's been reformatted.(Still has Smart errors)

Folding had been doing some strange things prior to this. All history now... and memory can play tricks !
Some Double Gromacs would be nice ...

bruce wrote:

new08 wrote:This unit does seem under priced
It's doing little better PPD than my old T21 700 @ a third the speed CPU.
Some posts on line show this has been around as a query a few years- but points increased from 300 ish sometime, maybe to compensate- but not enough as commented elsewhere. PS Normally @ about 95% or more cpu.

I'm not sure why you feel they're "under priced" The official standard is 110 PPD on a 2.8 GHz P4 and John says he's getting at least that much.

That would be perfect for small or normal sized WU ... but the upload size for these WU is 18 MB, which classify them in the Big WU category which should get a bonus (and yield about 220 PPD on the 2.8 GHz P4 mentionned above) ...

Previous WUs form the p179x series had very similar characteristics (processing time, upload size, ...), but were worth more than 2100 points ...

Previous WUs form the p179x series had very similar characteristics (processing time, upload size, ...), but were worth more than 2100 points ...

Yes that's right, I'm running them side by side on my other Phenom a 955 @ 3.6GHz, 1 getting 900ppd and the other 260ppd the 1799 seems to be close to the size timewise to complete (about 2 days), so it does seem a little lopsided.
I also seem to be getting them despite swapping my config to normal size units, while they actually seem to be big which was why I posted that I thought something was wrong.

Note that the points have been adjusted up since I started the thread so the initial numbers nolonger accurate.

Eric is investigating these projects : they shouldn't have such big results files ...

In the meantime, he suspended these projects from being assigned.