OFFICIAL CGMINER mining software thread for linux/win/osx/mips/arm/r-pi 4.11.0 - page 625.

The00Dustin

hero member

Activity: 807

Merit: 500

Quote from: kano on March 06, 2012, 09:49:05 AM

Actually - that's the smartest comment yet on the subject ...

Quote from: kano on March 06, 2012, 09:49:05 AM

Edit: though I should point out ...

The 7970 is a single GPU

Being smart without actually being smart is awesome... I forgot about the gpu ordering issue, so basically each of these needs tried if there's not an easier way to work out which is the display (since board order and driver order don't match and even differ between OS):

Code:

-I 9,7,9,9
-I 9,9,7,9
-I 9,9,9,7

Good thing it isn't dual-GPU, that would make it a real PITA.

EDIT: Even assuming there's no good way to determine which is which short of such testing, wouldn't display performance have been noticeably worse in the previously attempted

Code:

-I d,9,9,9

if it wasn't making the display GPU dynamic?

kano

legendary

Activity: 4592

Merit: 1851

Linux since 1997 RedHat 4

... and on a different subject ... ZTEX

I wasted an hour or so on it so far and worked out what appears to be the commands
(and changed all the cgminer code that needs changing and created a template ztex.c)

So if anyone has one it would be good if they could drop by #cgminer on FreeNode and I'll ask you to run a few commands so I can work out what the commands really are.

I asked the ZTEX people and got 2 useless replies in one email:
1) Read the java code, the mining commands for the bitstreams aren't documented
(yeah I had already done that that's why I asked to confirm what I had worked out)
2) "You cannot do this within a day."
lol - well I've only spend a couple of hours on it so far ...
(and then decided after than comment ... stuff it I'll do it tomorrow ... or whenever ...)

kano

legendary

Activity: 4592

Merit: 1851

Linux since 1997 RedHat 4

Quote from: The00Dustin on March 06, 2012, 06:39:33 AM

Isn't the 7970 a dual-gpu card? Maybe having the display on the first card means both of those GPUs affect Windows because of this? IOW, instead of

Code:

-I 7,9,9,9,9,9,9,9

maybe

Code:

-I 7,7,9,9,9,9,9,9

or with 1/3rd the 9's if you only have 2 cards.

Actually - that's the smartest comment yet on the subject ...
Because it could be that GPU0 isn't the display ...
i.e. you'd have to make sure which GPU's are the display and set it lower.
Then I'm almost certain the rest could be higher ...

Edit: though I should point out ...

The 7970 is a single GPU

The00Dustin

hero member

Activity: 807

Merit: 500

Isn't the 7970 a dual-gpu card? Maybe having the display on the first card means both of those GPUs affect Windows because of this? IOW, instead of

Code:

-I 7,9,9,9,9,9,9,9

maybe

Code:

-I 7,7,9,9,9,9,9,9

or with 1/3rd the 9's if you only have 2 cards.

kano

legendary

Activity: 4592

Merit: 1851

Linux since 1997 RedHat 4

Well I have actually discussed recently what the intensity actually does.

It simply divides up the work based on: a higher intensity means longer work for the GPU to do non-stop.

... now where was that post ...
https://bitcointalksearch.org/topic/m.771425

So each step up in intensity is making the GPU process twice as long between very short 'rests'
(though they are only cgminer GPU rests while the CPU readies the next GPU request, not OS rests)

kano

legendary

Activity: 4592

Merit: 1851

Linux since 1997 RedHat 4

Quote from: ?? on ??

Quote from: kano on March 05, 2012, 06:41:29 PM

Quote from: ?? on ??

...
Try it before you laugh.

The point he has already stated is that if you are using more than one GPU in your computer - you are wasting the others.

The stability issue is the display GPU.
Set the other GPU's to higher intensity.

You are reducing the performance of the other GPU's by using the same settings on all of them that you have determined are needed to keep the display GPU stable.

Kano,

You are not hearing me. Here are some visuals:

I: d,9,9 E: 1130 M: 1000
http://www.fileswap.com/dl/YWZrbwsLyU/d-9-9-1130-1000.png.html

I: d,11,11 E: 1130 M: 1000
http://www.fileswap.com/dl/d6Akac5gdI/d-11-11-1130-1000.png.html

I: 7 E: 1130 M: 1000
http://www.fileswap.com/dl/URJ9zQ5EDs/7-7-7-1130-1000.png.html

I'm just reporting what your software does on Windows with three 7970s not what it should do.

Best RATE and stability is achieved with intensity 7 for ALL GPUs. Maybe it is a bug, but that is what is happening.

Try it for yourself.

BTW, your "gpu" command is working great. I get status of each gpu by parsing your reply json string.
I wrote a little watchdog to monitor and restart cgminer when it crashes or status goes to sick or dead.

Heh

I only wrote the API (and a few patches here and there)

But yeah I didn't see any comments about using multiple intensities - that's why I said that.

Have you tried anything like 7,X,X for any of X=8, 9, 10 or 11?
(since d has other effects also)

That's very strange that the other cards doing nothing in the system except mining would require a setting of 7

Is there anything unusual about your setup?
(obviously not this: but what I mean would be like having 1 display on each card or some other unusual setup)

Edit: I'll also add that a few 100 share test is pretty small ... and could be unreliable.
Give it an hour at least to be sure of the value.

kano

legendary

Activity: 4592

Merit: 1851

Linux since 1997 RedHat 4

Quote from: cablepair on March 05, 2012, 10:20:16 PM

ckvolias:

forgive me if this has already been reported but I have been super busy lately and not able to follow the thread

using latest cgminer 2.3.1-2 win7 64x

after awhile I loose temps and fan speed displays both on screen and from API - a cgminer reboot fixes the issue

screenshot....

http://btcwebhost.com/images/anubistempsgone.png

Yep it is known.
It "appears" to be the ATI ADL library gives up after a long time so it's been resigned to the can't fix basket.
Of course it could be some obscure hard to find cgminer bug, but if it was and since it takes such a long time to happen, it's not likely to be found through any testing looking for the cause of it.

cablepair

hero member

Activity: 896

Merit: 1000

Buy this account on March-2019. New Owner here!!

ckvolias:

forgive me if this has already been reported but I have been super busy lately and not able to follow the thread

using latest cgminer 2.3.1-2 win7 64x

after awhile I loose temps and fan speed displays both on screen and from API - a cgminer reboot fixes the issue

screenshot....

outsidefactor

newbie

Activity: 22

Merit: 0

Not sure if this has been covered in previous posts, but MAN, 227 pages of posts to read, and searches don't turn up anything, so what the hey.

Just went from 2.2.6 to 2.3.1 and went from mid 500s on each 7970 and dropped to... 25.

At first I assumed it was to do with the new GPU interval setting it was annoying me about, but after wasting an hour playing with that to no avail I started looking for other causes. The first thing I noticed was it seemed to be taking 99% of each GPU to turn up 25 Mhash/sec. I tried phatk and went to 150 per GPU. POCLBM came in first with 552 at 12 intensity (I don't want the fans offensively loud).

So, not sure what's changed. I have not tried other kernels in the past so I can't tell if phatk has always sucked or, like Diablo, has just started sucking.

Anyway, I hope this helps anyone else with a 7970 confused by a sudden performance hit after an update!

outsidefactor

kano

legendary

Activity: 4592

Merit: 1851

Linux since 1997 RedHat 4

Quote from: ?? on ??

...
Try it before you laugh.

The point he has already stated is that if you are using more than one GPU in your computer - you are wasting the others.

The stability issue is the display GPU.
Set the other GPU's to higher intensity.

You are reducing the performance of the other GPU's by using the same settings on all of them that you have determined are needed to keep the display GPU stable.

DeathAndTaxes

donator

Activity: 1218

Merit: 1079

Gerald Davis

Quote from: ?? on ??

So you go and try it on WINDOWS with multiple 7970s and post your findings.

Why would I buy 7970s just to test out your flawed theory?

Try turning the SINGLE GPU connected to a display to dynamic as suggested in the README and you can increase the intensity to where it should be without issue. As far as 5 threads being optimal? LOLZ.

Vbs

hero member

Activity: 504

Merit: 500

Quote from: os2sam on March 05, 2012, 11:38:19 AM

Is the SDK 2.5 in 11.11 differenet/better than what shipped with 11.9?

They are all different between SDK packs or Catalyst versions. As for "better", that's more difficult to answer... Tongue

Awhile ago I checked some of them:

Code:

Package - Version Number
------------------------
SDK2.4  - SDK 2.4.595.10
11.6    - SDK 2.4.650.9   <- Newest 2.4
11.7    - SDK 2.5.684.213
SDK2.5  - SDK 2.5.684.213
11.8    - SDK 2.5.709.2
11.9    - SDK 2.5.732.1
11.10   - SDK 2.5.775.2
11.11   - SDK 2.5.793.1   <- Newest 2.5
SDK2.6  - SDK 2.6.831.4
11.12   - SDK 2.6 (10.0.831.4)
12.1    - SDK 2.6 (10.0.851.4)

Panda Mouse

member

Activity: 88

Merit: 10

Gliding...

Quote from: ?? on ??

Quote from: DeathAndTaxes on March 05, 2012, 10:40:49 AM

Quote from: ?? on ??

What is your intensity setting? I bet it is more than 7.

Intensity is the most critical setting for stability. Set the intensity low than play with the rest.

No reason Intensity needs to be less than 7 on a 7970. I run 8 on 5970 and that is only because I am using p2pool. With conventional pool intensity 9 is more appropriate. conman I believe found optimal intensity at 10 or 11 on 7970.

That was Linux. My Windows 7/GD70 experience is different. I was running 9, 10, 11, tried various other settings but I was constantly crashing cgminer. I thought that it was my core/mem/vddc settings, but by accident I found out that on WINDOWS, intensity is the most critical setting.

7 is optimal on Windows with multiple overclocked cards. Less than 7, runs well, but the hash rate suffers.
More than 7, CPU and GPU overloads are likely. Depending what else is installed and running.

On Windows 7, overclocked cards with 9+ sooner or later get "idle for 60 seconds", "too busy GPU event log" etc. That is on a clean, mean
and well tuned system. It might take an hour, it might take few hours, but within one day cgminer crashes when run with -I 9/10/11 and overclocked core/memory.

Now, with -I 7 and 5 threads, my GPU loads are constant at 99%.

BTW, cgminer gets C5 exception (werfault.exe) when GPU gets overloaded and restarted by the BIOS.

You might be lucky with your "-I 9" for a while until the system crashes in the middle of the night and you lose 10 hours of run time :-(
Running with more than 7 creates more problems than it solves. Hash rate does not improve by much as GPUs are at 99% already, so what is the point. To crash cgminer?

Thank a lot, I'll use your experience at once :-) Workss great (Win 7, 1170/1350 -I 7).

DeathAndTaxes

donator

Activity: 1218

Merit: 1079

Gerald Davis

Quote from: ?? on ??

You might be lucky with your "-I 9" for a while until the system crashes in the middle of the night and you lose 10 hours of run time :-(
Running with more than 7 creates more problems than it solves. Hash rate does not improve by much as GPUs are at 99% already, so what is the point. To crash cgminer?

GPU being at 99% has nothing to do with how much of that is USEFUL WORK (i.e. shares)
I have a windows workstation w/ 3x 5970 water cooled which has been running for 7 months now.

If you are driving a display it is a good idea to set that GPU (AND ONLY THAT GPU) to d intensity. Too high of intensity on the GPU used by the OS for driving the display can cause system instabilities. I set 1 core to d and the rest to 9 and the rig runs fast and efficient without crashes. This is heavily overclocked and overvolted. Lowering the intensity of all GPUs even those never used for anything other than mining is simply wasteful.

5 threads per GPU? What exactly do you think that accomplishes?

os2sam

legendary

Activity: 3583

Merit: 1094

Think for yourself

Quote from: Vbs on March 05, 2012, 06:50:20 AM

Quote from: os2sam on March 04, 2012, 10:23:06 PM

I just ran it up the flag pole. Catalyst 11.9 which has SDK 2.5 still has the 100% CPU utilization bug, which was what I kind of thought. I'll do some benchmarking to compare with what I was getting with the 11.6 and 100% CPU Utilization.
Sam

Since you are doing some benchies, also try this:
1) Run Catalyst 12.1 installer -> Custom install -> Unselect All button -> Select GPU Driver only -> Next... Untill it's installed
2) Run Catalyst 11.11 installer -> Custom install -> Unselect All button -> Select SDK Runtime only -> Next... Untill it's installed
3) Done!

That should give you a driver without the 100% CPU bug and the last version of SDK 2.5 runtime.

That's kind of my plan. I'm tinkering with 11.9 now and will then go to 11.11 and 12.1 per yours and others suggestions. I want to do some testing of my own before I settle on the combination you suggest.

Is the SDK 2.5 in 11.11 differenet/better than what shipped with 11.9?

I'm still perplexed that 11.6 has 100% utilization on Win7 but not on WinXP. I guess I'll have to let that one go

Thanks,
Sam

Panda Mouse

member

Activity: 88

Merit: 10

Gliding...

Quote from: ?? on ??

Quote from: Panda Mouse on March 05, 2012, 10:04:59 AM

Quote from: DeathAndTaxes on March 05, 2012, 09:50:04 AM

Quote from: Panda Mouse on March 05, 2012, 09:41:03 AM

I found interesting problem:

I have mobo Gigabyte with PCIe 1x16x and 3x1x, then I used resers.
I used 4xVTX7970.
Procesor Sempron 145 have not enauch power to pull all 4 with their 670MHs/s.
When I use 2 cards - there is no problem at 1190/1250

I think that is something wrong with cgminer (with diablominer engine). Processor usage even with 2 cards is 100%.
Pure diablominer in this configuration use 7-15% processor.

Is this only my problem ?
A better - 4 core AMD X4 FX-4100 will solve this situation?

First. COME ON MAN. Copying 1500 lines just to add " Mac OS X?" really? Please be respectful and edit down your asininely long quote.

As far as CPU. There is something else going on. I run 4x heavily overclocked 5970s (3.3 GH/s total) off the same Sempron 145 which I have underclocked and undervolted down to 1.2GHz. cgminer uses about 10% cpu time. I switched to diablo kernel to test it and it didn't change.

Which OS, which driver, which SDK?

First - I'm very sorry, never again :-) (1500 lines quote).
Second - Windows 7, AMD Catalyst 12.2, 2.6.
I've heard that started from 12.1, the Catalyst has an error and uses 100% of processor.
Unfortunately with my 7970 i have to use 12.2 :-(

What is your intensity setting? I bet it is more than 7.

Intensity is the most critical setting for stability. Set the intensity low than play with the rest.
I run Sempron downclocked with MSI "Cooling" profile, with 3 x 7970, CPU rarely gets above 30% with intensity set to 7 for all three GPUs.
I get a steady ~660 Mh/s per card. No more CPU spikes, "idle for 60 seconds" errors.

BTW, I run 5 threads per GPU as it gives me the right, steady load in MY setup.

Yes - you are right, mamy thanks. But The00Dustin have right too. When I increase Intensity > 9 then processor allways rich 100%.
I used 1960/1350, two threads (there is no difference when I used 5) per GPU, Powertune 0, I have 4x680, stable (till now :-()

The00Dustin

hero member

Activity: 807

Merit: 500

Quote from: DeathAndTaxes on March 05, 2012, 10:40:49 AM

conman I believe found optimal intensity at 10 or 11 on 7970.

However, in Windows intensity needs to be 9 or lower due to cpu usage. With the 7970, I'm not sure there are a lot of options, but PandaMouse may want to read the last 10 pages (or more) to see, as this stuff (7970s and Windows CPU bug versions, not necessarily together) was discussed pretty heavily at some point.

DeathAndTaxes

donator

Activity: 1218

Merit: 1079

Gerald Davis

Quote from: ?? on ??

What is your intensity setting? I bet it is more than 7.

Intensity is the most critical setting for stability. Set the intensity low than play with the rest.

No reason Intensity needs to be less than 7 on a 7970. I run 8 on 5970 and that is only because I am using p2pool. With conventional pool intensity 9 is more appropriate. conman I believe found optimal intensity at 10 or 11 on 7970.

kano

legendary

Activity: 4592

Merit: 1851

Linux since 1997 RedHat 4

Quote from: Panda Mouse on March 05, 2012, 09:41:03 AM

...

PLEASE, as D&T suggested, edit that post of yours and REMOVE/DELETE the ENTIRE quote.

Panda Mouse

member

Activity: 88

Merit: 10

Gliding...

Quote from: DeathAndTaxes on March 05, 2012, 09:50:04 AM

Quote from: Panda Mouse on March 05, 2012, 09:41:03 AM

I found interesting problem:

I have mobo Gigabyte with PCIe 1x16x and 3x1x, then I used resers.
I used 4xVTX7970.
Procesor Sempron 145 have not enauch power to pull all 4 with their 670MHs/s.
When I use 2 cards - there is no problem at 1190/1250

I think that is something wrong with cgminer (with diablominer engine). Processor usage even with 2 cards is 100%.
Pure diablominer in this configuration use 7-15% processor.

Is this only my problem ?
A better - 4 core AMD X4 FX-4100 will solve this situation?

First. COME ON MAN. Copying 1500 lines just to add " Mac OS X?" really? Please be respectful and edit down your asininely long quote.

As far as CPU. There is something else going on. I run 4x heavily overclocked 5970s (3.3 GH/s total) off the same Sempron 145 which I have underclocked and undervolted down to 1.2GHz. cgminer uses about 10% cpu time. I switched to diablo kernel to test it and it didn't change.

Which OS, which driver, which SDK?

First - I'm very sorry, never again :-) (1500 lines quote).
Second - Windows 7, AMD Catalyst 12.2, 2.6.
I've heard that started from 12.1, the Catalyst has an error and uses 100% of processor.
Unfortunately with my 7970 i have to use 12.2 :-(

Topic: OFFICIAL CGMINER mining software thread for linux/win/osx/mips/arm/r-pi 4.11.0 - page 625. (Read 5805546 times)