PrimeGrid Year of the Horse Races Jan 3-18

[Icecold]

My WR WU should finish within ten hours. I could finish another and then switch that video card to the Pony race. Or I could switch today.

Which would be best for the TeAm? Place better in Pony or make a poor showing in both? Or am I thinking wrong?

I was wondering the same thing. Once I finish my first round I'm pretty sure my 2 GTX 480's could do one more round of the WR WU's before the challenge is up.

 

[Plimogz]

I had to reduce the clock on my 780Ti.

Odd that the time to completion only went up about 8 minutes at the start of a new WU.

This.

What's the deal with large drops in core speed only resulting in small variations in time to completion? Has anybody noticed how memory speed affects things?

 

[Ken g6]

Day 5 stats. 1/3 of the way through the race!

Stallion stats:
Rank___Credits____Username
27_____391908_____zzuupp

Rank__Credits____Team
12____778765_____The Knights Who Say Ni!
13____768896_____Nemesis Germany
14____768647_____AMD Users
15____391908_____TeAm AnandTech
16____388708_____PrimeSearchTeam
17____388428_____Heinlein Fans
18____385849_____Hardware Canucks

zzuupp looks lonely there. Anybody besides me trying to join him?

pony stats:
Rank___Credits____Username
5______1861953____Lane42
20_____751099_____biodoc
57_____300404_____GLeeM
96_____180117_____plimogz
110____150129_____deerslayer_TA
130____120103_____SlangNRox
153____90155______geecee
282____29947______Ryan

Rank__Credits____Team
4_____5252090____Team 2ch
5_____4653476____BOINC@MIXI
6_____4204209____Aggie The Pew
7_____3483910____TeAm AnandTech
8_____1920876____L'Alliance Francophone
9_____1830752____PrimeSearchTeam
10____1770799____Ukraine

We seem to be settling into 7th. But beware of people like me with slow CPUs who will dump later on in the race. :sneaky:

Edit: More news:

IMPORTANT:

New Project Preferences for CPU app selection!

Those of you who have been running CPU apps for this challenge probably have noticed that the server hasn't been sending out the correct version of the app (SSE3/SSE4/AVX). It's supposed to choose the correct version, but it hasn't been. (It worked fine in testing when installed last year, of course.)

I have replaced the automatic method with a manual selection method. For Genefer-Short, you can now explicitly choose either the normal CPU-SSE3 app, the much faster AVX app (if your CPU and operating system support it), or the slower SSE2 app if you're running on a very old 64 bit Athon II.

Here's what you need to know:

WHICH APP SHOULD I USE?

AVX: This is the fastest, but if your CPU or operating system doesn't support AVX the app will crash. To use this app, you must have:

An appropriate CPU, which is any Intel CPU starting with the Sandy Bridge CPUs. (Sandy Bridge, Ivy Bridge, or Haswell). Although recent AMD CPUs support AVX, their implimentation is very slow and there's probably no benefit to running the AVX app.

An appropriate operating system: 64 bit Linux, OSX, Windows 7 SP1 or later, and Windows Server 2008 SP1 or later.

SSE3: This will run on almost any 64 bit CPU, so if you can't run the AVX app, run the SSE3 app. Only the very earliest Athon II CPUs lack SSE3 support. All Athlon II's since Revision E (Venice and San Diego), and all other 64 bit CPUs, support SSE3 and can run this version of the app.

SSE2: If you have a 64 bit Athon II that is older than Rev e (see above), you most run this version of the app. This is equivalent to Genefx64.


WHAT YOU NEED TO DO:


WINDOWS:

If you had the FORCE AVX box checked before, the AVX box will be checked now, and you'll be receiving the AVX app. No action is necessary.

If you had the CPU box checked before, the SSE3 box will be checked now, and you will be receiving the SSE3 app. Unless you have one of those ancient Athon II CPUs, no action is necessary on your part.

IF YOU DO have an old Athlon II, you must uncheck the SSE3 box and check the SSE2 box.


MAC:

If you have an AVX capable CPU, you must uncheck the SSE3 box and check the AVX box to get the AVX app.

If you do not have an AVX capable CPU, no action is necessary on your part.


LINUX:

If you have an AVX capable CPU, you must uncheck the SSE3 box and check the AVX box to get the AVX app.

If you don't have an AVX capable CPU and had the CPU box checked before, the SSE3 box will be checked now, and you will be receiving the SSE3 app. Unless you have one of those ancient Athon II CPUs, no action is necessary on your part.

IF YOU DO have an old Athlon II, you must uncheck the SSE3 box and check the SSE2 box.


Sorry for all the inconvenience and confusion, but after these adjustments are made you should reliably receive the correct CPU apps.

 

[deerslayer]

Thanks for the stats Ken! I tried to get one more Pony wu uploaded but I wasn't fast enough.

 

[lane42]

Thanks for the stats Ken g6 :awe:

I Wondered why I wasn't getting sse3 work on one rig
and another one was.

 

[Ken g6]

My WR WU should finish within ten hours. I could finish another and then switch that video card to the Pony race. Or I could switch today.

Which would be best for the TeAm? Place better in Pony or make a poor showing in both? Or am I thinking wrong?

Good question. Supposedly, TeAm position in the Stallion race will count for 50% more than TeAm position in the Pony race in the year-end overall standings. Make of that what you will...

 

[Ken g6]

so, a GTX 570/580 check the cuda box. a GTX 780 check the opencl box,
I have one box with a 580 and 780, what would I do.

Finally got an answer for this:

In this case I can exclude GTX 580 for GeneferOCL and GTX TITAN for GeneferCUDA in cc_config:

<exclude_gpu>
<url>project_URL</url>
[<device_num>N</device_num>]
[<type>NVIDIA|ATI|intel_gpu</type>]
[<app>appname</app>]
</exclude_gpu>

What's the deal with large drops in core speed only resulting in small variations in time to completion? Has anybody noticed how memory speed affects things?

Supposedly, lowering memory speed helps stability more than lowering core clock speed.

 

[Plimogz]

Only the very earliest Athon II CPUs lack SSE3 support. All Athlon II's since Revision E (Venice and San Diego)

It looks like he's got AthlonII and Athlon64 confused.

Supposedly, lowering memory speed helps stability more than lowering core clock speed.

I wonder if the larger increase in stability resulting from decreasing memory speed means that there's a tighter correlation between memory speed and time to complete WU's than between core speed and WU completion time.

 

[GLeeM]

Good question. Supposedly, TeAm position in the Stallion race will count for 50% more than TeAm position in the Pony race in the year-end overall standings. Make of that what you will...

OK - I'm doing another one!

Another question: This one says it will take a little longer (the last one was a few minutes short of five days), I won't be able to finish the third WU by about one or two hours. Unless I overclock at least a little. So do I chance the OC?

 

[Ken g6]

Another question: This one says it will take a little longer (the last one was a few minutes short of five days), I won't be able to finish the third WU by about one or two hours. Unless I overclock at least a little. So do I chance the OC?

I wouldn't. The WUs are just going to keep getting longer as the race goes on. I'd go ahead and jump over to the pony race after this next one's done. (In fact, I will after my next one's done, but it's taking me over 10 days.)

 

[zzuupp]

OK - I'm doing another one!

Another question: This one says it will take a little longer (the last one was a few minutes short of five days), I won't be able to finish the third WU by about one or two hours. Unless I overclock at least a little. So do I chance the OC?

I agree with Ken. Primegrid tasks are over sensitive to overclocking in the bad way.

... and who's next in the wr race???

 

[lane42]

Thanks Ken g6.
I don't know why but im seeing faster times with GeneferOCL on the gtx580
then GeneferCUDA, around 9 hours for a OCL workunit, 11 hours for a CUDA workunit.

 

[Plimogz]

I finally looked through Ken's WU juggling guide in anticipation of the race deadline, and while it looks straight forward enough, I don't know what variable I need to control in order to make sure that none of the tasks are 'high priority'.

On my 2500K system for instance, all four of the AVX tasks which I've most recently started up are running under high priority, even though the dead line is two weeks from now and they shouldn't need more than 5 days to complete.

Has BOINC decided to run these in high priority because most of the other tasks in the queue have their deadline a week later? Thus making the tasks with the closer deadlines higher priority? And if so, is there a way besides aborting high priority tasks which will enable me to only have normal priority tasks?

 

[biodoc]

Thanks Ken g6.
I don't know why but im seeing faster times with GeneferOCL on the gtx580
then GeneferCUDA, around 9 hours for a OCL workunit, 11 hours for a CUDA workunit.

I wonder why my 660Ti is such a dog?

12 hours to finish a GeneferOCL WU.

EDIT: Maybe this is why? GTX580 has better double precision performance

SP GFLOPS DP GFLOPS
1581.1_______244 GTX580
2459.52______150 GTX660Ti EDIT: This is incorrect info. Need to find the right numbers.

 

[Plimogz]

^ In all likelyhood. Your 580 is probably faster than 680/770s withal.

Oh, and I think that I've got the juggling worked out. Though how BOINC assigns high priority is still a bit of a mystery. And I don't quite know how to modify it. But it doesn't seem like it should make much difference in the end.

 

[biodoc]

I moved my 780Ti (downclocked memory and power to 90%) to WR
480 and 570 (finished one WU) remain on WR
660Ti remains on the ponies

Estimated time to completion for 780Ti on WR is just under 75 hours. Hopefully it is stable.

 

[deerslayer]

I wonder why my 660Ti is such a dog?

12 hours to finish a GeneferOCL WU.

EDIT: Maybe this is why? GTX580 has better double precision performance

SP GFLOPS DP GFLOPS
1581.1_______244 GTX580
2459.52______150 GTX660Ti EDIT: This is incorrect info. Need to find the right numbers.

Could be worse. You could be running my 7770 (30 hours per pony). That 560 ti I bought is blowing it out of the water.

 

[biodoc]

Could be worse. You could be running my 7770 (30 hours per pony). That 560 ti I bought is blowing it out of the water.

Good point! Too greedy I guess.

 

[GLeeM]

Though how BOINC assigns high priority is still a bit of a mystery.

It is a mystery to me too, but I have, on occasion, seen the WU go from high priority back to normal after a few hours.

 

[deerslayer]

Good point! Too greedy I guess.

When this race is over I definitely need to move the 560 ti to my main rig. I didn't want to move the 7770 in the middle of the race.

 

[Ken g6]

Day 6 stats:

Stallion:
Rank___Credits____Username
25_____782929_____Icecold
39_____391908_____zzuupp
52_____387496_____plimogz
60_____386012_____biodoc
72_____384484_____GLeeM

Rank__Credits____Team
5_____3866892____BOINC@MIXI
6_____3853084____Dutch Power Cows
7_____3098082____BOINC.Italy
8_____2332832____TeAm AnandTech
9_____2329497____Ukraine
10____2319211____Keep The Fire Alive!
11____1957029____Overclock.net

Looks like we're doing quite well now.

Pony:
Rank___Credits____Username
5______2403278____Lane42
27_____901678_____biodoc
63_____360658_____GLeeM
79_____299789_____plimogz
92_____240281_____deerslayer_TA
106____210177_____SlangNRox
157____120275_____geecee
170____120021_____Icecold
178____119906_____zzuupp
250____59900______Ryan

Rank__Credits____Team
4_____6785475____Team 2ch
5_____6370866____Aggie The Pew
6_____5914906____BOINC@MIXI
7_____4835969____TeAm AnandTech
8_____2641943____L'Alliance Francophone
9_____2581822____Ukraine
10____2402239____Crunching@EVGA

Still holding steady.

 

[deerslayer]

Thanks for the stats. All these pending results are making me nervous. I think I've had two total that were validated.

 

[Ken g6]

Pending results shouldn't make you nervous. They just mean you're the first one to return a completed result. Two identical results are necessary for validation.

Inconclusive results should make you more nervous. They mean someone else got a different result, and at least one of the results is wrong. Although in cases like this, for instance, it's obvious that the very fast result is wrong.

 

[GLeeM]

Thanks for the stats, ken_g6

Looks like we're doing quite well now.

And without Peter! (I'm not saying we don't need/miss him)

All these pending results are making me nervous. I think I've had two total that were validated.

I know what you mean - me too. We just have to trust that the ones that validated are like the rest.

I have a feeling that all teams will have some invalid results - more than other races. If you look at your WUs you will see many that have error-ed out many times. I have one that was sent out 15 times! And I saw you had one that was sent out 13 times.

 

[deerslayer]

Thanks for the stats, ken_g6


And without Peter! (I'm not saying we don't need/miss him)


I know what you mean - me too. We just have to trust that the ones that validated are like the rest.

I have a feeling that all teams will have some invalid results - more than other races. If you look at your WUs you will see many that have error-ed out many times. I have one that was sent out 15 times! And I saw you had one that was sent out 13 times.

Thanks for the comforting words

Where do you find the error info? Is it in the text file that gives the time estimation?