[ANN] 1GH.COM - HVC / MAX / MMC / PTS / VIA Anonymous Pool ★ - page 8.

hdzw

member

Activity: 80

Merit: 10

Quote from: reorder on March 27, 2014, 10:31:39 AM

Quote from: braxx on March 27, 2014, 10:29:30 AM

ccminer 0.5 is out... hashspeeds are much better than amd cards... 750ti reports 13Mh/s (that is better than a 280x)

@reorder: sounds like a challenge to optimize cgminer better...
don't let nvidia beat amd Cheesy

greetings

It is underway, this is for non-oc 280x:

Code:

GPU 1: 68.0C 2848RPM | 14.54M/13.70Mh/s | R:0.1% HW:0 WU:0.0/m T:1 xI:105

However, I am still busy adjusting vardiff.

I still get 11M for 280x. Can you share your config to reach 14M?

braxx

sr. member

Activity: 392

Merit: 250

Quote from: reorder on March 27, 2014, 10:31:39 AM

Quote from: braxx on March 27, 2014, 10:29:30 AM

ccminer 0.5 is out... hashspeeds are much better than amd cards... 750ti reports 13Mh/s (that is better than a 280x)

@reorder: sounds like a challenge to optimize cgminer better...
don't let nvidia beat amd Cheesy

greetings

It is underway, this is for non-oc 280x:

Code:

GPU 1: 68.0C 2848RPM | 14.54M/13.70Mh/s | R:0.1% HW:0 WU:0.0/m T:1 xI:105

However, I am still busy adjusting vardiff.

np

love your work Cheesy

reorder

sr. member

Activity: 462

Merit: 250

Quote from: braxx on March 27, 2014, 10:29:30 AM

ccminer 0.5 is out... hashspeeds are much better than amd cards... 750ti reports 13Mh/s (that is better than a 280x)

@reorder: sounds like a challenge to optimize cgminer better...
don't let nvidia beat amd Cheesy

greetings

It is underway, this is for non-oc 280x:

Code:

GPU 1: 68.0C 2848RPM | 14.54M/13.70Mh/s | R:0.1% HW:0 WU:0.0/m T:1 xI:105

However, I am still busy adjusting vardiff.

braxx

sr. member

Activity: 392

Merit: 250

ccminer 0.5 is out... hashspeeds are much better than amd cards... 750ti reports 13Mh/s (that is better than a 280x)

@reorder: sounds like a challenge to optimize cgminer better...
don't let nvidia beat amd Cheesy

greetings

khm

newbie

Activity: 19

Merit: 0

Quote from: reorder on March 27, 2014, 08:46:05 AM

My bad: I have made a wrong adjustment and vardiff stopped upping the difficulty at all, overloading servers with huge number of low-diff shares. This is fixed now. On the bright side, an improved miner is soon to be released.

Now I think it's worse. It goes to extremes from such low diff that it spams server and hashrate drops to 7.xx Mhash (280x) to such high diff that all I get are reject/stales.

khm

newbie

Activity: 19

Merit: 0

Quote from: reorder on March 27, 2014, 09:21:00 AM

Quote from: khm on March 27, 2014, 05:57:57 AM

Quote from: reorder on March 26, 2014, 06:21:50 PM

I have now adjusted vardiff to push up more agressively than down, that should help with this.

It doesn't help. We need ~0.2 diff
Maybe use string in password to force higher diff?

What kind of hashrate is 0.2 optimal for? For ~22Mhs vardiff now settles at about 0.14 and this means 4 shares/min, anything above is probably prone to generating quite a few stales.

I'm seeing diff 4.xx & 2.xx now. It gives 1Mhash/s more per 280x (rig with 4x 280x)

So:
diff <0.1 - 9.60Mhash per 280x
diff 0.1-0.18? - 9.63Mhash per 280x
diff 0.19+ - 10.55Mhash per 280x
diff 2.xx, 4.xx - 11.4x Mhash per 280x

reorder

sr. member

Activity: 462

Merit: 250

Quote from: khm on March 27, 2014, 05:57:57 AM

Quote from: reorder on March 26, 2014, 06:21:50 PM

I have now adjusted vardiff to push up more agressively than down, that should help with this.

It doesn't help. We need ~0.2 diff
Maybe use string in password to force higher diff?

What kind of hashrate is 0.2 optimal for? For ~22Mhs vardiff now settles at about 0.14 and this means 4 shares/min, anything above is probably prone to generating quite a few stales.

reorder

sr. member

Activity: 462

Merit: 250

Quote from: black_angel1 on March 27, 2014, 08:36:40 AM

why i cant connect ?

[2014-03-27 14:35:55] stratum_subscribe timed out
[2014-03-27 14:35:55] ...retry after 30 seconds

now its ok

My bad: I have made a wrong adjustment and vardiff stopped upping the difficulty at all, overloading servers with huge number of low-diff shares. This is fixed now. On the bright side, an improved miner is soon to be released.

black_angel1

newbie

Activity: 20

Merit: 0

why i cant connect ?

[2014-03-27 14:35:55] stratum_subscribe timed out
[2014-03-27 14:35:55] ...retry after 30 seconds

now its ok

khm

newbie

Activity: 19

Merit: 0

Quote from: reorder on March 26, 2014, 06:21:50 PM

I have now adjusted vardiff to push up more agressively than down, that should help with this.

It doesn't help. We need ~0.2 diff
Maybe use string in password to force higher diff?

jakiman

legendary

Activity: 1638

Merit: 1011

jakiman is back!

Quote from: reorder on March 26, 2014, 06:21:50 PM

I have now adjusted vardiff to push up more agressively than down, that should help with this.

I just tried it out and it's not any better unfortunately. (I've left it for good 20mins also)

7850 @ 1GH = 5.38 MH/s (before)
7850 @ 1GH = 5.40 MH/s (now) (diff hovers around 0.010549)
7850 @ ZHP = 6.30 MH/s (fix diff 0.20000)

280x @ 1GH = 10.50 MH/s (before)
280x @ 1GH = 10.50 MH/s (now) (diff hovers between 0.022220 - 0.029832)
280x @ ZHP = 11.49 MH/s (fix diff 0.20000)

ibinsad

sr. member

Activity: 616

Merit: 250

Quote from: reorder on March 26, 2014, 06:21:50 PM

I have now adjusted vardiff to push up more agressively than down, that should help with this.

zhpool still better by now
with 1gh 33Mh/s
with zhpool 40Mh/s

jakiman

legendary

Activity: 1638

Merit: 1011

jakiman is back!

Quote from: reorder on March 26, 2014, 06:21:50 PM

Quote from: jakiman on March 26, 2014, 04:37:57 PM

Quote from: reorder on March 26, 2014, 04:25:15 PM

Quote from: jakiman on March 26, 2014, 04:15:07 PM

I get much higher hash rate (about 10%) across all my GPU mining rigs when mining HVC on zhpool with a higher diff of 0.2+ compared to 1GH.
Is there a possibility to have a higher diff port for the GPU miners on 1GH also?

What is a typical hashrate of one of your rigs where you have the difference? We'd rather adjust vardiff properly than set up multiple ports as this confuses users unnecessarily.

It could be vardiff but starting at 0.2 as I did see it change to 0.4 also. (Not 100% sure)
Here are my results using your latest 3.25 cgminer and waiting at least 30mins to stabilize:

7850 @ 1GH = 5.38 MH/s
7850 @ ZHP = 6.30 MH/s

280x @ 1GH = 10.50 MH/s
280x @ ZHP = 11.49 MH/s

I have 4 rigs and they all got a noticeable boost.

Others are noticing similar:
https://bitcointalksearch.org/topic/m.5914122

I have now adjusted vardiff to push up more agressively than down, that should help with this.

Thank you! I will try this tonight and will let you know.

reorder

sr. member

Activity: 462

Merit: 250

Quote from: jakiman on March 26, 2014, 04:37:57 PM

Quote from: reorder on March 26, 2014, 04:25:15 PM

Quote from: jakiman on March 26, 2014, 04:15:07 PM

I get much higher hash rate (about 10%) across all my GPU mining rigs when mining HVC on zhpool with a higher diff of 0.2+ compared to 1GH.
Is there a possibility to have a higher diff port for the GPU miners on 1GH also?

What is a typical hashrate of one of your rigs where you have the difference? We'd rather adjust vardiff properly than set up multiple ports as this confuses users unnecessarily.

It could be vardiff but starting at 0.2 as I did see it change to 0.4 also. (Not 100% sure)
Here are my results using your latest 3.25 cgminer and waiting at least 30mins to stabilize:

7850 @ 1GH = 5.38 MH/s
7850 @ ZHP = 6.30 MH/s

280x @ 1GH = 10.50 MH/s
280x @ ZHP = 11.49 MH/s

I have 4 rigs and they all got a noticeable boost.

Others are noticing similar:
https://bitcointalksearch.org/topic/m.5914122

I have now adjusted vardiff to push up more agressively than down, that should help with this.

jakiman

legendary

Activity: 1638

Merit: 1011

jakiman is back!

Quote from: reorder on March 26, 2014, 04:25:15 PM

Quote from: jakiman on March 26, 2014, 04:15:07 PM

I get much higher hash rate (about 10%) across all my GPU mining rigs when mining HVC on zhpool with a higher diff of 0.2+ compared to 1GH.
Is there a possibility to have a higher diff port for the GPU miners on 1GH also?

What is a typical hashrate of one of your rigs where you have the difference? We'd rather adjust vardiff properly than set up multiple ports as this confuses users unnecessarily.

It could be vardiff but starting at 0.2 as I did see it change to 0.4 also. (Not 100% sure)
Here are my results using your latest 3.25 cgminer and waiting at least 30mins to stabilize:

7850 @ 1GH = 5.38 MH/s
7850 @ ZHP = 6.30 MH/s

280x @ 1GH = 10.50 MH/s
280x @ ZHP = 11.49 MH/s

I have 4 rigs and they all got a noticeable boost.

Others are noticing similar:
https://bitcointalksearch.org/topic/m.5914122

reorder

sr. member

Activity: 462

Merit: 250

Quote from: jakiman on March 26, 2014, 04:15:07 PM

I get much higher hash rate (about 10%) across all my GPU mining rigs when mining HVC on zhpool with a higher diff of 0.2+ compared to 1GH.
Is there a possibility to have a higher diff port for the GPU miners on 1GH also?

What is a typical hashrate of one of your rigs where you have the difference? We'd rather adjust vardiff properly than set up multiple ports as this confuses users unnecessarily.

jakiman

legendary

Activity: 1638

Merit: 1011

jakiman is back!

I get much higher hash rate (about 10%) across all my GPU mining rigs when mining HVC on zhpool with a higher diff of 0.2+ compared to 1GH.
Is there a possibility to have a higher diff port for the GPU miners on 1GH also?

meyns

newbie

Activity: 31

Merit: 0

Quote from: reorder on March 25, 2014, 07:37:00 PM

Quote from: meyns on March 25, 2014, 07:29:38 PM

Quote from: reorder on March 25, 2014, 07:03:48 PM

Quote from: meyns on March 25, 2014, 07:01:16 PM

Can anyone explain this problem?

Quote

[00:56:43] Log date is now 2014-03-26
[00:56:43] Started cgminer 3.7.3

[00:56:43] Log date is now 2014-03-26
[00:56:43] Started cgminer 3.7.3
[00:56:43] Probing for an alive pool
[00:56:44] Pool 0 difficulty changed to 0.000244
[01:00:54] Waiting for work to be available from pools.
[01:00:54] Setting GPUs to idle performance.

Running Ubuntu 12.04 on 4x HD7950, with latest source from https://github.com/reorder/cgminer_keccak. Using pool 1gh or even nonce-pool gives the same problem.

Just give it a few more minutes: it is compiling the kernel. Apparently you have not only 7950s in the system?

Only 4x 7950. There are no other gpu's, not even on board. After 10 minutes it isn't even started.

In this case it probably hangs starting the compiled kernel on one of cards. The likely cause for that is hardware, but just to be sure please try deleting *.bin files in cgminer directory and starting over. Another possible reason could be a misbehaving compiler, the miner was only tested with Catalyst versions 13.11 beta and newer.

I can't find any .bin files in the directory. I'm running catalyst 13.12 stable.
I'm running 2 gpus on the motherboard and 2 with a usb riser. Disconnecting the 2 usb risers I'm able to start mining HVC through 1GH. With the 2 usb risers I'm unable. Scrypt goes fine btw.

reorder

sr. member

Activity: 462

Merit: 250

Quote from: meyns on March 25, 2014, 07:29:38 PM

Quote from: reorder on March 25, 2014, 07:03:48 PM

Quote from: meyns on March 25, 2014, 07:01:16 PM

Can anyone explain this problem?

Quote

[00:56:43] Log date is now 2014-03-26
[00:56:43] Started cgminer 3.7.3

[00:56:43] Log date is now 2014-03-26
[00:56:43] Started cgminer 3.7.3
[00:56:43] Probing for an alive pool
[00:56:44] Pool 0 difficulty changed to 0.000244
[01:00:54] Waiting for work to be available from pools.
[01:00:54] Setting GPUs to idle performance.

Running Ubuntu 12.04 on 4x HD7950, with latest source from https://github.com/reorder/cgminer_keccak. Using pool 1gh or even nonce-pool gives the same problem.

Just give it a few more minutes: it is compiling the kernel. Apparently you have not only 7950s in the system?

Only 4x 7950. There are no other gpu's, not even on board. After 10 minutes it isn't even started.

In this case it probably hangs starting the compiled kernel on one of cards. The likely cause for that is hardware, but just to be sure please try deleting *.bin files in cgminer directory and starting over. Another possible reason could be a misbehaving compiler, the miner was only tested with Catalyst versions 13.11 beta and newer.

meyns

newbie

Activity: 31

Merit: 0

Quote from: reorder on March 25, 2014, 07:03:48 PM

Quote from: meyns on March 25, 2014, 07:01:16 PM

Can anyone explain this problem?

Quote

[00:56:43] Log date is now 2014-03-26
[00:56:43] Started cgminer 3.7.3

[00:56:43] Log date is now 2014-03-26
[00:56:43] Started cgminer 3.7.3
[00:56:43] Probing for an alive pool
[00:56:44] Pool 0 difficulty changed to 0.000244
[01:00:54] Waiting for work to be available from pools.
[01:00:54] Setting GPUs to idle performance.

Running Ubuntu 12.04 on 4x HD7950, with latest source from https://github.com/reorder/cgminer_keccak. Using pool 1gh or even nonce-pool gives the same problem.

Just give it a few more minutes: it is compiling the kernel. Apparently you have not only 7950s in the system?

Only 4x 7950. There are no other gpu's, not even on board. After 10 minutes it isn't even started.

Topic: [ANN] 1GH.COM - HVC / MAX / MMC / PTS / VIA Anonymous Pool ★ - page 8. (Read 75205 times)