SRBMiner Cryptonight AMD GPU Miner V1.9.3 - native algo switching

[dannyprats]

Feature request:

Switch ALGO -on the fly or via remote API  (preferably )- with key combination ALT + # (number defined in file: algo.txt)

Code:

algo.txt:

1: normalv7
2: heavy
3: lite
4: ipbc
....

For this to work: pools.txt should have a structure like this:

Code:

{
"pools" :
[
   "normalv7":
    [
{"pool" : "xmr-eu1.nanopool.org:14444", "wallet" : "4A5hJyu2FvuM2azexYssHW2odrNCNWVqLLmzCowrA57xGJLNufXfzVgcMpAy3YWpzZSAPALhVH4Ed7xo6RZYyw2bUtbm12g.donation", "password" : "x"},
{"pool" : "xmr-eu2.nanopool.org:14444", "wallet" : "4A5hJyu2FvuM2azexYssHW2odrNCNWVqLLmzCowrA57xGJLNufXfzVgcMpAy3YWpzZSAPALhVH4Ed7xo6RZYyw2bUtbm12g.donation", "password" : "x"}
    ],
    "heavy":
    [
{"pool" : "loki.ingest.cryptoknight.cc:7733", "wallet" : "L5YYTvuGdMVDx52AWZK6kwJMvc9bN2Q7FGHkzW2Xd8iReXtRnBXRm9a8pTYB9TD9NDSjQnaRmnFQ1SDvetAjWLvfJVtSY1p.75000", "password" : "rig1"},
{"pool" : "haven.ingest.cryptoknight.cc:5533", "wallet" : "hvxyJGKFHq9XLia5Up4GTuW6cGZhVnUZhgLPqxxeRfXe8H5ZcLYKjYFME9mg679C1LZAQBkv4gdQtg2gyWZ2hYkM475sJKf6sp.75000", "password" : rig1"}
    ],
    "lite":
    [
{"pool" : "aeon.miner.rocks:5555", "wallet" : "eXaMpLeEyXACPVTCQZ2bGjg9k7Y9y8rLAGBRBaAJVDnLLbLGAeGLqvn2k82aXzXnEwXUXh72amhAKoRQ.50000", "password" : "w=rig1"},
    ]
]
}

[doktor83]

Feature request:

Switch ALGO -on the fly or via remote API  (preferably )- with key combination ALT + # (number defined in file: algo.txt)

algo can't really be switched at the moment without restarting miner and re-initialising buffers.

[hesido]

Feature request:

Switch ALGO -on the fly or via remote API  (preferably )- with key combination ALT + # (number defined in file: algo.txt)

Code:

algo.txt:

1: normalv7
2: heavy
3: lite
4: ipbc
....

For this to work: pools.txt should have a structure like this:

Code:

{
"pools" :
[
   "normalv7":
    [
{"pool" : "xmr-eu1.nanopool.org:14444", "wallet" : "4A5hJyu2FvuM2azexYssHW2odrNCNWVqLLmzCowrA57xGJLNufXfzVgcMpAy3YWpzZSAPALhVH4Ed7xo6RZYyw2bUtbm12g.donation", "password" : "x"},
{"pool" : "xmr-eu2.nanopool.org:14444", "wallet" : "4A5hJyu2FvuM2azexYssHW2odrNCNWVqLLmzCowrA57xGJLNufXfzVgcMpAy3YWpzZSAPALhVH4Ed7xo6RZYyw2bUtbm12g.donation", "password" : "x"}
    ],
    "heavy":
    [
{"pool" : "loki.ingest.cryptoknight.cc:7733", "wallet" : "L5YYTvuGdMVDx52AWZK6kwJMvc9bN2Q7FGHkzW2Xd8iReXtRnBXRm9a8pTYB9TD9NDSjQnaRmnFQ1SDvetAjWLvfJVtSY1p.75000", "password" : "rig1"},
{"pool" : "haven.ingest.cryptoknight.cc:5533", "wallet" : "hvxyJGKFHq9XLia5Up4GTuW6cGZhVnUZhgLPqxxeRfXe8H5ZcLYKjYFME9mg679C1LZAQBkv4gdQtg2gyWZ2hYkM475sJKf6sp.75000", "password" : rig1"}
    ],
    "lite":
    [
{"pool" : "aeon.miner.rocks:5555", "wallet" : "eXaMpLeEyXACPVTCQZ2bGjg9k7Y9y8rLAGBRBaAJVDnLLbLGAeGLqvn2k82aXzXnEwXUXh72amhAKoRQ.50000", "password" : "w=rig1"},
    ]
]
}

Changing the pools tree structure is not necessary (and harder for everyone else)
{"pool" : "xxx:111", "algorithm" : "heavy" ...... }
should suffice. Algorithm would be an optional property, as the setting would be taken from config normally.

Since each algo would have different settings, instead of algo, you could also go for:
{"pool" : "xxx:111", "config" : "config-heavy.txt" ... }

This shoots two birds with one stone.

If this is implemented, the job json could also support: "_SRBMiner_Algorithm" : "heavy"  to dynamically switch, so no API with write support is necessary because proxy's can append it in the JSON (The added _SRBMiner_ is to avoid any namespace collisions until a standard can be agreed upon by the miners and pools)

Giving API switching capability is good but it brings added burden to Doktor Proxy's could handle the switch until a "write" API could be programmed by Doktor.

Edit:
Actually this naming convention could be used for algo negotiation:
https://github.com/xmrig/xmrig-proxy/blob/master/doc/STRATUM_EXT.md

Edit2: This is basically an extension to stratum protocol, employed by XMRig, but there's no reason why SRBminer cannot support this! Using watchdog, SRBMiner can easily restart to switch to new algos, this not only provides unattended algo switching for hardforked coins, but also shall provide coin switching using proxy's.

When an algo negotiation is initiated and a new algo is required, watchdog could restart the miner, so no API programming would be necessary. The only problem to solve is to decide a way to do per algo config, which shouldn't be hard, really.

[WolframDeEst]

I don't know this is bug from new driver or from your miner, but my RX 560 identified as RX 470. Make srbminer crash and still not sure what make this happening. I will continue from my side, i am not sure if new driver make my OC not stable or something else make your miner not stable. But still curious about that bug. Thanks!

https://i.imgur.com/aFt5xXK.png

Get some update. Try less agresive OC with more voltage and make thread setting in RX 560 from 2 to 1 make it stable now. Not sure what make my previous setting not stable in this one. I think something in driver make my oc less stable and not sure why i need to make thread in RX 560 become 1.

[hieunguyen]

my old card r9 370 mining with sbrminer is slow and usually restart, can anybody please help me?

[altsay]

my old card r9 370 mining with sbrminer is slow and usually restart, can anybody please help me?

What is your hashrate?

[ripcurrent]

Question for you guys.

I have a new card a Vega 56 Powercolor red devil. Samsung memory.

When I try to apply a powerplay table the card becomes "not a valid gpu" in OverDriveNT tool and pulls a error 43 in device manager.

I never came across this before any thoughts, i have applied the power play tables many times never saw this error on other cards.

Any thoughts.

In addition I find that the 120 intensity for vega 56 and 64 in SBR is the best do you agree?

first try without powertable.

second you sure u use good .reg file and change this:
[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Class\{4d36e968-e325-11ce-bfc1-08002be10318}\0000]

mybe your card not 0000 mybe is 0001 0002 *

vega 56/64bios - v7 intensity/worksize : 112/16 or 120/8
heavy : 56/8
lite 235/8

for other  algo etc arto-aloy see my  old post

Thanks for the help..yes it works fine no power play, I checked the number and it corresponds to the reg file. Do you have a link to a vega 56 powerplay that you know is good I will check this next. Is just strange after I reboot and it read it it in no longer good. Hashes fine under normal stock drives but use a lot of power.

try this:
http://vega.miningguides.com/
http://vega.miningguides.com/p/faq.html#TDRFailure
https://www.reddit.com/r/MoneroMining/comments/74hjqn/monero_and_vega_the_definitive_guide/

Update: Still no go..no matter what powerplay table I try to apply, it throws the cards into "gpu not supported" and error code 43 in the device manager, if I  dont apply the PP table they run fine. I have another rig same cards run perfect...maybe ill do a reinstall of windows...at this point im at a loss.

[Iamtutut]

Windows code 43 means driver issue.

What's your hardware ?

[Larvitar]

Some of these results are not making sense at all.

I have tweaked my Asus RX580 (8GB) according to best guide and not getting over 1000 hash on each.
I dont think its possible unless you having clocked it to have hundreds of memory errors and card then fuck up..

So! If you don't know how to do it then it is impossible?

Didnt say its impossible, just saying with the dodgy overlocks your not considering power draw.
Considering the safety of your cards, then good luck with your power bill vs profit of mining.
Good luck with the life of your cards..

You need good timing straps to reach 1k+ on CN-V7.  8gb cards can do 1k+ on CN-Heavy easily.  
I have also got 1260 H/s on an RX 570 8gb card (1466/2300 clocks).  They do 1140 H/s @ 1250/2240 clocks and 900mV core (105w from wall).  So it's pretty much possible.

Also, All 4gb cards that I have Elpida, Hynix do 1k+ on CN-V7.

Your Micron straps are impressive.

But did you made some "magic" with your Hynix straps? I could just reach 940H/s with custom timings :/

[Kgonla]

I have results...
First of all, I'm using Server 2008 what is about the same than W7
I completely removed my firewall & now:

[2018-06-13 05:04:02] CryptonightV7 mode enabled
[2018-06-13 05:04:44] DevFee pools loaded
[2018-06-13 05:04:44] DevFee pool SET
[2018-06-13 05:04:44] DevFee address SET
[2018-06-13 05:04:44] Starting init of mining threads

Devfee pools are loaded but still taking too long to start, about 42 seconds what is best than previous 78 sec.
This does not happens in W10

Also the self shutdown of SRBminer after 4th reconnect only happens in Server not in W10

The good part of Server is that with the RX 580 4GB I can set intensity to 61 in V7 while in W10 I can only set 59 loosing some H/s

Update:

Now with "DevFee pools loaded" the program does not self close after the 4th reconnect under Server 2008, so it was related to the firewall blocking.
Sounds estrange, I just tested once for now but it must be that.

Update 2:

The program self close randomly, only when max_difficulty is used.
I've seen after 15-20 minutes & near 2 hours. Maybe when the devfee is going to start?

No more tests for now, I'll wait next version to see if it works fine.

[UnclWish]

It's strange, but I allready lowered memory frequency on my 580 by 50MHz and this has no affect on speed... Even 1 h/s no drops. I try to remove calculate errors and figured out that on SRB miner core freq. is more important mem. freq. Other miners such as Claymore or GGS are sensitive to memory clock...

[dingdongtobias]

Anyone played with persistent_memory ?
I got a 3% speedup on v7 with 580 8g hynix mem on i 72, w 8, t 2 and persistent_memory true

[MadKernel]

my old card r9 370 mining with sbrminer is slow and usually restart, can anybody please help me?

Try my Open Source miner : https://github.com/Enoch72/sgminer-MK/releases/tag/MK-1.0.7

Seems that performs very well in older AMD architecture.
The rawintensity to setup in the bat file is (24-1) *16 = 368
If someone with newer cards will test it will be welcome.
PS: I try to write you a private message but seems that you have choose not to allow message from newbie, as I'm.

[UnclWish]

Anyone played with persistent_memory ?
I got a 3% speedup on v7 with 580 8g hynix mem on i 72, w 8, t 2 and persistent_memory true

I checked on 580 8Gb hynix on heavy algo. No difference on enabled/disabled option.

[vmozara]

Guys,

anybody tried new Adrenalin driver 18.6.1 with Vegas? Is there any difference?

[igotek]

I need the gpu settings for SRBMiner config.

My GPU is Sapphire Nitro Radeon RX470 4GB.

[dragonmike]

@Doktor: should we use the same intensity parameters with the new "Haven" algo as we were with "Heavy"? Just trying to set up my config files before the fork goes live...

[ripcurrent]

Windows code 43 means driver issue.

What's your hardware ?

I don't think its a driver error as I can mine fine without the PP tables, but soon as I try to install them the cards pull the errors. Running Power Color Vega 56's with 18.5.1 drivers. Tried a few PP table too and even the safe one from the guide, funny thing is I have another rig with he same card and works fine with the PP

[danskin]

anyone body not seeing CPU mining info in hash statistics (pressing h)?

[kellygeefx]

anyone body not seeing CPU mining info in hash statistics (pressing h)?

Only for AMD GPU. use a CPU miner