Bitcoin Forum
June 26, 2024, 04:25:45 PM *
News: Latest Bitcoin Core release: 27.0 [Torrent]
 
   Home   Help Search Login Register More  
Pages: [1]
  Print  
Author Topic: [NEWBIE] CGMiner doesn't detect my two gpus (scrypt mining)  (Read 2284 times)
bernardbot (OP)
Newbie
*
Offline Offline

Activity: 5
Merit: 0


View Profile
May 06, 2013, 03:17:54 PM
 #1

Hello guys,
I have my shiny new rig and I decided to mine using Ubuntu 12.04. I actually found this forum really helpful, and would like to reply to some post explaining my configuration for auto-start services and /usr/bin/screen. It works like a charm from that point of view and I am really happy with it.

The only problem I have, apart from overheating, is that cgminer doesn't recognize one of the two cards, disabling it at startup.
The message it spits out is:

Code:
 [P]ool management [G]PU management [S]ettings [D]isplay options [Q]uit
 GPU 0:  41.0C 2861RPM | OFF   / 0.000h/s | A:0 R:0 HW:0 U:0.00/m I:19
 GPU 1:  70.0C 2932RPM | 359.3K/1.572Mh/s | A:0 R:0 HW:0 U:0.00/m I:19
--------------------------------------------------------------------------------

 [2013-05-06 00:37:27] Started cgminer 3.1.0
 [2013-05-06 00:37:27] Probing for an alive pool
 [2013-05-06 00:37:29] Network diff set to 26M
 [2013-05-06 00:37:29] Error -5: Enqueueing kernel onto command queue. (clEnqueu
eNDRangeKernel)
 [2013-05-06 00:37:29] GPU 0 failure, disabling!
 [2013-05-06 00:37:29] Thread 0 being disabled

As you can see, to keep the temperature low, I just mine at around 360-400 kH/s at the moment.
How can I debug the problem? I have already checked the crossfire settings and the cards are NOT in
crossfire at the moment (I read that it is not necessary).

Thanks for your help guys!
haarts
Newbie
*
Offline Offline

Activity: 23
Merit: 0


View Profile
May 07, 2013, 08:35:11 AM
 #2

How much RAM does your system have? I too have this problem, 3x7950 with 2GB RAM. I suspect this is not enough.
bernardbot (OP)
Newbie
*
Offline Offline

Activity: 5
Merit: 0


View Profile
May 07, 2013, 01:41:32 PM
 #3

It is a 4 GB system. I solved the error lowering the --thread-concurrency parameter to 8192...
Now I have another problem though...the engine freq on one card is always stuck to 300MHz...don't know why...
bernardbot (OP)
Newbie
*
Offline Offline

Activity: 5
Merit: 0


View Profile
May 07, 2013, 11:13:39 PM
 #4

And this is the really weird thing (frequencies @ 300MHz  Huh) :

Code:
GPU 1: 302.2 / 211.2 Kh/s | A:3  R:0  HW:6  U:2.03/m  I:15
50.0 C  F: 6% (1016 RPM)  E: 300 MHz  M: 150 Mhz  V: 0.950V  A: 0% P: 0%
Last initialised: [2013-05-07 18:51:27]
Intensity: 15
Thread 1: 357.8 Kh/s Enabled ALIVE

Below with the conservative conf with --thread-concurrency 8192, -I 15 (going to 84-85 very quickly on GPU 0).
Code:
 [P]ool management [G]PU management [S]ettings [D]isplay options [Q]uit
 GPU 0:  73.0C 3023RPM | 190.6K/233.5Kh/s | A:0 R:0 HW:0 U:0.00/m I:15
 GPU 1:  49.0C 2316RPM | 190.9K/241.9Kh/s | A:1 R:0 HW:0 U:3.82/m I:15

Here another weird thing. All the load on the second (the 0 is OFF), but the sensor reports 93 for the first.

Code:
 [P]ool management [G]PU management [S]ettings [D]isplay options [Q]uit
 GPU 0:  93.0C 4799RPM | OFF   /129.1Kh/s | A:2 R:0 HW: 6 U: 0.53/m I:15
 GPU 1:  53.0C 1017RPM | 524.8K/300.1Kh/s | A:4 R:0 HW:15 U: 1.06/m I:19
Pages: [1]
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!