SILENTARMY v5: Zcash miner, 115 sol/s on R9 Nano, 70 sol/s on GTX 1070

[zawawa]

SAv5 works on windows7 and on r9280x. 75-80sol/s. same hashrate with claymore but without fees.

Update:

Needs further testing. My initial conclusion is that Claymore finds more solution/shares than SAv5 running at the same speed. Can anyone confirm this?

I would love to find out, too. My suspicion is that this is a pool-specific issue as I did not experience it, but I may be wrong. I will definitely take a closer look at it.

[1714628039]

1714628039

[1714628039]

1714628039

[1714628039]

1714628039

[beeboop]

https://i.imgur.com/PM3s0Pp.png 6 rx470 8GB

yeah 1070 is clearly not optimized btw what is your consumption, mine is 580w

About 700w

[laik2]

SAv5 works on windows7 and on r9280x. 75-80sol/s. same hashrate with claymore but without fees.

Update:

Needs further testing. My initial conclusion is that Claymore finds more solution/shares than SAv5 running at the same speed. Can anyone confirm this?

I would love to find out, too. My suspicion is that this is a pool-specific issue as I did not experience it, but I may be wrong. I will definitely take a closer look at it.

Python code needs serious polishing.
More verbose logging and statistics(jsonrpc implementation). GPU temperatures and specifying -nv -amd probable options for api.
Which means basicly that I will have to learn  python within a few hours because monitoring is crucial for me...

[mrb]

I think in terms of speed claymore and sav5 are almost the same, but with share finding, I can conclude that claymore does find more. I tracked all the shares with SA5 and compare it with the logs on claymore miner and the difference is 27% for 24 hours.

Either you are wrong, or you are using a fork of silentarmy that is broken/buggy. I can only vouch for code and binaries obtained from https://github.com/mbevand/silentarmy

It's very easy to verify: because pools compute the sol/s based on the number of shares submitted, if the pool shows you an average sol/s rate that matches what silentarmy reports then it means silentarmy is finding the expected number of shares.

Also you can't compare the # of shares between different runs of a miner (even different runs of the same miner), because many pool use vardiff which means different miners mining at the same sol/s rate will find a different number of shares that depends on the target sent to the miner.

[oslak]

SAv5 works on windows7 and on r9280x. 75-80sol/s. same hashrate with claymore but without fees.

Update:

Needs further testing. My initial conclusion is that Claymore finds more solution/shares than SAv5 running at the same speed. Can anyone confirm this?

I would love to find out, too. My suspicion is that this is a pool-specific issue as I did not experience it, but I may be wrong. I will definitely take a closer look at it.

I dont think cygwin add latency or had any issues with SA-solver. I use the same difficulty level for comparison and can conclude that claymore does find more.

[jk_14]

yeah 1070 is clearly not optimized btw what is your consumption, mine is 580w

About 700w

~540W from the wall for 6x1070, ~80sol/s each, powerlimit 75W


btw: quick install procedure for KopiemTu (linux rig for newbies):
https://bitcointalk.org/index.php?topic=520998.msg16858580#msg16858580

[laik2]

yeah 1070 is clearly not optimized btw what is your consumption, mine is 580w

About 700w

~540W from the wall for 6x1070, ~80sol/s each, powerlimit 75W


btw: quick install procedure for KopiemTu (linux rig for newbies):
https://bitcointalk.org/index.php?topic=520998.msg16858580#msg16858580

There is another way and I am currently testing it...more pip dependencies is the - here...

[oslak]

I think in terms of speed claymore and sav5 are almost the same, but with share finding, I can conclude that claymore does find more. I tracked all the shares with SA5 and compare it with the logs on claymore miner and the difference is 27% for 24 hours.

Either you are wrong, or you are using a fork of silentarmy that is broken/buggy. I can only vouch for code and binaries obtained from https://github.com/mbevand/silentarmy

It's very easy to verify: because pools compute the sol/s based on the number of shares submitted, if the pool shows you an average sol/s rate that matches what silentarmy reports then it means silentarmy is finding the expected number of shares.

Also you can't compare the # of shares between different runs of a miner (even different runs of the same miner), because many pool use vardiff which means different miners mining at the same sol/s rate will find a different number of shares that depends on the target sent to the miner.

Thanks Marc for clarifying. There is no certainty to this other than the one I use, for windows, is pre-compiled. But it shows same hashrate on the same static difficulty. I set a fix difficulty for comparison. Will post further if percentage improves.

[zawawa]

I think mrb is right in saying that things should be OK as long as the reported and effective speeds match, and that has been the case with my Windows builds. I'm running some tests just to be on the safe side, though.

[oslak]

I think mrb is right in saying that things should be OK as long as the reported and effective speeds match, and that has been the case with my Windows builds. I'm running some tests just to be on the safe side, though.

Thanks very much. I also want to find out more.

[mrb]

IT IS POSSIBLE TO ADD SUPPORT TO www.miningrigrentals.com?

IM GETTING INVALID USERNAME/PASSWORD

nm no idea  tried and get the same thing .

What is their stratum endpoint? I can't find it after 30 sec looking around www.miningrigrentals.com

[jk_14]

What is their stratum endpoint? I can't find it after 30 sec looking around www.miningrigrentals.com

Hi, mrb

Could you implement some very usefull, ccminer-like option: -S --syslog

  -S, --syslog          use system log for output messages

that would be nice for KopiemTu support, i.e. for monitoring

[oslak]

@mrb your latest silentarmy can reach 102sol/s on 280x. very good performance increase.

[adaseb]

@mrb your latest silentarmy can reach 102sol/s on 280x. very good performance increase.

So those hacked bios's really break the 100Sol/s barrier? Nice.

Hynix or Elpida?

Must suck having a nano which is like triple the cost and almost the same speed.

[oslak]

@mrb your latest silentarmy can reach 102sol/s on 280x. very good performance increase.

So those hacked bios's really break the 100Sol/s barrier? Nice.

Hynix or Elpida?

Must suck having a nano which is like triple the cost and almost the same speed.

Yes hack bios from you. 
102 Hynix.
90+ elpida

[laik2]

What is their stratum endpoint? I can't find it after 30 sec looking around www.miningrigrentals.com

Hi, mrb

Could you implement some very usefull, ccminer-like option: -S --syslog

  -S, --syslog          use system log for output messages

that would be nice for KopiemTu support, i.e. for monitoring

So far I manged to get it colored and renamed dev to GPU:

Received job "0e68c2a973b2d174fa4d"
Total 178.8 sol/s [GPU:0 87.3, GPU:1 84.9] 57 shares

Change line 262 to:
            info_gpus = ', '.join(['GPU:\033[91m%s\033[0m \033[93m%.1f\033[0m' % (x[0], x[1])
And line 265 to:
            print("Total \033[92m%s\033[0m sol/s [%s] \033[92m%d share%s\033[0m" % \

Google ANSI python color codes and replace them as you wish... logging is still desirable but hard to implement due to lack of knowledge of python.

[nerdralph]

Have anyone try this miner https://github.com/Optiminer/OptiminerZcash ?

yup...faster than SAv5. But I'm sure SA will catch up soon

It's not going to be tomorrow.  SA5 has no more easy optimizations left; the current version maxes out the memory bandwidth of the card.  Additional significant performance improvements are going to require big changes to the algorithm.  My top idea involves reducing NR_SLOTS to 8 (or maybe even 6 or 4) from 12, and adding a collision overflow table for each round.  By reducing the table size it reduces the time to modify the table (shifting the indexes and updating the XOR'd hashes) and could improve performance by 50-75%, putting SA in the performance range of OptiminerZcash.  This idea would require changes in almost every part of the kernel including ht_store, equihash_round, and expand_refs, and would require many hours of coding.

[toptek]

IT IS POSSIBLE TO ADD SUPPORT TO www.miningrigrentals.com?

IM GETTING INVALID USERNAME/PASSWORD

nm no idea  tried and get the same thing .

What is their stratum endpoint? I can't find it after 30 sec looking around www.miningrigrentals.com

i made miner and tried to rent it to help this what they give you .


equihashminer -s us-east01.miningrigrentals.com:3333 workername -u

If your software is not listed: mining software follows the above command line example, select one that includes the algorithm you are mining and check its readme on how to select the algorithm into the command line.

Please allow up to 3 minutes for your miner to establish its connection

Please use at least one backup server in your local config (if possible):

    eu-de01.miningrigrentals.com:3333
    us-west01.miningrigrentals.com:3333
    us-central01.miningrigrentals.com:3333
    eu-01.miningrigrentals.com:3333
    ap-01.miningrigrentals.com:3333

[du44]

Have a trouble with miner - launching and see:
Connected to the pool
Mining on 2 devices
Pool sent first job
And nothing more - no first share or hashrate - no one more symbol.. Please help

[zawawa]

I think mrb is right in saying that things should be OK as long as the reported and effective speeds match, and that has been the case with my Windows builds. I'm running some tests just to be on the safe side, though.

Thanks very much. I also want to find out more.

There is no notable discrepancies so far between speeds reported by miners and effective speeds on the pool side. I will continue testing for the next 24 hours just in case.