where in miner avg speed?
4ch/h ... 10ch/h are the speed values. 10ch/h * 24 is chains per days (cpd) value, but it takes several days of mining to get this value correctly. You can use 7ch/h value instead, check my message above how to estimate your cpd value.
|
|
|
Today I calculated some stats for my rigs for last four days:
Rig #1: 0.098 10ch/h per one card, x24 = 2.35 cpd. Rig #2: 0.077 10ch/h per one card, x24 = 1.84 cpd. Rig #3: 0.089 10ch/h per one card, x24 = 2.13 cpd. Rig #4: 0.067 10ch/h per one card, x24 = 1.6 cpd.
So average performance of v2.0 on a single 290X is about 1.98 chains per day. Every rig is 4770 CPU (load 25%) and 4 290X cards core=1050MHz, mem=1350MHz, average values 270 TRT, 165K 4ch/h, 115 7ch/h. You can estimate your cpd value using your 4ch/h and 7ch/h values. I will add estimated and average cpd values in miner stats in next version to simplify calculations.
|
|
|
I'm mining good with 3 rigs - R9 270,but sometime it suddenly stopped the driver then i must restart the rig then run the bat file again,any comment with that weird error ? ( donated 10% also ![Cheesy](https://bitcointalk.org/Smileys/default/cheesy.gif) ) Try to disable overclocking and see if it still happens, that's the most probable reason.
|
|
|
trying to mine on my i7's hd 4000 graphics without luck. the device listed as pci bus 0, device 2, function 0 my 7950 is pci bus 1, device 0, function 0
so i would assume di would need to be set to 2 since 0 runs on my 7950, however says device can not be found. anyone have a clue how to make this work?
Miner supports only AMD cards, please read readme file.
|
|
|
Even in new version of XPM client there is option to send some part of block found to address specified.
Anyway I dont think we will have solo miner fast, people are greeeeeeeedy ;-d and it seems there is hard co-operation of our beloved dev. and ypool for their 13%; dont forget that if miner is for only 1 pool that have bigger profit. I dont know for others are doing, but for me solo mining would mean more than double profit.
"-donationaddress" won't work, because there is no way to check from the miner if you use this option. I'm not cooperated with ypool in any way, I selected it because I already mine there for a few months and this is the only pool that supports dev fee.
|
|
|
@Claymore: Are you considering getting donations for releasing a solo miner version? I think a lot of people would want to solo mine with your miner and that would be healthy for the network.Also, do you consider supporting older cards, like 5xxx series? Ty!
No, sorry, I don't want to make it solo right now, may be later.
|
|
|
The default miner config is too heavy for my 7870.
My 7950 are working with no hiccup and desctop is workable at the same moment.
My 7870 on the other hand is realy strugling, like mining scrypt at extreme difficulty.
Can I edit the .bat in order to make it alittle lighter?
Right now you cannot do it, but I will add "-hc" option in next release to specify number of hashes that are processed at once, it will help.
|
|
|
I uploaded "Claymore GPUs Control Utility v1.0" to my MEGA folder. I know that there are more powerful tools, but I use this minimal utility to control GPUs temp and OC them. Perhaps it will be useful for some people. AMD GPUs only.
|
|
|
Have a very strange question! The same workers I 3X280X, how YPOOL shows 4 of 5 workers were digging?
Just wait a minute or two. It can show more workers for some time if you disconnected/closed miners hard and started again.
|
|
|
Yes, I forgot to mention it previously, I tried it shortly yesterday, result is pretty much the same. At the moment I have actually 2 cards in one rig and 3 card in other and 4ch value is pretty much same for all. So, it is kind of interesting, that amount of GPUs per CPU is not limiting speed but still there is somekind of bottleneck.
I don't have 1-core CPU, so I cannot test it. But 99% the bottleneck is 1-core CPU. If I set affinity to limit miner to 1 core on my CPU it still shows good values, but it can be because opencl kernel use other cores as well.
|
|
|
My question/point from preavious posts was, bottleneck for rig 2 and 3 is CPU?! Is it still possible to somehow improve without buyng better processor?
Try to start just one card and see 4ch/h value on it.
|
|
|
Yes, AMD Sempron 145 handles 3 GPU with ss 1(70% load), its just that card get less 4ch/h. Although it handles GPUs, i guess it is still slower and only gives 75k per 280x, when faster processor rig gives 115k.
Press "m", what numbers do you see? Also make sure you use Catalyst v13.12, it is very important.
|
|
|
I have 3 280x cards with i5 processor and I get 115k 4ch and on other rig I have also 3 280x card but with amd sempron 145 processor and with -ss 1 I get 75k per card. Is there any way to improve it. Like changing some parameters?
I dont get those who put -nofee 1, really cheap. And you still come here for advice? Thats the people who ruin crypto.
Personal experience, you need at least a quad core cpu to run properly. Since the miner runs default of 4 threads. I will try to change threads then. Even 1-core CPU can handle multiple threads without any problems. The only problem is the speed of such CPU, it will execute threads one by one. So if you have 2-cores CPU you can set -t 2 or -t 4, both settings will have same speed if CPU already loaded completely. What CPU load do you have? If 100%, the problem is that your CPU is too slow to handle 3 280X cards. You can set two cards for XPM mining and use third card for something else. PS. Actually even latest Celerons can handle up to 4 280X cards in v2.0. May be sempron is even slower, I don't know about AMD CPUs performance much.
|
|
|
What if remove 4 of 6 cards, i.e. leave only two cards? Do they both work properly? In my tests I used up to 4 cards, never tried >4. I removed 4 cards,one plug in pcie 16x running @ 60k 4ch,another plug in pcie 1x @ 0k 4ch and GPU useage 0%. Whatever, only the card plug in pcie 16x get work. Well, now the only option is to reinstall Windows, it helps sometimes, see message above from "gilad215".
|
|
|
@Clay If you need other information I'd glad to help.
What if remove 4 of 6 cards, i.e. leave only two cards? Do they both work properly? In my tests I used up to 4 cards, never tried >4.
|
|
|
I have 2 * 280x Asus DC2T and 1 280x Sapphire Vapor-X. i5 processor. 13.2 AMD driver. Is this normal? 4ch/h for each is around 21k. Running 2.0v for 4 hours.
It is not normal, check screenshots of other people here, for your card you should get about 120K of 4ch/h, you get only 21K. Reinstall catalyst v13.12, reboot computer and disable overclocking, it helps in 99% cases
|
|
|
@claymore:
1) Could you make option to send 6-ch shares too?
2) I have 4x280X rig When I set GPU_MAX_ALLOC_PERCENT=100 after some time (from seconds to few minutes, sometimes longer) some of miners start to eat RAM without end. Then they ofc crash. I'm talking about Giga Bytes of RAM. Without setting that variable all works just fine, but efficiency is lower.
Yes, I have correct drivers, etc.
3) What should be gpu core/memory settings? Core as high and memory as low as it could be done, or like with scrypt?
PS. Good work, rly good work :-)
1. Ypool does not accept 6ch shares. 2. I saw this issue only twice on one of my rigs. It is very difficult to duplicate, so it is not fixed yet. But hope I will find the reason soon. 3. Core clock is the most important thing. But don't overclock too much.
|
|
|
Is it necessary to have: setx GPU_MAX_ALLOC_PERCENT 100 setx GPU_USE_SYNC_OBJECTS 1
?
At starting miner will tell you about GPU_MAX_ALLOC_PERCENT when necessary. If you don't see any message anything about it (it means that it will show "HashCnt = 512") - everything is ok already.
|
|
|
So it is imperative to be on 24/7? To say it simple If I break operation for say 3 hours per day, will I lose more than 3 hours work? Moreover I am on 14.3 and everything is working fine, should i revert to 13.12?
If you don't mine for 3 hours you will lose earning only for 3 hours. Anyway you can install 13.12 and compare mining speed to 14.3. Then select best driver version. Or just keep current version that you use if you think it's ok for you.
|
|
|
|