Topic: CCminer(SP-MOD) Modded NVIDIA Maxwell / Pascal kernels. - page 920. (Read 2347659 times)
The problem seems to be that when the CPU is working in the backround, the cards are producing very little.
But there seems to be a bug somewhere. The gpu is stopping quite often:
http://i60.tinypic.com/8xsdpu.png
I can't reproduce this on my system with release 69.http://i60.tinypic.com/8xsdpu.png
But then I only have one card.
Will try to compile the latest commit tomorrow.
Here is what I am getting off commit 1103 compared to commit 1056 after 1500 accepts
GTX960 dropped down to 6140 from 6150.
GTX980 dropped down to 11760 from 11775.
SP one thing that I noticed, after around 200-250 accepts the 980 was peaking out over 11800. I was kind of excited but it settle back to the numbers posted above.
BTW, how do you post screenshots? Do I need to host them somewhere?
You can upload to http://postimage.org/GTX960 dropped down to 6140 from 6150.
GTX980 dropped down to 11760 from 11775.
SP one thing that I noticed, after around 200-250 accepts the 980 was peaking out over 11800. I was kind of excited but it settle back to the numbers posted above.
BTW, how do you post screenshots? Do I need to host them somewhere?
In above post, sp used http://tinypic.com/
Here is what I am getting off commit 1103 compared to commit 1056 after 1500 accepts
GTX960 dropped down to 6140 from 6150.
GTX980 dropped down to 11760 from 11775.
SP one thing that I noticed, after around 200-250 accepts the 980 was peaking out over 11800. I was kind of excited but it settle back to the numbers posted above.
BTW, how do you post screenshots? Do I need to host them somewhere?
GTX960 dropped down to 6140 from 6150.
GTX980 dropped down to 11760 from 11775.
SP one thing that I noticed, after around 200-250 accepts the 980 was peaking out over 11800. I was kind of excited but it settle back to the numbers posted above.
BTW, how do you post screenshots? Do I need to host them somewhere?
When I am running in quiet mode and remove cpu verification, it runs at full speed.
Latest git, the 970 peaks higher but average hashrate is lower.
Maybe the usual scantime issue? Will test as soon as my rig stops crashing (faulty card) :-(
EDIT: no it's not scantime.
Maybe the usual scantime issue? Will test as soon as my rig stops crashing (faulty card) :-(
EDIT: no it's not scantime.
But there seems to be a bug somewhere. The gpu is stopping quite often:
This is my numbers. I got the Windforce G1 970 up to 10MHASH
The asus strix 750ti is also peaking over 5
With a light overclock.
windows (x86 with the head at github without the intensity parameter set)
The asus strix 750ti is also peaking over 5
With a light overclock.
windows (x86 with the head at github without the intensity parameter set)
Lyra2v2 has increased from 0.7 GH/s to 19.88 GH/s today at nicehash.
Mona effect?
Mona effect?
The biggest Mona pool have 18 ghs Lyra2rev2 hashrate...
I have tested all releases of lyra2rev2 with my little rig of 980/2x970/960, win7x64.
I got the best results with versions 62 and 65 around 38900Kh. Followed by release 66, slower by 50Kh. All without I or X parameter.
The difference is most notable on 970.
The best option right now is to use 970 with release 65 and 980/960 with release 69, specifying the -X 30 for 980.
Those are the numbers right now:
BTW the clocks are:
980 G1 gaming: 1450
970 G1 gaming: 1560
970 ASUS strix oc: 1440
960 mini cheap something: 1411
The highest temp and fan rate are on the 960 - 65°C and 65%
Maybe someone finds my data usefull
I got the best results with versions 62 and 65 around 38900Kh. Followed by release 66, slower by 50Kh. All without I or X parameter.
The difference is most notable on 970.
The best option right now is to use 970 with release 65 and 980/960 with release 69, specifying the -X 30 for 980.
Those are the numbers right now:
BTW the clocks are:
980 G1 gaming: 1450
970 G1 gaming: 1560
970 ASUS strix oc: 1440
960 mini cheap something: 1411
The highest temp and fan rate are on the 960 - 65°C and 65%
Maybe someone finds my data usefull
Lyra2v2 has increased from 0.7 GH/s to 19.88 GH/s today at nicehash.
Mona effect?
Mona effect?
I submitted a bether launchconfig in the lyra2v2 algo. (commit 1103)
The gtx 750 (overclocked 1500 core) is up from 3650 to 4400 (source rednow)
The 750ti unchanged
The 970 up from 9200 to 9400 (windforce G1 standard clocks)
950, 960,980,980ti untested.
please build and test
The gtx 750 (overclocked 1500 core) is up from 3650 to 4400 (source rednow)
The 750ti unchanged
The 970 up from 9200 to 9400 (windforce G1 standard clocks)
950, 960,980,980ti untested.
please build and test
Information purpose. Speeds for Lyra2v2 using SP version around commit 1056. All at stock mfg clocks.
EVGA FTW GTX960
6,150 kH/s
EVGA Hydro Copper GTX980
11,775 kH/s
EVGA FTW GTX960
6,150 kH/s
EVGA Hydro Copper GTX980
11,775 kH/s
Lyra2rev2 was added in 1.5.60 release...
Monacoin just hardforked from scrypt to Lyra2v2. Pretty big in Japan..
Ya-ya, Guano Apes - Big In Japan ))) nice song
no, it's Alphaville :-)
https://en.wikipedia.org/wiki/Big_in_Japan_(Alphaville_song)
EDIT: you must be young!
Lord of the boards comes to mind which I find amazing!
Monacoin just hardforked from scrypt to Lyra2v2. Pretty big in Japan..
Ya-ya, Guano Apes - Big In Japan ))) nice song
no, it's Alphaville :-)
https://en.wikipedia.org/wiki/Big_in_Japan_(Alphaville_song)
EDIT: you must be young!
Information purpose. Speeds for Lyra2v2 using SP version around commit 1056. All at stock mfg clocks.
EVGA FTW GTX960
6,150 kH/s
EVGA Hydro Copper GTX980
11,775 kH/s
EVGA FTW GTX960
6,150 kH/s
EVGA Hydro Copper GTX980
11,775 kH/s
Good job sp(and others) with .69
Those are my lyra2rev2 hashrates, all gpus overclocked:
The hashrate is very constant, the temperatures very low compared to quark.
Those are my lyra2rev2 hashrates, all gpus overclocked:
The hashrate is very constant, the temperatures very low compared to quark.
as "kh" in 69 versions and give gtx960 gtx950 to Quark and Lyra2REv2.
p.s. Why in the first post is not to make a sign for speed and do not update it as an update miner?
p.s. Why in the first post is not to make a sign for speed and do not update it as an update miner?
I use English as my second language but I am unable to understand what your question is.
The gtx960 can do around
quark: 10,5 MHASH
Lyra2REv2: ??
The gtx950 can do around
quark: 8,4 MHASH
Lyra2REv2: ??
The lyra hashrate is not tested on the 950 and the 960
as "kh" in 69 versions and give gtx960 gtx950 to Quark and Lyra2REv2.
p.s. Why in the first post is not to make a sign for speed and do not update it as an update miner?
p.s. Why in the first post is not to make a sign for speed and do not update it as an update miner?
You can set the opt_scantime with the '-s' switch. My default is 25. In the DJM34 version it is set to 60.
Does it help on the pool?
ccminer -s 60 -a lyra2v2 --benchmark.
Does it help on the pool?
ccminer -s 60 -a lyra2v2 --benchmark.
Jump to: