Thanks for the warm welcome, even though I've been lurking far longer than I have created an account on this forum. I probably can't get my CPU to fold well because my CPU gets real toasty at around 70C on stock clocks (3.5 GHz). I did prior research about the limits of my 6 core i7-3930k, the intel specifications say that 66.8C is the highest temperature intended for the chip.
My room is cool enough to fold with my GPU at around 65C . My next WU should finish by the end of this hour.
My wife's box (3930k) has all 12 threads pretty well pegged at all times at stock clock and core temp shows them typcially between 55 and 68 degrees C. Mine pretty well run 24/7 on BOINC projects including the math ones that push the CPU's harder than most. It has been rock solid for over a year now. There are 4 GPU's crunching in the box too. Granted the thing has a single Delta in the case making it sound like a jet engine, but otherwise is all air cooled.
LOL, those fans sure are loud. I don't think I can drastically improve the air flow in my desktop. The thing is I want my CPU to last me for a while, so I've held off overclocking. Also, I've tried running fah6 on linux and I got no response from the server. I really liked my console interface for folding at home.
Well damn. Maybe I can do 1M PPD with my two 970s and a stock 980. Temps have been pretty cool the past week, but I want them to be a little colder before I start folding. Hopefully nothing changes in the next few months.
I can already imagine PPD deflation within the next few months. Thanks for the greeting! The more, the merrier! Anyone got any estimates for completing a work unit for Core_A4?
It varies between 1 hour and several hrs. and does not have the best PPD but it will help get your 10 done faster.
I got 414 PPD (I can't check how many PPDs I got for each WUs) for submitting 2 Core_A4 WUs in around 40 minutes (I used a different computer ). Update: I got my first 10 WUs done, seems like the CPU WUs did the job Grandpa!
3 months ago I was comfortably aiming for the number 3 spot with a goal of number 2. The way things are going i'll be lucky to stay in the top 10. Which in a lot of ways is a good thing
This is true. I did hit 1M in a single day once last week. Luck of the draw clearly though. I have been considering adding another 970 or two to my NAS so maybe #4 and #3 is possible. 1 and 2 are on lock though.
I agree, I keep getting 30k point WUs that take 3 minutes (TPF) instead of 100k point WUs that take around 6 minutes (TPF).
Good to see things ramp up around here. A few weeks ago I was #7 in PPD for the team with just over 250K, now over 500K barely gets me in top 10. Keep up the good work people, Go [H]orde!
It has definitely picked up around here. We were averaging around 12M PPD a few weeks ago and now we're hitting 20M PPD. Not bad at all!
Something seems to be up with my boxen. I'm getting like half the PPD I'm supposed to all of a sudden.
Not sure what borg'd means, but it appears I'm getting a mix of 18 and 21s mostly 21s. I think the issue was Internet connectivity yesterday. I'll see what happens today. So far it seems to be running okay. I'm at 384K for the day so far. For some reason 3PM has been skipped the past few days, but yesterday I hit 398K for the whole day when I've been averaging over 800K.
Have a look in the log files; some of the newer projects/cores have a tendency to run into "Bad States" and retry at a last checkpoint (leading to lost of time and points). Another issue sure will do your connectivity as the points only count after uploaded/accepted by backends
ChristianVirtual there is a cure for those running 980's that are running into that problem on Windows rigs unfortunately there is no cure for those that are running Linux. Unless someone like tear makes us one. This is a post I made over at FF that addresses the problem on 980's it is a NVidia problem not a core 21 problem. https://foldingforum.org/viewtopic.php?f=19&t=28226
I've got two EVGA 970 and one EVGA 980 both running Windows 7. What do I search for in the logs to see if this is my problem. Search term "bad state" turned 0 results. edit Pulled up the full log in the directory and sure enough I am getting bad state errors yesterday. Line 693: 23:33:27:WU02:FS01:0x21:Bad State detected... attempting to resume from last good checkpoint Line 839: 02:13:31:WU02:FS01:0x21:Bad State detected... attempting to resume from last good checkpoint Line 881: 04:08:43:WU02:FS01:0x21:Bad State detected... attempting to resume from last good checkpoint Line 3352: 20:31:50:WU00:FS02:0x21:Bad State detected... attempting to resume from last good checkpoint Line 3453: 22:04:14:WU00:FS02:0x21:Bad State detected... attempting to resume from last good checkpoint Line 6272: 18:33:52:WU02:FS02:0x21:Bad State detected... attempting to resume from last good checkpoint Line 6554: 22:27:44:WU02:FS02:0x21:Bad State detected... attempting to resume from last good checkpoint Line 6664: 00:08:28:WU00:FS01:0x21:Bad State detected... attempting to resume from last good checkpoint Line 6685: 00:32:27:WU00:FS01:0x21:Bad State detected... attempting to resume from last good checkpoint Line 6708: 00:56:45:WU00:FS01:0x21:Bad State detected... attempting to resume from last good checkpoint
Guess what it is your problem go here http://www.guru3d.com/files-details/nvidia-inspector-download.html and get Inspector and drop the memory speed in P2 state, try 100Mhz reduction first. If you can not figure out how to use Inspector I will make a tutorial latter for you.
Went -100Mhz on both the 970s. I'll take a look at the 980 log in the HTPC when I am on it and see if it's got the same issue. Thanks guys.
The 980 will probably have a worse issue than the 970, at least all 3 of mine did By the way the up side to all of this is you will most likley be able to get a bigger OC out of the cards after doing this, I am currentley running 25 to 35Mhz higher than I prevousley could on my 980's I am now folding stably at 1500+ Mhz on all 3 of my 980 Classifieds
Yes I did my avg TPF went down 1 to 2 sec depending on the achievable OC and the lack of the BS gave me quite a boost in the daily PPD. You will need to find your sweet spot I started at -100 on the memory and raised it +10Mhz at a time until I started getting the BS message again. It varied from card to card my best will run at 5975Mhz without BS or slowdown the worst is 5953Mhz
... must ... resist ... installing ... Windows ... Hope one day in the future nV will allow/document such setting under Linux, too. Just don't want to switch
Winter is coming which means I need to heat my room. Last year I went the boinc route because of everything that happened with PG however this year I'm going back to f@h. I just built my skylake build and will be folding on my 980ti when I'm not using my PC. Last night when I was watching it before bed it said it should be around 500k-600k PPD depending on WU. I wont be using it to fold 24/7 but it should be enough to put me in the top 20 for now...
Well I believe it is time to turn that heater back on. Welcome back to the fold our daily production is still on the rise and with your starting back up along with any others that may follow only one thing can happen, we have already gone from 8 mil. PPD a little over a month ago, I think that made us #36 in production at the time to right at 22 mil. daily production now which I believe is #6 in daily production.
![[H]ard|Forum](styles/hardforum/xenforo/logo_dark.png){kind=logo}