FAHClient V7.1.21 released (1st Open-Beta)

Nitro,

Yeah - can't run the FAHClient as a service and fold on a GPU. On my headless box (which isn't running FAHControl) I have it set to start as a console application at logon time.

H.
 
The 11293 WU's are test units - few atoms, which apparently leads to the poor performance/high CPU utilization. "Real" units may be better performing. There's still work going on tuning Core_16.

H.
 
The 11293 WU's are test units - few atoms, which apparently leads to the poor performance/high CPU utilization. "Real" units may be better performing. There's still work going on tuning Core_16.

H.

I seriously hope so, because this isn't exactly what I expected. I guess this means that the ATI GPU3 client is officially in open beta now? I'll give it a whirl if they release new work units for it to see if anything changes.
 
Anybody have the remote control feature working? Us there a way to monitor ppd of all the rigs on my network?
Posted via Mobile Device
 
Is there a way to monitor ppd of all the rigs on my network?
Posted via Mobile Device

Not that im aware of, because the log is now different. There is a single log for all Slots on a computer.

Reading it, makes my eyes want to bleed.
 
Anybody have the remote control feature working? Us there a way to monitor ppd of all the rigs on my network?
Posted via Mobile Device

I got it working, but the ppd wasn't accurate. It doesn't seem to be accurate even on the local machine. If I remember right it was only showing a ppd value, without bonus, on the advmethod units. The normal ones had no ppd value, just blank for that field with a base value shown for the unit.
 
Not that im aware of, because the log is now different. There is a single log for all Slots on a computer.

Reading it, makes my eyes want to bleed.

just think of the nighmare "atlasfolder" would have.

I really hope they give us an option for individual client logs vice group logs. troubleshooting a single hickuping client will be a nightmare for me, I have 4 clients on one machine, and I know some of us have even more.

 
Anybody have the remote control feature working? Us there a way to monitor ppd of all the rigs on my network?
Posted via Mobile Device
Remote control works fine, I've been doing it since much earlier beta versions. However, PPD is not accurate. Currently, FAHControl does not accurately calculate bonuses. Additionally, for PPD calculations to work, the work server has to be upgraded to V7. There are many servers out there that will not be upgraded to V7 for many months to come.
 
just think of the nighmare "atlasfolder" would have.

I really hope they give us an option for individual client logs vice group logs. troubleshooting a single hickuping client will be a nightmare for me, I have 4 clients on one machine, and I know some of us have even more.



i doubt he even pays attention to that stuff. but theres ways around it by installing multiple v7 clients and setting hfm to read each client. aka the old fashion way..
 
More trouble than it's worth.
eh, it's close.

I'm conflicted. I need HFM-like monitoring. My farm requires too much babysitting (due to my tweaking) to run reliably if unmonitored.

However, I would really like the ability to start crunching the next WU before the current one uploads.
 
More trouble than it's worth.


of course it is. but he runs all gpu's and no SMP so the v7 clients pretty useless for him anyways right now. plus who the hell would want to spend 2 weeks converting all those systems to v7.
 
how hard can it be to have a single radio box that allows us to split the logs up into different files. Sure for general monitoring I suppose it could be simpilar to have 1 log. But come on, seriously stanford, a lot of us would really benefit from multiple logs, and ultimately, you will benefit too.

I for one am not on the fence. IMO this software is not mature enough for me to install it. Sure there are features that are missing, there are even some significant improvements and the v6 client is a piece of dogsh!t. At least I know how to work around the v6 problems. My time is more valuable than to research the best way to optomize and control a piece of beta, new, unsupported software
 
connection untrusted

fah-web.stanford.edu uses an invalid security certificate.

The certificate is not trusted because no issuer chain was provided.

(Error code: sec_error_unknown_issuer)

trying to download the new client...is this normal??
 
I got that also (yesterday), but then again HP's site that I use every week to put in my timecard also has a screwed up cert, so I just proceeded through and scanned the download just to be (more) sure.
 
OK. so maybe not the NSA spoofing stanford to own my box?
 
I've got two 6970s and figured I'd give this a try just to see how it is...does GPU folding bring anyone else's PC to a grinding halt? I get about 12% CPU usage per GPU but my whole system would just freeze for like 5-7 seconds (can't click on anything - almost like stuttering on the first-gen SSDs), then allow me to do stuff for 2-3 seconds: wash, rinse, repeat.

Edit: seems I needed to lower the CPU usage from 100%. But my GPUs are grabbing FahCore 0x11 WUs and not Core_16 WUs...thought it was supposed to be a simple process this time.
 
Last edited:
Edit: seems I needed to lower the CPU usage from 100%. But my GPUs are grabbing FahCore 0x11 WUs and not Core_16 WUs...thought it was supposed to be a simple process this time.

Here's how to get the Core 16 WUs working: http://timverry.com/?p=661

My system (I have one 6950 O/Ced and an SMP client, both folding) doesn't feel any slower, but anything flash or GPU intensive that I accidentally start (I forget I'm folding) will sometimes crash the system requiring a hard reset. Really can't wait until that bug is gone!
 
Started working with v7 & just got my first core 16 wu 11294 which my 5870 is currently working on at between 94% & 99% usage. It's random if you get a core 16 wu which are limited release right now.

Will see how it works out, takes quite a bit of resources at 400mb ram & 15% CPU on my i920 w. HT on. If it works out well I will move back to folding again.

Tested & flash video does not have any adverse reactions.
 
I've dumped V7 running my GPUs for now as the configuration disappears upon rebooting. Maybe I'll take a look at it later on to see if I screwed up the install but everything is working with the systray client.
 
Started working with v7 & just got my first core 16 wu 11294 which my 5870 is currently working on at between 94% & 99% usage. It's random if you get a core 16 wu which are limited release right now.

Will see how it works out, takes quite a bit of resources at 400mb ram & 15% CPU on my i920 w. HT on. If it works out well I will move back to folding again.

Tested & flash video does not have any adverse reactions.

If you set the specific advanced methods flag you will only get Core 16 WUs and not the older ones. That 5870 can do 10k+ PPD if you overclock it btw ;) :)

Interesting that you're not running into the flash issue. What version of flash are you running and do you already have hardware acceleration disabled? just curious, but it'd be a nice surprise if it's been fixed and they somehow snuck a bug fix in on me lol.
 
Interesting that you're not running into the flash issue. What version of flash are you running and do you already have hardware acceleration disabled? just curious, but it'd be a nice surprise if it's been fixed and they somehow snuck a bug fix in on me lol.

AFAIK - The flash issue is only an issue with 69xx boards, anything prior doesn't experience the issue.

H.
 
AFAIK - The flash issue is only an issue with 69xx boards, anything prior doesn't experience the issue.

H.

Really, wow that sucks (for me)! The 5870 is getting better PPD /and/ doesn't have the flash issue!? Oh well, I'll pull ahead once they opitimize the new client more towards final release so I suppose I can live w/ it for now :) Alwaysbefolding after all.
 
I just hope they figure it out soon, if I had a farm it wouldn't matter, but with this thing being the fancy schmancy new box on the block the flash bug is cramping my style yo.
 
I have yet to try this yet but I might have found a solution to my issue of losing my configuration stuff when I reboot. When I set up my V7 I selected "Install for anyone using this computer." Based on what I can gather from Folding Forum is that there is a bug with this selection due to permissions and that one should use "Install just for me." I'll give that a go sometime soon and see what happens.
 
If you set the specific advanced methods flag you will only get Core 16 WUs and not the older ones. That 5870 can do 10k+ PPD if you overclock it btw ;) :)

Interesting that you're not running into the flash issue. What version of flash are you running and do you already have hardware acceleration disabled? just curious, but it'd be a nice surprise if it's been fixed and they somehow snuck a bug fix in on me lol.

Seems to be running ok. I am running my 5870 at 900mhz when doing distributed computing and I'm getting about 9000 PPD, seems to scale pretty linearly if 1ghz nets +/-10,000 PPD. Can push it higher, but 900mhz with fan at 40% gives 70 C, acceptable for me. Actually the 70 C tells me it could be working harder, full out it should hit 74 C. Also I have nothing optimized for folding so that may affect my results somewhat.

The 5870 has more raw shaders so that's where the increase in speed comes from I'd guess (along with development I would assume started in the 5xxxx series), but the 6xxx series has some nice new commands available which should make it go faster once optimized properly.

I'll continue folding & see how this develops!
 
Second beta out:

http://foldingforum.org/viewtopic.php?f=67&t=18302

Here are the change logs:

v7.1.22:
Added proxy support with authentication types: none, basic & digest.
Added proxy configuration tab to FAHControl.

v7.1.23:
Fixed bug caused by ignoring WU return code after a quick pause/unpause.
Set default verbosity to 3.
Explictly white-list Fermi GPUs & downgrade if CUDA driver is insufficient.
Added new GTX 470, 485 and 590 to GPU white-list.
Moved FAHControl single app port within 0-65535 range. #604.
Filtered out 'Theater' from GPU list.
White-listed Nvidia Quadro G8x cards.
Move all ATI R700s GPUs to species 3 and R800s to species 4.
Don't allow passwordless access even if password is not set.
Don't automatically allow IPs in command-allow-no-pass.
Don't call battery status code if not necessary.
Release resources related to battery query in OSX. #593
Allow changing the number of CPU threads mid run but warn. #292
Fixed client new and client connection options saving. #617
Lock database on startup to stop multiple runs of FAHClient in same dir.
Pass '-service' option to core when running as a Win32 service. #592
Fix permissions for All Users Windows install. #595.

v7.1.24:
Don't download a new WU if max-units is reached. #607
Added GeForce GTX 460 SE to the white-list.
Fixed 'core-priority' in FAHControl.
Fixed options save.
Don't allow changing 'local' client name or IP.
Don't try to autostart the local client once online.
Fix permissions problem in RPM. #627
Hide 'Status' tab in novice mode.
Added project info. In novice mode only by default.
 
I'll have to try it again since they addressed the All Users install issue.
 
I'm having unrelated MB issues (http://hardforum.com/forumdisplay.php?f=40) that are giving my some fits, but I have to say that I really like teh GUI that tehy have put up, other than a few small things. I'd love to see better logfile viewing options for those that are SLI or GPU/SMP folding, and the option panel is a mess. Other than that it's pretty nice.
 
I found this thread by accident - why isn't it stickied?

at any rate I just picked up 7.1.24 for win 7 x64

its working like a charm with my new dual 6950's

I took some getting use to the change option types so the first WU I picked up wasn't what I wanted but at any rate it looks like each card is going to finish its first WU in about 3hrs. I'll update PPD once it finishes one.

So currently what's the best PPD client type option to be running??

advanced, bigbeta, bigadv, or beta

Also I wanted to clarify / verify - you can set the above client type options either in each SLOT individually or globally under Configure - Expert tab correct?

p.s. i did have on hiccup when I was trying to add the client type to a slot (while it was folding) the window locked up - I could move it around but clicking OK, cancel, min, and X did nothing. I had to end the process to get out of it. I restarted the client - this time paused folding - set the client type - resumed folding - no issues.

edit:
this client is estimating 8000 PPD however that was BEFORE I set bigadv
 
Last edited:
I found this thread by accident - why isn't it stickied?

at any rate I just picked up 7.1.24 for win 7 x64

its working like a charm with my new dual 6950's

I took some getting use to the change option types so the first WU I picked up wasn't what I wanted but at any rate it looks like each card is going to finish its first WU in about 3hrs. I'll update PPD once it finishes one.

So currently what's the best PPD client type option to be running??

advanced, bigbeta, bigadv, or beta

Also I wanted to clarify / verify - you can set the above client type options either in each SLOT individually or globally under Configure - Expert tab correct?

p.s. i did have on hiccup when I was trying to add the client type to a slot (while it was folding) the window locked up - I could move it around but clicking OK, cancel, min, and X did nothing. I had to end the process to get out of it. I restarted the client - this time paused folding - set the client type - resumed folding - no issues.

edit:
this client is estimating 8000 PPD however that was BEFORE I set bigadv

If you're running this client using GPUs only, you will only get GPU WUs (I know, duh...) not bigadv, bigbeta, etc. Based on this site http://www.overclock.net/overclock-net-folding-home-team/475163-folding-home-gpu-ppd-database.html I would estimate your 8K PPD is about on target for those cards.
 
btw... this thread is linked to in the main sticky... you can only have so many stickies per subforum so all the guides are linked to in the main guide...
 
cbarbour71 and patriot - tyvm

I failed to realize the GPU WU's were a different animal than bigadv, etc.
 
@ INFINITE

Unless things have changed you will find that you will make more ppd letting the Caymans cool off and running just bigadv on the CPU. I tried the GPU client last winter on my dual unlocked 6950-i7 setup, it heated the house nicely, but hacked my bigadv bonus so bad that it wasn't worth it points wise. There is lots of speculation that Stanford is in bed with Nvidia on this. I do not know if that is true...but I do know that the red cards kick the snot out of the green ones at crunching bitcoin, another highly parallel process.
 
jojo has it right. You might be better off folding with your CPU, though without knowing what you have, I couldn't say for sure.
 
as always thx for the helpful insight

that's good news for two fold - I recently discovered my rig was locking up after folding about 3hrs (of gpu folding) upon further examination I found one of the 6950's is running 20C hotter @ idle than the other!! I swapped position of cards no change - there must be something wrong with the cooler on this card.

So at any rate I'll try the CPU today and see what type of results it yields. The good ol i7 2500k won't let me down ;)

Out of curiosity can you run BOTH cpu and gpu folding at the same time in this new client? I'm assuming you could by adding slots and changing their type from gpu to cpu?

edit got the cpu slot all configured and its still crunching its first bigadv WU but the estimate ppd only in the 4100 range. Does that seem normal? Seems awfully low to me. Or does it just need to crunch more wu to properly figure out the total?
 
Last edited:
Back
Top