These are good numbers to have.
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
-
Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.
Windows BIGADV is Here
- Thread starter amdgamer
- Start date
These are good numbers to have.
EvilAlchemist
2[H]4U
- Joined
- Jan 11, 2008
- Messages
- 2,732
From Dr Kasson over at FF
bigadv has been temporarily turned off for linux due to problems with the linux A3 core on bigadv workunits
bigadv has been enabled for windows.
Depending on the work unit availability, machines requesting bigadv may roll over to normal -advmethods.
bigadv has been temporarily turned off for linux due to problems with the linux A3 core on bigadv workunits
bigadv has been enabled for windows.
Depending on the work unit availability, machines requesting bigadv may roll over to normal -advmethods.
I'm not either. I never thought I'd ask this, but what the heck did you guys do to change your clients' config and receive -bigadv WUs??
- Joined
- Oct 23, 2007
- Messages
- 6,105
I'm not either. I never thought I'd ask this, but what the heck did you guys do to change your clients' config and receive -bigadv WUs??
Add -bigadv flag
IDK if it's mandatory but I also have my client config set to always have advmethods on, and my client flags are -bigadv -smp.
- Joined
- Jul 5, 2004
- Messages
- 1,859
More stats:
W3520 @ 3.8Ghz
1GB used (assigned it 5GB but its just not using it)
2.19 core
2684 project
TPF=48:17
PPD = 18.2k (HFM)
So I'm only 2k ppd short from regular SMP2.
W3520 @ 3.8Ghz
1GB used (assigned it 5GB but its just not using it)
2.19 core
2684 project
TPF=48:17
PPD = 18.2k (HFM)
So I'm only 2k ppd short from regular SMP2.
OK, well I have the -bigadv flag, client type=3 as well as big packets in the config and the -advmethods flag. It's not just one client but two clients that are receiving standard A3 WUs only.Add -bigadv flag
IDK if it's mandatory but I also have my client config set to always have advmethods on, and my client flags are -bigadv -smp.
Last edited:
OK, well I have the -bigadv flag, client type=3 as well as big packets in the config and the -advmethods flag. It's not just one client but two clients that are receiving standard A3 WUs only.
I had the same settings as you and I was able to get a couple of them, at least the first day, just fine. [Clienttype]type=0 in my config though. I don't know if that makes a difference or not.
I missed the big packets setting at first, but it showed up in the log as packet too big or something to that effect when trying to download a WU.
freeloader1969
2[H]4U
- Joined
- Dec 13, 2003
- Messages
- 2,238
Absolutely nothing yet. Still getting 6072's.
Code:
[02:56:50] + Processing work unit
[02:56:50] Core required: FahCore_a3.exe
[02:56:50] Core found.
[02:56:50] Working on queue slot 05 [July 1 02:56:50 UTC]
[02:56:50] + Working ...
[02:56:50] - Calling '.\FahCore_a3.exe -dir work/ -nice 19 -suffix 05 -np 6 -checkpoint 15 -verbose -lifeline 3216 -version 629'
[02:56:50]
[02:56:50] *------------------------------*
[02:56:50] Folding@Home Gromacs SMP Core
[02:56:50] Version 2.22 (Mar 12, 2010)
[02:56:50]
[02:56:50] Preparing to commence simulation
[02:56:50] - Looking at optimizations...
[02:56:50] - Created dyn
[02:56:50] - Files status OK
[02:56:50] - Expanded 1761611 -> 2243881 (decompressed 127.3 percent)
[02:56:50] Called DecompressByteArray: compressed_data_size=1761611 data_size=2243881, decompressed_data_size=2243881 diff=0
[02:56:50] - Digital signature verified
[02:56:50]
[02:56:50] Project: 6072 (Run 0, Clone 156, Gen 66)
[02:56:50]
[02:56:50] Assembly optimizations on if available.
[02:56:50] Entering M.D.
[02:56:56] Completed 0 out of 500002 steps (0%)
[03:00:54] Completed 5001 out of 500002 steps (1%)
Last edited:
I know that you can only throw 8 threads at the bigadv due to the limitations of the VM, but are you throwing all 12 threads at the bigadv running natively in Windows? If you are, I would have thought that you would see much better TPF's than that.
I'm not either. I never thought I'd ask this, but what the heck did you guys do to change your clients' config and receive -bigadv WUs??
Ever since I left the Linus bigadv to just run straight Windows SMP2 A3 stuff, i've been running the SMP2 client with nothing other than advmethods set on by default, and with the command line switches -smp 8 -verbosity 9 -bigadv. It happened to pick one up by itself. Based on what Evil just said, maybe you just got unlucky and all of us early folks grabbed up the bigadv's?
Maybe you can delete the client and reinstall and set it up again using a different machine number to see if it changes things? I didn't do anything to pick up that bigadv unit, so it is wierd that you can't pick it up even thought you added the -bigadv switch. I hope you get it resolved as I think Windows bigadv has been rock stable so far as noone has reported crashes.
- Joined
- May 21, 2008
- Messages
- 5,840
17 here, but I'm going back to A3s to raise my OC, then back to bigadv
Posted via [H] Mobile Device
Posted via [H] Mobile Device
tjmagneto
[H]ard DCOTM x2
- Joined
- Aug 6, 2008
- Messages
- 3,299
For the 2684 in the VM I was specifying 8 processors and using the -smp 8 flag, for the Windows 2684 I am using the smp flag which light up all 12 cores.
Yeah, that is what I thought you were doing. I thought that your TPF's would have been much better in native windows since you are throwing all 12 threads at it while the VM limited you to 8.
OK, well I have the -bigadv flag, client type=3 as well as big packets in the config and the -advmethods flag. It's not just one client but two clients that are receiving standard A3 WUs only.
Out of my 5 Core I7 boxes doing BigAdv, the only one that is not crunching BigAdv, is the one that I've set -advmethods flag on. It's on a hideous P6701.
The Windows native client runs at low priority while the VM runs at normal priority. Changing the priority of FahCore_A3.exe in Windows Task Manager has a pretty big effect.
I did this on a box running WinSMP/BigAdv and 3xGPU clients. I left the GPU clients on low priority and changed the WinSMP to "normal", but when I did that, the GPUs went down to 20% output, while TPF went down by nearly 3 minutes (from 44 to 41).
Needless to say I switched it back.
Interesting. Someone earlier posted they didn't have the flag and only set the client type. Perhaps setting both the flag and the client type in the config is negating the setting. I took off the flag for now and I'll see what effects if any there are. BTW, you have 5 i7 rigs? And, you're not in the top 20...?
Yes, this is what I have always done since the SMP2 client was released months ago. It should be just as effective and in some cases even necessary with -bigadv like it was with prior standard A3.
With mixed client configs, it is absolutely necessary or you'll encounter inordinate CPU cycle bleed off starving the GPU clients. In my machines, it is not sufficient to assign the SMP idle priority status, I need to remove a core from the flag and dedicate it to the GPU clients, or the SMP client undergoes a significant drop in performance because the GPU clients will starve it in turn when there's shared cores. That's probably the only gripe I have about the SMP2 client. It's a lose-lose proposition with mixed client configs or at least with the systems I run, which are admittedly different than 95% of the folders out there.
OK, confirmed, you're right. I removed the -advmethods flag and left the client type at 3 and received a P2684. Both clients are now running -bigadv however, they are the exact same WU, 5-22-0. Does anyone know if two clients working on the same WU will both be credited??
Memory usage is extremely low in comparison to even native Linux with previous -bigadv. I'm reading a hair under 1.5GB total memory usage (OS+processes) in Win XP-32 on each of my systems. Very, very impressed.
I am seeing the same thing on both of my rigs. Very nicely done. Definitely compared to the VM world where I was eeking by with 6GB
- Joined
- Apr 25, 2001
- Messages
- 15,930
I'm dropping one this evening.
- Joined
- May 21, 2008
- Messages
- 5,840
Dropping one in 6 hours, around 80k I think, then more ovrrclocking!
Posted via [H] Mobile Device
Posted via [H] Mobile Device
Well, I have to say that after my initial excitement I am left quite disappointed. My fastest -bigadv machine is c r a w l i n g through these units. It's probably the only one capable of processing the Windows version of the P2684 out of my 4 original -bigadv systems. I don't understand why people have stated their TPFs are lower than they were in Linux. Not in my case. I'm barely making the bonus deadline at 52+ minutes when I was doing mid-40s min in native Linux previously. This is not a slow machine. We're talking dual Harpertowns at nearly 2.9GHz. The TPF I'm seeing is not acceptable and even had superior PPD from the fastest standard A3 WUs...
Interesting.......
I have my SMP client priority set to idle and maybe that is why my TPF's on the 2684 A3 bigadv are absolutely no faster and exactly the same as it was under Linux via the VM. Next time I have a chance, i'm going to do what you did and change priority to normal and see if it won't speed things up a bit.
- Joined
- May 21, 2008
- Messages
- 5,840
Apollo did you juststart them? Waituntil like the 7th frame for a good time, my fist two showed 38-39 minutes, now I'm down to high 28's or 29 even (currently 28:55)
Posted via [H] Mobile Device
Posted via [H] Mobile Device
Unfortunately, it's on the 12th frame. I started the WU late last night and left it running through the early part of the day. I checked the system right after midday, and it had only accomplished about 10 or 11 frames. It's crazy slow. Keep in mind this is my fastest system out of 4 dual quads.
I can't continue running -bigadv on this system (or any other) when it's so close to the bonus deadline. Any significant downtime and that's nearly 4 days of work with only partial credit - not worth it. More than ever, I would need to frequently monitor and make sure to use the -oneunit flag just for routine things like maintenance to avoid lost time. Stranger still, I have seen regular A3 WUs outperform the Windows P2684 PPD-wise. That's something I haven't heard anyone claim yet. I know my systems are a little different than the vast majority here, but come on!!
Last edited:
I'm at 20% complete now with my 1st one, 48:34 TPF and beating deadline by about 13 hours. Also, my CPU temp has dropped about 5c which I'm pretty happy about. I figured out that with my unmolested TPF being approx 49 minutes, I can actually USE my PC about 3 hours a day for Gaming and still clear deadline. Things like netflix and web browsing don't seem to bother it at all
Actually, has anyone noticed that their cpu's are running cooler when doing windows bigadv? My processor is running about 2-3 degrees cooler than when it was running the regular A3 work units. Room temps are climate controlled so they stay constant.
404 - File not found
Gawd
- Joined
- Apr 2, 2007
- Messages
- 525
This is what I'm getting:
i7 920 C0 @ 3927Ghz (187x21) 1.296v in cpuz. 80C max temp.
1GB used
2684 project
TPF=48:28
PPD = ~18k
This is on my workstation tho and I've been using it all day. I noticed no lag, which is really nice. Gaming was laggy, but unchecking a thread for the a3 process did the trick.
i7 920 C0 @ 3927Ghz (187x21) 1.296v in cpuz. 80C max temp.
1GB used
2684 project
TPF=48:28
PPD = ~18k
This is on my workstation tho and I've been using it all day. I noticed no lag, which is really nice. Gaming was laggy, but unchecking a thread for the a3 process did the trick.
- Joined
- Oct 23, 2007
- Messages
- 6,105
I've noticed my TPF (actually all my stats in HFM) go up and down pretty consistently on my primary/2 gpu client workstation. One frame it'll lowball me, then I check it a few hours later and it's back up to where I think it should be (around 18-19k PPD). Guess I was incorrect when I posted previously about lower figures. I assume that's because at the time that particular frame is being analyzed for figures, the gpu client is hitting the cpu harder.
I assume that's because at the time that particular frame is being analyzed for figures, the gpu client is hitting the cpu harder.
- Joined
- May 21, 2008
- Messages
- 5,840
Broke my hand today so upgrades are on hold but I think I got 79k for that unit
yes temps are lower, and draw too, jumps around a lot tho
Posted via [H] Mobile Device
yes temps are lower, and draw too, jumps around a lot tho
Posted via [H] Mobile Device
sirmonkey1985
[H]ard|DCer of the Month - July 2010
- Joined
- Sep 13, 2008
- Messages
- 22,412
did they tell you not to punch your commanding officer in the face in bootcamp?
lolDude!
I'm sorry to hear that, but I hope your hand gets better soon. I hope it is your non dominant hand.
- Joined
- May 21, 2008
- Messages
- 5,840
It's my right lol, I'm screwed, no more gaming or building boxes
on a good note I got even 79k on that unit, seems very stable
Posted via [H] Mobile Device
on a good note I got even 79k on that unit, seems very stable
Posted via [H] Mobile Device
Yeah, you are screwed. However at least this will give you a good amount of time to catch up on all of those DVD's that you meant to watch, or to totally abuse a Netflix account(they don't allow it and will throttle you). Your rig is stable, so at least no worries there.
I abuse my netflix account and they send me extra disks from time to time with fast turnaround.
On another note, on uploads I noticed my connection was still usable, not fast mind you but I could still use the interwebs. This was not the case with the previous -bigadv. Anyone else confirm this?
- Joined
- Apr 25, 2001
- Messages
- 15,930
I had no problems using the net when my -bigadv was uploading. I have 2meg up.
- Joined
- Apr 25, 2001
- Messages
- 15,930
well.... there are those of us that live in BFE and are not so lucky
I'm in BFE, but I'm fortunate to have a WISP that offers good service and speeds. No cable or DSL out here.
Yeah, that was why I cancelled my Netflix subscription. A buddy of mine also cancelled after a while for the same reason. During some months when I go through discs really fast, they started to throttle me and wait a few days before sending out the next movie in my queue.
On my rig, my connection is just plain unusable when any work unit(bigadv or regular) is in the process of uploading. However, my uploads are limited to only 384kbps max
They have a new Fiber Optic internet service, but it will probably never come to my neighborhood in the ghetto where I probably am the only computer user.
tjmagneto
[H]ard DCOTM x2
- Joined
- Aug 6, 2008
- Messages
- 3,299
I have one dropping this evening but the downside is that I was having GPU problems again so I was killing the ppd a bit while I was fighting with the 275s. They don't like the 105xx WUs and throw EUEs. The rest of the story will be for another thread, maybe.
They don't like the 105xx WUs and throw EUEs. The rest of the story will be for another thread, maybe.