os2sam
Legendary
Offline
Activity: 3586
Merit: 1099
Think for yourself
|
 |
June 29, 2013, 03:50:10 AM |
|
both GPU, run OpenCL benchmark, no problem. Latest test:  I read Diablo Readme for GPU vector size, and diablo return opencl worksize 256, and apply it here. I try use other kernel, seem mining no respond at all, like null? this one on Catalyst 13.4, OpenCL 1.2 Uh, I think the "-k diakgcn" is a diagnostic kernel that doesn't actually submit any shares. try using "-k phatk -v 2 -w 256"
|
A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing on usenet and in e-mail?
|
|
|
SleeperUnit
Newbie
Offline
Activity: 52
Merit: 0
|
 |
June 29, 2013, 05:41:46 AM |
|
both GPU, run OpenCL benchmark, no problem.
...
I read Diablo Readme for GPU vector size, and diablo return opencl worksize 256, and apply it here. I try use other kernel, seem mining no respond at all, like null?
this one on Catalyst 13.4, OpenCL 1.2
Uh, I think the "-k diakgcn" is a diagnostic kernel that doesn't actually submit any shares. try using "-k phatk -v 2 -w 256" "diakgcn" is the Diablo GCN kernel, it is designed for GPUs built on GCN technology (Radeon HD7000 serries) CG miner auto-selects the best kernel for your hardware and rarely gets it wrong so just use the default kernel and you'll be fine.
|
|
|
|
Anime4000
Newbie
Offline
Activity: 30
Merit: 0
|
 |
June 29, 2013, 09:04:31 AM |
|
other kernel give cgminer crash, both my Overclocked Series GPU can't mining, but, normal GPU can mining and cgminer work with Catalyst 13.4. Why? because overclocked GPU remove BFI_INT from hardware? I did try disable BFI_INT on phoenix.
|
|
|
|
os2sam
Legendary
Offline
Activity: 3586
Merit: 1099
Think for yourself
|
 |
June 29, 2013, 12:46:47 PM |
|
"diakgcn" is the Diablo GCN kernel, it is designed for GPUs built on GCN technology (Radeon HD7000 serries)
CG miner auto-selects the best kernel for your hardware and rarely gets it wrong so just use the default kernel and you'll be fine.
Ah, OK. At one time there was a diagnostic kernel probably started with diag* instead of diak*. Thanks for the correction. Sam
|
A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing on usenet and in e-mail?
|
|
|
-ck (OP)
Legendary
Offline
Activity: 4522
Merit: 1669
Ruu \o/
|
 |
June 29, 2013, 03:20:02 PM |
|
"diakgcn" is the Diablo GCN kernel, it is designed for GPUs built on GCN technology (Radeon HD7000 serries)
CG miner auto-selects the best kernel for your hardware and rarely gets it wrong so just use the default kernel and you'll be fine.
Ah, OK. At one time there was a diagnostic kernel probably started with diag* instead of diak*. Thanks for the correction. Sam diakgcn is the Diapolo Kernel for GCN. It has nothing to do with diablo and pretty much never outperformed any other kernel so it is never selected automatically by cgminer.
|
Developer/maintainer for cgminer, ckpool/ckproxy, and the -ck kernel 2% Fee Solo mining at solo.ckpool.org -ck
|
|
|
Karin
|
 |
June 29, 2013, 06:32:17 PM Last edit: July 04, 2013, 08:53:50 PM by Karin |
|
[problem solved]
|
|
|
|
bernard75
Legendary
Offline
Activity: 1316
Merit: 1003
|
 |
June 29, 2013, 06:55:46 PM |
|
3.1.1 doesn't use the WinUSB driver for 'Icarus' devices - it uses the serial-USB driver on windows. 3.2* uses WinUSB that Zadig will install for you.
(Icarus devices are: Icarus, Lancelot, Asicminer USB and Cairnsmore1)
|
|
|
|
Anime4000
Newbie
Offline
Activity: 30
Merit: 0
|
 |
June 29, 2013, 07:17:02 PM |
|
By the way, currently I have 4670 GPU, run PCI-E 1.1 x16 (32MH/s with 750MHz > 800MHz clock using cgminer). So, if I move GPU to motherboard that capable PCI-E 2.0 x16, dose Hash rate will increase? i think not
|
|
|
|
Miner-TE
|
 |
June 29, 2013, 07:25:08 PM |
|
By the way, currently I have 4670 GPU, run PCI-E 1.1 x16 (32MH/s with 750MHz > 800MHz clock using cgminer). So, if I move GPU to motherboard that capable PCI-E 2.0 x16, dose Hash rate will increase? i think not
You are correct, It would not. Bandwidth on the PCI bus does not help hash rate.
|
BTC - 1PeMMYGn7xbZjUYeaWe9ct1VV6szLS1vkD - LTC - LbtcJRJJQQBjZuHr6Wm7vtB9RnnWtRNYpq
|
|
|
os2sam
Legendary
Offline
Activity: 3586
Merit: 1099
Think for yourself
|
 |
June 29, 2013, 08:33:04 PM |
|
By the way, currently I have 4670 GPU, run PCI-E 1.1 x16 (32MH/s with 750MHz > 800MHz clock using cgminer). So, if I move GPU to motherboard that capable PCI-E 2.0 x16, dose Hash rate will increase? i think not
I was getting 35Mhs on my AGP 4670. So I would expect you should be getting higher than that already. But no changing the slot/MB won't help.
|
A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing on usenet and in e-mail?
|
|
|
kano
Legendary
Offline
Activity: 4676
Merit: 1858
Linux since 1997 RedHat 4
|
 |
June 29, 2013, 09:49:52 PM |
|
By the way, currently I have 4670 GPU, run PCI-E 1.1 x16 (32MH/s with 750MHz > 800MHz clock using cgminer). So, if I move GPU to motherboard that capable PCI-E 2.0 x16, dose Hash rate will increase? i think not
You are correct, It would not. Bandwidth on the PCI bus does not help hash rate. For BTC sha256 yes.
|
|
|
|
Loft
|
 |
June 29, 2013, 10:45:03 PM |
|
I can't change the voltage across the config file. I have 2 7950 MSI and 2 7870 Gigabyte, but with the help of MSIAfterburner I can't go to change the voltage. Tried using a bat file to change нaпpяeниe avail. Version CGminer 3.1.1, 3.2.1, 3.3.0 What I should do?
|
|
|
|
xgtele
|
 |
June 30, 2013, 06:34:56 AM |
|
I have 2 7970 GPU in my system and they are required a different GPU threads: -g 1 for the first card and -g2 for the second card.
How I can set it correctly in cgminer bat file? Settings like -g 1,2 does not work
|
|
|
|
Krak
|
 |
June 30, 2013, 06:49:42 AM |
|
I have 2 7970 GPU in my system and they are required a different GPU threads: -g 1 for the first card and -g2 for the second card.
How I can set it correctly in cgminer bat file? Settings like -g 1,2 does not work
The only way you can do that is to set the GPU that requires 1 thread to dynamic intensity (-I d).
|
BTC: 1KrakenLFEFg33A4f6xpwgv3UUoxrLPuGn
|
|
|
xgtele
|
 |
June 30, 2013, 07:22:59 AM |
|
I have 2 7970 GPU in my system and they are required a different GPU threads: -g 1 for the first card and -g2 for the second card.
How I can set it correctly in cgminer bat file? Settings like -g 1,2 does not work
The only way you can do that is to set the GPU that requires 1 thread to dynamic intensity (-I d). Hashing speed is lowered 10 times.  I can simply do custom GPU threads for each cards in GUIminer and it is based on cgminer, so it should be solution somewhere around..
|
|
|
|
-ck (OP)
Legendary
Offline
Activity: 4522
Merit: 1669
Ruu \o/
|
 |
June 30, 2013, 07:27:44 AM |
|
I have 2 7970 GPU in my system and they are required a different GPU threads: -g 1 for the first card and -g2 for the second card.
How I can set it correctly in cgminer bat file? Settings like -g 1,2 does not work
The only way you can do that is to set the GPU that requires 1 thread to dynamic intensity (-I d). Hashing speed is lowered 10 times.  I can simply do custom GPU threads for each cards in GUIminer and it is based on cgminer, so it should be solution somewhere around.. I have 2 7970 GPU in my system and they are required a different GPU threads: -g 1 for the first card and -g2 for the second card.
How I can set it correctly in cgminer bat file? Settings like -g 1,2 does not work
You cannot set separate thread counts without running 2 instances. However you can set an intensity for dynamic as well, with gpu-dyninterval.
|
Developer/maintainer for cgminer, ckpool/ckproxy, and the -ck kernel 2% Fee Solo mining at solo.ckpool.org -ck
|
|
|
Anime4000
Newbie
Offline
Activity: 30
Merit: 0
|
 |
June 30, 2013, 11:30:03 AM |
|
both GPU, run OpenCL benchmark, no problem.
this one on Catalyst 13.4, OpenCL 1.2
Try catalyst 13.1 Uninstall any drivers you have cleanly with driversweep, including the opencl installed and start again with what 13.1 installs. Then start with a freshly extracted cgminer or delete any .bin files you have in your cgminer directory. I did uninstall driver, what AMD Installer did, they also uninstall motherboard driver, and restart, blue screen. I go safe mode, sweep left over driver, reboot, blue screen, go safe mode again, remove driver manually on Device Manager, now, install back everything, and OpenCL have problem, https://fbcdn-sphotos-e-a.akamaihd.net/hphotos-ak-ash3/q71/10346_10201359330448163_818314584_n.jpgcgminer still crash on start.
|
|
|
|
|
bitpop
Legendary
Offline
Activity: 2954
Merit: 1065
|
 |
June 30, 2013, 02:34:12 PM |
|
Just wanted to say, 331 doesn't display all devices especially when detected right after startup.
|
|
|
|
tew
Newbie
Offline
Activity: 3
Merit: 0
|
 |
June 30, 2013, 08:50:42 PM |
|
I am interested in trying my hand at creating yet another FPGA cluster hasher, with each cluster connected via a single USB connection to the host (rather than 1 per FPGA).
I've looked at the Cairnsmore 1, Ztex, and Icarus drivers; each one has hard-coded some of the idiosyncrasies of those particular devices. I'm wondering if there is something like a generic CGminer driver skeleton I could fill in / develop against.
Is there a simple block of code that easily links with CGminer, passes up the relevant info from the device chain (estimated duration of one hash-cycle iteration, number of hashes completed per cycle) to the main miner program on initialization, downloads base header data and range to be hashed against from the miner to the device with each getwork or stratum fetch, receives proof of work from the device to send to the mining pool, and sends an interrupt to the device when assigned work becomes invalid?
Given the proliferation of mining devices, there's almost a need for a "miner driver" specification just like the old "packet driver" specification back in the days of Win32s enabling Win3.11 to connect to the internet, so that the people creating the new mining devices would be responsible for driver creation, rather than the mining stack software creators.
Thanks, -Tom
|
|
|
|
|