tomashas
Newbie
Offline
Activity: 8
Merit: 0
|
|
March 11, 2014, 10:50:22 PM |
|
Did you try gpu-thread: 2 and xintensity with your hw-configuration and one of the experimental kernels in sgminer ? You can reach over 90% of your hashrate for WU:
I have got 1.18MH/s and 1100WU 1.40MH/s and 1297WU
I noticed there's a /kernel folder with a couple files in there (alexkarnew.cl alexkarold.cl ckolivas.cl psw.cl zuikkis.cl) but to be honest I wouldn't know how to tell sgminer to use one of them. My config is: "intensity" : "20", "vectors" : "1", "worksize" : "512", "kernel" : "scrypt", "gpu-dyninterval" : "7", "queue" : "1", "retry-pause" : "5", "scan-time" : "30", "shaders" : "2816", "thread-concurrency" : "24550", "gpu-powertune" : "18", "gpu-thread" : "1", "sharethreads" : "32", "lookup-gap" : "2", "gpu-vddc" : "1.000", "gpu-engine" : "950", "gpu-memclock" : "1500" I'm pretty happy the graphs on Clevermining are now a lot more realistic, I guess patience isn't my strongest point http://dump.chiko.nl/g/a82334.png
|
|
|
|
pengoau
|
|
March 11, 2014, 10:54:15 PM |
|
--kernel zuikkis
Gives me good performance. Also you need a bit of time for WU to stabilise.
|
|
|
|
tomashas
Newbie
Offline
Activity: 8
Merit: 0
|
|
March 11, 2014, 11:09:18 PM |
|
--kernel zuikkis
Gives me good performance. Also you need a bit of time for WU to stabilise.
Took it for a spin, this happened http://dump.chiko.nl/g/a4e744.png I do like how those .cl files have a function called shittify()
|
|
|
|
FCR
Newbie
Offline
Activity: 31
Merit: 0
|
|
March 11, 2014, 11:16:43 PM |
|
The "only" config you can change for this pool is intensity and even then its not possible (e.g. 280x).
People pointing at your config, where totally underestimating how overloaded the (only) server was.
Some people had already good configs and after trying elsewhere got good performance and only when they were here they had high rejects. Higher than the expected 10%~ due to the coins that are mined...
Whats a bad config? What setting on the miner did you change? You know that queue, scan time and expiry have no affect and its all driven by the server.
In my subtle way and thanks to Kalroth, it was proved that really the only setting available to you is intensity.
Actually to clarify, Kalroth verified that cgminer code deferred to server derived settings for expiry, queue and scantime when those values are available versus what is set in your config/bat files. What gets buried in this thread is his statement that he doesn't know whether Clevermining's servers provide these settings as part of the stratum handshake. In truth, most Altcoin pools do NOT pass these settings down to the client. So these settings can actually still be set on the clientside and have some effect since there are no server settings provided by the pool to default to. Kalroth - feel free to chime in here to help clarify your statements further. If you don't specify them the client uses the default setting, which if set by the server is used. After test I found using various settings including these didn't really help. Once I changed to the Asia server, performance improved. Proof is in the pudding, the US server at the time was overloaded it had nothing to do with a "bad config". I was being reasonable to, I was willing to put up with no more than 10% rejects. That's actually reversed... the way the code is written is that the server settings takes precedence. This was the point that Kalroth was making. But that only holds true if the settings are actually delivered by the server, which in the case of this pool and quite a few others (I dare say most Altcoin pools), they are not. At that point it's either your custom settings or the defaults builtin to CGMiner/SGMiner. Just because they had minimal effect in your particular setup doesn't mean they won't work for someone else, especially when the server is not setting your q, e, s for you. I can't say for sure if they work on here, but I was getting fairly high rejects, usually higher than average, before I added q, s, e. After I added them, my rejected shares went down considerably to 2.4%. Now on US East server I am at 0.3% rejected shares after 24 hours with those settings. The average on that server seems to be 2-3%. This is just my observation on my own rigs; it may or may not be coincidence. I am using queue 5 scan-time 1 expiry 1. I picked these because these are the suggested settings on the Middlecoin wiki.
|
|
|
|
Kifissia
Member
Offline
Activity: 81
Merit: 10
|
|
March 11, 2014, 11:50:00 PM |
|
I am using queue 5 scan-time 1 expiry 1. I picked these because these are the suggested settings on the Middlecoin wiki.
How do you add these commands? in the .conf file? Can you paste these commands here please? Also, is it possible to use these commands in a .bat file? If so, what are the commands? Thanks.
|
|
|
|
tomashas
Newbie
Offline
Activity: 8
Merit: 0
|
|
March 11, 2014, 11:57:02 PM |
|
How do you add these commands?
Add: "scan-time": "1", "expiry": "1", "queue": "5",
To your cgminer.conf file.
|
|
|
|
pengoau
|
|
March 12, 2014, 12:05:21 AM |
|
fwiw i'm using the default values and not specifying expiry, scantime or queue either via config file or cmd line arguments.
AFter 45mins I got 100% accepted....
I'm in Australia using the Asia server.
I believe the --low-mem option is helping as it won't cache shares. My logic is that it make sure to get the most up to date share as its not caching anything.
|
|
|
|
Kifissia
Member
Offline
Activity: 81
Merit: 10
|
|
March 12, 2014, 01:12:10 AM |
|
Thanks to both of you. I ll try both settings.
|
|
|
|
feofhweufheuiwhfef
Newbie
Offline
Activity: 25
Merit: 0
|
|
March 12, 2014, 01:42:22 AM |
|
I noticed that in about 1 month, BTC per day on CleverMining has gone down over 50%.
I was planning to set up a mining rig, but seeing a 50% drop in 1 month makes me seriously reconsider this idea.
Does anyone know why profitability has gone down so much, and any predictions for the future?
Thanks
|
|
|
|
zneww
|
|
March 12, 2014, 01:43:24 AM |
|
I noticed that in about 1 month, BTC per day on CleverMining has gone down over 50%.
I was planning to set up a mining rig, but seeing a 50% drop in 1 month makes me seriously reconsider this idea.
Does anyone know why profitability has gone down so much, and any predictions for the future?
Thanks
READ AND YOU SHALL FIND YOUR ANSWER.
|
|
|
|
Cryptos2go
Member
Offline
Activity: 81
Merit: 10
|
|
March 12, 2014, 02:04:12 AM |
|
I noticed that in about 1 month, BTC per day on CleverMining has gone down over 50%.
I was planning to set up a mining rig, but seeing a 50% drop in 1 month makes me seriously reconsider this idea.
Does anyone know why profitability has gone down so much, and any predictions for the future?
Thanks
I would reconsider, seriously, what you have been planning on considering. I believe this reply should be amicable, don't you agree?
|
|
|
|
adrenaline_x
Newbie
Offline
Activity: 6
Merit: 0
|
|
March 12, 2014, 02:11:32 AM |
|
Nope. the pool can have high rejects compared to other pools.. But this pool stats can sometime also be off. over the past week i went back to a different pool with less stats and lower mh/s / btc and even with the lower published stats of mh/s to btc on that pool versus the higher rate on this pool, i made more on the other pool.. Try another pool after a few days and see what the difference is.
|
|
|
|
pengoau
|
|
March 12, 2014, 02:22:20 AM |
|
Couldn't help myself... now rejects are about 32% with the defaults after an hour.
Will trial:
"scan-time": "1", "expiry": "1", "queue": "5",
|
|
|
|
Kifissia
Member
Offline
Activity: 81
Merit: 10
|
|
March 12, 2014, 02:40:01 AM |
|
Couldn't help myself... now rejects are about 32% with the defaults after an hour.
Will trial:
"scan-time": "1", "expiry": "1", "queue": "5",
Hmm.. My WUs dropped after these settings. I ll try the low mem command.
|
|
|
|
pengoau
|
|
March 12, 2014, 02:45:54 AM |
|
Set queue to 0 as with 5 I was getting 40%+ rejects.
|
|
|
|
Kifissia
Member
Offline
Activity: 81
Merit: 10
|
|
March 12, 2014, 03:02:52 AM |
|
Set queue to 0 as with 5 I was getting 40%+ rejects.
"scan-time": "1", "expiry": "1", "queue": "0", This??? I just changed queue on my own to 0 What about expiry? I thought we dont want our work to expire after 1 sec because the server accepts some stales. Anyway, 1-1-0 settings?
|
|
|
|
pengoau
|
|
March 12, 2014, 03:05:38 AM |
|
Good point, I'll just leave the expiry at its default of 28 seconds.
So I'm just using --scan-time 1 --queue 0
Your settings are ok Kifissia
|
|
|
|
Kifissia
Member
Offline
Activity: 81
Merit: 10
|
|
March 12, 2014, 03:07:13 AM |
|
The default for cgminer is 120sec. And I have no god damn clue what scan-time is
|
|
|
|
pengoau
|
|
March 12, 2014, 03:10:32 AM |
|
I'm using sgminer v4.1 Fork of cgminer made for gpus. Get it here: https://mega.co.nz/#!rVAT1C7T!hY5TQnUQybBLWzPSglaAZCJlxTHIT0W1pvFjubcfVuQ Thats my version that I made an installer for. Run it thru an antivrius.
|
|
|
|
ibfragalot
Member
Offline
Activity: 98
Merit: 10
|
|
March 12, 2014, 03:11:25 AM |
|
Can someone explain this for me?
> After several hours I can't find any stats for my username(bitcoin address) ?
> My 900 khs system is showing 5Mh/s!?
(5s):5.155M (avg):5.105Mh/s | A:0 R:0 HW:0 WU:0.1/m
Thanks.
|
|
|
|
|