evolvia31
Member
Offline
Activity: 74
Merit: 10
|
|
April 26, 2014, 12:33:48 PM |
|
nicehash.com
When there pool switch to currency i have the problem.
|
|
|
|
pjcltd
Legendary
Offline
Activity: 1778
Merit: 1003
NodeMasters
|
|
April 26, 2014, 03:10:33 PM |
|
And you read the warnings Popular cgminer 3.7.2 is not supported due to the bug in the code. You have to use newer miner. Read FAQ to learn more about it. Found here https://nicehash.com/?p=gstarted
|
|
|
|
evolvia31
Member
Offline
Activity: 74
Merit: 10
|
|
April 26, 2014, 06:04:46 PM |
|
Ok that's recent this warning, i don't see it this week. But there are a big problem with the gridseed G-Blade. The only soft which is 100% compatible is fork from jmordica, all other have bug because G-Blade have 40 chip and not 5 I try to patch bfgminer to support 40 chip but i'm not good programmer and my patch don't work very well, pool detect 7,8Mh/s against 10,6Mh/s with the jmordica fork of cgminer.
|
|
|
|
Blisk
|
|
April 27, 2014, 08:34:11 AM |
|
I have two blades and with bfgminer they work only on 2.5 to 2.6 MHz and also one of them after 10 minutes goes DEAD and after restarting says can not write to com port 7 I use 4 port HUB not powered.
With cgminer there is a big problem to make it working.
I use it on windows so is there any advice what to do? Frequency is 800Mhz
|
|
|
|
darkfriend77 (OP)
|
|
April 27, 2014, 09:21:39 AM |
|
I have two blades and with bfgminer they work only on 2.5 to 2.6 MHz and also one of them after 10 minutes goes DEAD and after restarting says can not write to com port 7 I use 4 port HUB not powered.
With cgminer there is a big problem to make it working.
I use it on windows so is there any advice what to do? Frequency is 800Mhz
cgminer works flawless on a raspberry ... super stable at 5.5 on the pool
|
|
|
|
salab
Newbie
Offline
Activity: 21
Merit: 0
|
|
April 27, 2014, 11:22:18 AM |
|
Hi
I played with my blades for few hours, read ton of online materials but did not find solution for my problem.
I connected blades to my home Windows server and the problem is when I start cgminer it does not detect blades. I have to unplug the blades, start cpuminer and then plug usb cables from blades quite fast so cgminer detect the hardware. It is frustrating as server is in the part of the house I have a long walk to and going there to do such stuff manually instead of remote desktop :/
|
|
|
|
evolvia31
Member
Offline
Activity: 74
Merit: 10
|
|
April 27, 2014, 03:59:52 PM |
|
without tweak in GC3355.c file BFGminer isn't ready to fully works with gridseed G-Blade,. Bfgminer support the first gridseed(300Kh/s) but not completely the G-blade (5.5Mh/s).
The reason is first gridseed use 5 chips GC3355 on board, against 40 chips by board with the G-Blade.
I work on a patch but i'm not a guru programmer so actually my patch don't work very well.
I push it to github when it's done.
|
|
|
|
GenTarkin
Legendary
Offline
Activity: 2450
Merit: 1002
|
|
April 27, 2014, 08:13:57 PM |
|
I have two blades and with bfgminer they work only on 2.5 to 2.6 MHz and also one of them after 10 minutes goes DEAD and after restarting says can not write to com port 7 I use 4 port HUB not powered.
With cgminer there is a big problem to make it working.
I use it on windows so is there any advice what to do? Frequency is 800Mhz
cgminer works flawless on a raspberry ... super stable at 5.5 on the pool not for me =( ... id give anything to have cgminer working on my rpi but the USB bus resets after a few minutes of mining and miners stop =( I can only get minerd(cpuminer) to work stable on the USB ports since it seems to communicate differently through them or something... but cpuminer is like 1MH slower and for some reason it doesnt seem to report the pool accepting all the shares generated. Like 2 nonce found messages would scroll past but only 1 accepted at pool ... with no reject meessage either. Happens a few times a minute. So, the avg reported pool hashrate by minerd is like 1-2MH less than cgminer 3.7.2 hosted on a laptop. I wish I could try bfgminer see if it works more stable on the rpi The hub I have it all plugged into is powered as well. So, I dont know wtf the deal is =(
|
|
|
|
darkfriend77 (OP)
|
|
April 27, 2014, 10:04:07 PM |
|
...
not for me =( ... id give anything to have cgminer working on my rpi but the USB bus resets after a few minutes of mining and miners stop =( I can only get minerd(cpuminer) to work stable on the USB ports since it seems to communicate differently through them or something... but cpuminer is like 1MH slower and for some reason it doesnt seem to report the pool accepting all the shares generated. Like 2 nonce found messages would scroll past but only 1 accepted at pool ... with no reject meessage either. Happens a few times a minute. So, the avg reported pool hashrate by minerd is like 1-2MH less than cgminer 3.7.2 hosted on a laptop. I wish I could try bfgminer see if it works more stable on the rpi
The hub I have it all plugged into is powered as well. So, I dont know wtf the deal is =(
I used a fresh debian install ... and the orginal cgminer from andreed ... wrong hashrate ... but doesn't matter ... and changed the cmdline ... slub_flag or something like this ... and it works stable 48h and more
|
|
|
|
GenTarkin
Legendary
Offline
Activity: 2450
Merit: 1002
|
|
April 28, 2014, 02:12:31 AM |
|
...
not for me =( ... id give anything to have cgminer working on my rpi but the USB bus resets after a few minutes of mining and miners stop =( I can only get minerd(cpuminer) to work stable on the USB ports since it seems to communicate differently through them or something... but cpuminer is like 1MH slower and for some reason it doesnt seem to report the pool accepting all the shares generated. Like 2 nonce found messages would scroll past but only 1 accepted at pool ... with no reject meessage either. Happens a few times a minute. So, the avg reported pool hashrate by minerd is like 1-2MH less than cgminer 3.7.2 hosted on a laptop. I wish I could try bfgminer see if it works more stable on the rpi
The hub I have it all plugged into is powered as well. So, I dont know wtf the deal is =(
I used a fresh debian install ... and the orginal cgminer from andreed ... wrong hashrate ... but doesn't matter ... and changed the cmdline ... slub_flag or something like this ... and it works stable 48h and more Yeah I tried that flag, same behaviour =( ... oh well guess its fucked.
|
|
|
|
Blisk
|
|
April 28, 2014, 07:01:38 AM |
|
I have noticed alot of duplicated shares by gridseed blade, have anyone else that problem?
|
|
|
|
sandor111
|
|
May 01, 2014, 01:38:48 AM |
|
Anyone using cpuminer? Care to post your results?
|
|
|
|
GenTarkin
Legendary
Offline
Activity: 2450
Merit: 1002
|
|
May 01, 2014, 02:02:41 AM |
|
Anyone using cpuminer? Care to post your results?
I tried it and its way slower than cgminer 3.7.2 =( ... at least using my pi as host, which come to find out it was having major USB issues w/ the gblades... so maybe thats why it was slow.
|
|
|
|
tc61
|
|
May 01, 2014, 02:21:34 AM |
|
Sandor111,
First thanks for all the work.
So far the new 90a work fine with my usb pod gridseeds, but it has a few problems with my gridseed blade.
1. Each 40 chip board starts off fine, 10-30 minutes in one will slowly start to slow down until no hashes are reported, then the other will stop also. 2. I am running the Pods and the Blade on separate instances. 3. Is there a way to run them both on the same instance , where the 10 pods can be auto tuned and the two 40 chip blades can be set for a hard freq?
Thanks
TC
|
|
|
|
sandor111
|
|
May 01, 2014, 02:42:21 AM |
|
Sandor111,
First thanks for all the work.
So far the new 90a work fine with my usb pod gridseeds, but it has a few problems with my gridseed blade.
1. Each 40 chip board starts off fine, 10-30 minutes in one will slowly start to slow down until no hashes are reported, then the other will stop also. 2. I am running the Pods and the Blade on separate instances. 3. Is there a way to run them both on the same instance , where the 10 pods can be auto tuned and the two 40 chip blades can be set for a hard freq?
Thanks
TC
@Anyone with G-Blades: If you could upload the full miner log with debug output, that would be helpful. For now, seperate instances is the only way, but it's a good idea to specify which devices should be autotuned. I'll think about adding that.
|
|
|
|
tc61
|
|
May 01, 2014, 03:00:58 AM |
|
Sandor111,
Can you tell me how to capture the debug for you?
|
|
|
|
DocGonzzo
Newbie
Offline
Activity: 27
Merit: 0
|
|
May 02, 2014, 06:32:31 PM |
|
If anyone with a blade is willing to try this out, I have been working to include blade support on Scripta, a RPi controller for Scrypt hashing. Here is link to my latest attempt to incorporate the blade: https://mega.co.nz/#!YZMRUZgA!aCRXr1W860jcic5bl12fbSRpHZwSX0WEgmu4PcuZlDEThe image is unzipped at 15gb, but that is just the size card I was using, it should easily fit on 8gb This is still in alpha/beta development, but without a blade of my own it is tough to test. Here is the repo for the development branch of Scripta: https://github.com/DocGonzzo/scriptaAnd here is the original Scripta announcement: https://litecointalk.org/index.php?topic=9908.0Please be aware that although the BTC cores should not be running, it would probably be wise to keep an eye on the device temp when running this distro
|
|
|
|
sandor111
|
|
May 02, 2014, 09:54:33 PM |
|
Anyone tried the latest cpuminer (0.9c)? Should be working fine with G-Blades now.
|
|
|
|
|
digitizing
Member
Offline
Activity: 108
Merit: 10
|
|
May 03, 2014, 12:52:59 AM |
|
Is this one blade or 20 unit of gridseed blade (total 40 blade) ?
|
|
|
|
|