While checking on Boinc b/c of the client upgrade, I noticed that I have a new Rosetta client (as per the news) and I new type of WU.

On my comp, an "old" WUs would take around 2hr, but this new WU says it's going to take 9+ hrs!?!

Now it could be that with the new 5.2.2 the deadlines are all messed up again, but I also noticed that the naming convention has changed for WUs too. This new WU is called "1hz6A_abrelaxmode_random_gauss_02083" instead of the usual "1pvaA_abrelax_31702".

I'm not complaining or anything just observing. :)

Cheers!

---

I have two of these at the moment; one running, one ready-to-run. The one running stands at 70% complete after 18 minutes; the second one shows a time-to-completion of over five hours...so I'd guess that the "completion time bug" is still with us.

Interesting to note that the percent complete jumps in 10% increments, instead of 8.33% as in Rosetta 4.77.


(EDIT--WU completed in just over thirty minutes.)

---

Don't pay any attention to the time-to-completion. These new WU's should finish in less time than the previous ones. I will fix this in the next set of WU's. It has nothing to do with the client but has to do with the -rsc_fpops_est value I set when generating the work units. I will set it to a closer value.

Interesting... thanks for the info. :)

They're definitely small - just completed one in under 19 minutes on my new Athlon. Had me gpoing for a while there, thinking I got myself a supercomputer :-)

---

*** Join BOINC@Australia today ***

Yes, I'm getting tons of tiny WUs on all my boxes. Three point WU anyone?(https://boinc.bakerlab.org/rosetta/workunit.php?wuid=345218)

:-)

---

Team CFVault.com
http://www.cfvault.com

Hmmm... I completed the new wu in about 24 mins.

So the UW went from 2hr to 24 mins for me... that's a nice factor of 5. :)

So David... how's your server load??? hehehe

---

This new set of WU's are in fact smaller than the previous ones. Smaller in the sense that it does not require as much computation. This set is testing new sampling methods for ab initio fold prediction and does not use constraints. It is a smaller protein also.

This new set of WU's are in fact smaller than the previous ones. Smaller in the sense that it does not require as much computation. This set is testing new sampling methods for ab initio fold prediction and does not use constraints. It is a smaller protein also.

Thanks for the info, David. :-)

We always get worried when we see something new!

---

[b]"I'm trying to maintain a shred of dignity in this world." - Me[/b]

Three point WU anyone?

Luxury!

2.35

Three point WU anyone?

Luxury!

2.35

2.20

---

LUE

2.20

"And you try and tell the young people of today that ..... they won't believe you"

2.22's the "best" I've had so far - I yield to Desti's woosh-ier wu's!

Earlier today, I saw several WUs that completed in about 20 minutes, and then another that took the usual 1.5-2 hours...

---

There is a workunit batch with 'sim_aneal' in the name that takes longer to calculate. Expect ranges similar to the first batch of WU's. I see a range of around 2-10 hours.

"People of Earth, your attention please"

1.82

"People of Earth, your attention please"
1.82

That's going to be hard to beat. I now look silly with my 'huge' 2.87 WUs.

---

Team CFVault.com
http://www.cfvault.com

"People of Earth, your attention please"
1.82

That's going to be hard to beat. I now look silly with my 'huge' 2.87 WUs.

Yes, that one must have taken a splitsecond to process!

It's even smaller than some of the LHC WU's where the particle smash into the wall almost imidiately!! :-O

---

[b]"I'm trying to maintain a shred of dignity in this world." - Me[/b]