Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux) - page 1374.

phzi

hero member

Activity: 700

Merit: 500

Quote from: Claymore on April 20, 2016, 11:50:19 AM

Quote from: phzi on April 20, 2016, 11:31:41 AM

Stratum method is called, "eth_submitHashrate." I posted about this earlier in this thread as well (post #569).

See here (eth-proxy relaying method):
https://github.com/Atrides/eth-proxy/blob/d993fc8db84dc09f25b72a34ce69b6977c88ac64/mining_libs/getwork_listener.py#L58-L62

Edit:
Also see (ethminer creating eth_submitHashrate commands):
https://github.com/ethereum/libethereum/blob/8b389757cfc2757965671fa8e080b32aaf65771b/ethminer/MinerAux.h#L536
https://github.com/ethereum/libethereum/blob/6d65d5d1412013381324826ea583e8b3fa9f5f3c/ethminer/FarmClient.h#L37-L47

Yes I found "eth_submitHashrate" but I'd like to check some miner that sends this command to see how it works. There is no any documentation about it but it seems it has two params: first is the hashrate, second is filled by random and it's strange, it must be worker name so pool can know who reports the hashrate.

I confirmed with dr_pra that the 2nd param is ignored (at least by ethpool/ethermine). If more then one worker connects via a stratum proxy, submitted hashrate doesn't work. If a single worker is used on a connection, the hashrate is assosiated with that worker.

PandaPanda!

newbie

Activity: 35

Merit: 0

@Claymore I have a request.

I've noticed that you have an auto restart option now when one of the GPUs stops talking to the miner. However, when the program restarts, it hangs at the POOL connection part (i'm mining in ETH only mode). What I usually have to do when I encounter this, is close the program, and then restart it manually, then it works.

I have automated batch files to start your miner. Can you please add a command line switch which just quits the program rather than restarting it automatically when a GPU fails? Would be a huge help to me (and probably many others). I love your miner, and praise it on the eth forums. Much thanks

Claymore

donator

Activity: 1610

Merit: 1325

Miners developer

Today some Decred pools have problems, at least suprnova on dashboard page showed me a message about problems in their backend.

qqqq

legendary

Activity: 1596

Merit: 1011

Quote from: avirdoz on April 21, 2016, 07:17:42 AM

Quote from: qqqq on April 21, 2016, 07:09:21 AM

Recently i got on my all pc and all ver of miner on different pools

Code:

Stratum - Cannot connect  to dcr.suprnova.cc port 2252
 DCR: Stratum - Failed to connect, retry in 20 sec...

Code:

Stratum - Cannot connect  to dcr.pool.mn port 4800
 DCR: Stratum - Failed to connect, retry in 20 sec...

and

Code:

DCR: Job timeout, disconnect, retry in 20 sec...

ETH with ethermine pool is ok!

Same here.
Upd. coinmine works.

What the hell, looks like stratum down on that pools simultaneously. Or maybe some new client of DCR was released.

Haryford

newbie

Activity: 61

Merit: 0

Quote from: restless on April 21, 2016, 07:14:44 AM

I tried to use -wd 1, but can't make it to close the miner
tried -r -1
tried -r 0
When I mistype pool name (simulating dns /connectivity problem) it just keeps retrying to connect every 20sec, won't close the miner Sad

My miner does not close. It will hang the system. So I use -wd 0 to disable the watchdog, and -r -1 not to restart.

Claymore

donator

Activity: 1610

Merit: 1325

Miners developer

Quote from: restless on April 21, 2016, 07:14:44 AM

I tried to use -wd 1, but can't make it to close the miner
tried -r -1
tried -r 0
When I mistype pool name (simulating dns /connectivity problem) it just keeps retrying to connect every 20sec, won't close the miner Sad

Watchdog closes/restarts miner only if GPU failed.
I'm working on failover, will be added in a day.

avirdoz

legendary

Activity: 1191

Merit: 1001

Quote from: qqqq on April 21, 2016, 07:09:21 AM

Recently i got on my all pc and all ver of miner on different pools

Code:

Stratum - Cannot connect  to dcr.suprnova.cc port 2252
 DCR: Stratum - Failed to connect, retry in 20 sec...

Code:

Stratum - Cannot connect  to dcr.pool.mn port 4800
 DCR: Stratum - Failed to connect, retry in 20 sec...

and

Code:

DCR: Job timeout, disconnect, retry in 20 sec...

ETH with ethermine pool is ok!

Same here.
Upd. coinmine works.

restless

legendary

Activity: 1151

Merit: 1001

I tried to use -wd 1, but can't make it to close the miner
tried -r -1
tried -r 0
When I mistype pool name (simulating dns /connectivity problem) it just keeps retrying to connect every 20sec, won't close the miner Sad

qqqq

legendary

Activity: 1596

Merit: 1011

Recently i got on my all pc and all ver of miner on different pools

Code:

Stratum - Cannot connect  to dcr.suprnova.cc port 2252
 DCR: Stratum - Failed to connect, retry in 20 sec...

Code:

Stratum - Cannot connect  to dcr.pool.mn port 4800
 DCR: Stratum - Failed to connect, retry in 20 sec...

and

Code:

DCR: Job timeout, disconnect, retry in 20 sec...

ETH with ethermine pool is ok!

Code:

 DCR: Stratum - connecting to 'dcr.pool.mn' <136.243.50.159> port 4800
ETH: 04/21/16-14:13:42 - New job (target 0x0000000112e0be82) from eu1.ethermine
org:4444
ETH - Total Speed: 28.801 Mh/s, Total Shares: 115, Rejected: 0, Time: 04:36
ETH: GPU80 28.801 Mh/s
 DCR - Total Speed: 792.021 Mh/s, Total Shares: 1063, Rejected: 0
 DCR: GPU0 792.021 Mh/s
ETH: 04/21/16-14:13:57 - SHARE FOUND - (GPU 0)
ETH: Share accepted!
ETH: 04/21/16-14:14:00 - New job (target 0x0000000112e0be82) from eu1.ethermine
org:4444
ETH - Total Speed: 28.832 Mh/s, Total Shares: 116, Rejected: 0, Time: 04:36
ETH: GPU0 28.832 Mh/s
 DCR - Total Speed: 792.884 Mh/s, Total Shares: 1063, Rejected: 0
 DCR: GPU0 792.884 Mh/s
Stratum - Cannot connect  to dcr.pool.mn port 4800
 DCR: Stratum - Failed to connect, retry in 20 sec...

it trying to connect and then it show this error. All of this working a few hours ago, i even did.t touching my miners.

restless

legendary

Activity: 1151

Merit: 1001

Quote from: Trimegistus on April 21, 2016, 06:53:57 AM

Quote from: restless on April 21, 2016, 06:52:50 AM

Had anyone made some kind of reliably working "pool failover" with Claymore miner?

I guess he will do it himself in the next release.

I think using "-wd X" when miner is started with
start /wait
and then in bat file anoher start & goto may do the thing :S

DaveyJones

hero member

Activity: 768

Merit: 505

Quote from: Claymore on April 20, 2016, 11:50:19 AM

Quote from: phzi on April 20, 2016, 11:31:41 AM

Quote from: Claymore on April 20, 2016, 11:15:33 AM

v3.1:

- reduced delays at shares sending, it must improve speed on pools a bit.

It is not a major update, just found some additional delays in v3.0 so I uplodaded an update. No other changes.

PS. I see dwarfpool has "Sent hashrate" column. How to make it work? I tried eth-proxy + ethminer but it is still empty. If someone can point me how to make it work for Stratum I will add it to my miner.

Stratum method is called, "eth_submitHashrate." I posted about this earlier in this thread as well (post #569).

See here (eth-proxy relaying method):
https://github.com/Atrides/eth-proxy/blob/d993fc8db84dc09f25b72a34ce69b6977c88ac64/mining_libs/getwork_listener.py#L58-L62

Edit:
Also see (ethminer creating eth_submitHashrate commands):
https://github.com/ethereum/libethereum/blob/8b389757cfc2757965671fa8e080b32aaf65771b/ethminer/MinerAux.h#L536
https://github.com/ethereum/libethereum/blob/6d65d5d1412013381324826ea583e8b3fa9f5f3c/ethminer/FarmClient.h#L37-L47

Yes I found "eth_submitHashrate" but I'd like to check some miner that sends this command to see how it works. There is no any documentation about it but it seems it has two params: first is the hashrate, second is filled by random and it's strange, it must be worker name so pool can know who reports the hashrate.

or you ask the user atrides here on bitcointalk for help... he is the admin of the dwarfpool Wink

also is there a way you can use your old packer too for those 1% of us where it is not running?

Charloz24

hero member

Activity: 966

Merit: 501

V 3.1 now works very well with my 370 rigs. Only thing is that I must not use clocking option (cclock or mclock) because the program stop. I would like to find a way to optimize my hashing...

Trimegistus

legendary

Activity: 1564

Merit: 1027

Quote from: restless on April 21, 2016, 06:52:50 AM

Had anyone made some kind of reliably working "pool failover" with Claymore miner?

I guess he will do it himself in the next release.

restless

legendary

Activity: 1151

Merit: 1001

Had anyone made some kind of reliably working "pool failover" with Claymore miner?

cryptominer_fuad88

newbie

Activity: 3

Merit: 0

Quote from: Q_R_V on April 21, 2016, 06:16:29 AM

Quote from: cryptominer_fuad88 on April 21, 2016, 04:30:40 AM

this is my setting on 3x msi r7 370 2gb

bat conf:
setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100

EthDcrMiner64.exe -epool eth-asia.dwarfpool.com:8008 -ewal 0xwallet -epsw x -dpool http://dcr.pool.mn:4722 -dwal worker -dpsw x

http://imgur.com/EATZZ99

Can you tell us your power consumption?

around 530watt..

Q_R_V

sr. member

Activity: 428

Merit: 250

Inactivity: 8963

Quote from: cryptominer_fuad88 on April 21, 2016, 04:30:40 AM

this is my setting on 3x msi r7 370 2gb

bat conf:
setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100

EthDcrMiner64.exe -epool eth-asia.dwarfpool.com:8008 -ewal 0xwallet -epsw x -dpool http://dcr.pool.mn:4722 -dwal worker -dpsw x

http://imgur.com/EATZZ99

Can you tell us your power consumption?

Trimegistus

legendary

Activity: 1564

Merit: 1027

Quote from: Walrusbonzo on April 21, 2016, 03:13:38 AM

I'm running my 2x7950s at 1130/1700(1.2v/1.65v) and 1190/1700(1.23v/1.65v). They get me 23mh/s and 23.5mh/s respectively.

Usually with a standard 7950 the ROM has different memory timings for different ranges of memory clockspeed. I custom tuned the ROM files for my 7950s with a hex editor and set the same tight timings from 1376MHz to 1750MHz+. That way I know the memory timings are not changing between any clock speed I wish to try.

I could edit your ROMs too if you are comfortable with flashing them. But I would only recommend you do it if your GFX cards have a backup ROM switch in case something goes wrong. You should then find you'll get great performance at 1500MHz and beyond.

You need good cooling to maintain stability with the kind of overclocks I'm doing.

I appreciate the offer but that's not exactly my goal Grin

I don't think the extra heat + extra noise + extra power consumption will be compensated by a few MHs. Besides, I want to have full stability because I don't have the time to nurse my rigs every hour.

My goal was to find settings in line with what I saw mentioned multiples times as a way to save a few watts and increase hashrate at the same time; some kind of underclock.

Anyway, thanks for your sugestions.

cryptominer_fuad88

newbie

Activity: 3

Merit: 0

this is my setting on 3x msi r7 370 2gb

bat conf:
setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100

EthDcrMiner64.exe -epool eth-asia.dwarfpool.com:8008 -ewal 0xwallet -epsw x -dpool http://dcr.pool.mn:4722 -dwal worker -dpsw x

http://imgur.com/EATZZ99

Gataway

newbie

Activity: 54

Merit: 0

Quote from: devlin on April 21, 2016, 03:11:13 AM

Quote from: o00o on April 21, 2016, 02:10:05 AM

Quote from: chup on April 20, 2016, 08:50:28 AM

Quote from: o00o on April 20, 2016, 05:32:55 AM

Quote from: cryptoc19 on April 20, 2016, 04:46:24 AM

Quote from: o00o on April 19, 2016, 06:21:03 PM

I use Sapphire Trixx to downclock my memory to 150 Mhz which also increases hashrate due to tighter timings. Wink

how does downclocking the memory actually help? and, to me, this value seems insane!

are you sure you meant 150 Mhz ?

Quote from: chup on April 20, 2016, 02:48:40 AM

Quote from: o00o on April 19, 2016, 06:21:03 PM

I use Sapphire Trixx to downclock my memory to 150 Mhz which also increases hashrate due to tighter timings. Wink

Using 15.7.1 and Trixx? Are You sure it's underclocked? Checked in Trixx card info or GPU-Z?

I'm absolutely certain, Trixx seems to be the only utility which allows you to do so without any hassle anyway. A wide variety of algorithms benefit from setting the memory to 150 Mhz [600 Mhz effective] such as Quark & Qubit which rely on latency rather than throughput. This sets your memory bus to a strap with tighter timings which will ultimately lower power consumption and heat and also allows for more overclocking headroom. To prove my point, I managed to yield 23 MH/s for Quark, a similar figure for Qubit and 37 MH/s Ethereum & 620 MH/s for Decred in dual mining mode on a 390x without rejected shares.

Has anyone tried Stilit's bios or any other modded bios for that matter on Ethereum?

15.12 Trixx and Win7 x64 I can only overclock, but not downclock. Trixx accepts downclock, pretends that it's done, but if I check in card info or GPU-Z clocks are default values.

MSI Afterburner reports that my memory is indeed running at 150 Mhz once set via Trixx. I have no other utilities present in this system which is why I cannot provide any other kind of confirmation. Try launching a game or anything else which can put a substantial load on your memory bus for that matter, such as Scrypt and you'll immediately notice a drastic decrease in performance. Out of curiosity, I "tried" to play a variety of games including Fallout 4, BF4 and GTA5 and let me say that what was once smooth sailing at 60 FPS is now a slideshow, which eliminated the last shred of doubt. 150 Mhz has proved to be optimal on my 290s, 290xs and 390xs as anything lower or higher seems to offset timings.

By curiosity I set memory for R9 290 to run at 150Mh, speed for ETH is aprox. 3Mh .... You are sure memory is indeed run at 150Mh, look at MSI Afterburner Hardware Monitor graph not in main window.

I do not think the memory shall run 150 MHz to mine the Ethereum. It is OK just for Decred mining.

Walrusbonzo

hero member

Activity: 682

Merit: 500

Quote from: Trimegistus on April 20, 2016, 06:32:34 PM

Guys, how are you doing with v3.1? Any speed improvement?

Can anyone share some optimized settings for the HD 7950? Mine are now at ~18 MHs on stock settings.

Someone mentioned reducing memory clock speed to get better timings, right? I did it but the only result was loosing a few MHs Huh

Currently I have a two HIS 7950 IceQ 3GB GDDR5 but I have several other 7950s on their way.

Any ideas?

Thx!

I'm running my 2x7950s at 1130/1700(1.2v/1.65v) and 1190/1700(1.23v/1.65v). They get me 23mh/s and 23.5mh/s respectively.

Usually with a standard 7950 the ROM has different memory timings for different ranges of memory clockspeed. I custom tuned the ROM files for my 7950s with a hex editor and set the same tight timings from 1376MHz to 1750MHz+. That way I know the memory timings are not changing between any clock speed I wish to try.

I could edit your ROMs too if you are comfortable with flashing them. But I would only recommend you do it if your GFX cards have a backup ROM switch in case something goes wrong. You should then find you'll get great performance at 1500MHz and beyond.

You need good cooling to maintain stability with the kind of overclocks I'm doing.

Topic: Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux) - page 1374. (Read 6590565 times)