Gateless Gate Sharp 1.3.8: 30Mh/s (Ethash) on RX 480!

[bytewise]

The thing about SA's Python implementation is that it does not work well at all on Windows.
In fact, the miner got faster by 10% just by switching from SA to sgminer-gm while keeping the same Equihash kernel.
I am not entirely satisfied with sgminer as I said before, and I will probably write my own backend and user-interface in Clojure in a long run.

Python on Windows always has been a less than preferable situation. Haven't personally run silentarmy on anything but Linux so I can't comment as to what the root cause of that would be. The solver and work distributor being separate was what made it have those characteristics I described earlier.

Clojure is an not a great choice imo. It's difficult for outsiders to approach, requires a JVM, and runs slower than Java in benchmarks. It's your thread and project, just my two cents.

Give 4 threads a go one day on sgminer. One card is randomly selected to display roughly double the rate it should be getting (even though the 5s and avg show otherwise) on my 3 card rig.

3 threads and xI 673 give the highest hash rate from my testing.

[1714165140]

1714165140

[1714165140]

1714165140

[1714165140]

1714165140

[1714165140]

1714165140

[1714165140]

1714165140

[1714165140]

1714165140

[cryptominer420]

@zawawa
I will test out gg again in a week or two. I have to find a few blocks to justify testing again on the whole farm, I am however trying to fine tune for 4 card rigs and not having a ton of luck so far. Single card it runs great but my very week cpu's in my 4 card rigs are having issues.

[zawawa]

Single card it runs great but my very week cpu's in my 4 card rigs are having issues.

Ah, high CPU usage explains higher power consumption, too. I will get to it today.

[zawawa]

The thing about SA's Python implementation is that it does not work well at all on Windows.
In fact, the miner got faster by 10% just by switching from SA to sgminer-gm while keeping the same Equihash kernel.
I am not entirely satisfied with sgminer as I said before, and I will probably write my own backend and user-interface in Clojure in a long run.

Python on Windows always has been a less than preferable situation. Haven't personally run silentarmy on anything but Linux so I can't comment as to what the root cause of that would be. The solver and work distributor being separate was what made it have those characteristics I described earlier.

Clojure is an not a great choice imo. It's difficult for outsiders to approach, requires a JVM, and runs slower than Java in benchmarks. It's your thread and project, just my two cents.

Give 4 threads a go one day on sgminer. One card is randomly selected to display roughly double the rate it should be getting (even though the 5s and avg show otherwise) on my 3 card rig.

3 threads and xI 673 give the highest hash rate from my testing.

I will definitely try 3 threads on Linux later. Thanks! As for Clojure, the only performance issue I see is the slow startup time, which should not affect the overall performance of the miner. Also, I am way more productive in this language and its cross-platform support for multi-threading is excellent. The obscurity of the language and JVM requirement are real drawbacks, but they are trade-offs I am willing to make. In any case, switching to Clojure is a long shot, and it will be a sub-project to work on in my spare time, if there is any, for the foreseeable future.

[Ursul0]

Single card it runs great but my very week cpu's in my 4 card rigs are having issues.

Ah, high CPU usage explains higher power consumption, too. I will get to it today.

ghm... don't think so...

[laik2]

The thing about SA's Python implementation is that it does not work well at all on Windows.
In fact, the miner got faster by 10% just by switching from SA to sgminer-gm while keeping the same Equihash kernel.
I am not entirely satisfied with sgminer as I said before, and I will probably write my own backend and user-interface in Clojure in a long run.

Python on Windows always has been a less than preferable situation. Haven't personally run silentarmy on anything but Linux so I can't comment as to what the root cause of that would be. The solver and work distributor being separate was what made it have those characteristics I described earlier.

Clojure is an not a great choice imo. It's difficult for outsiders to approach, requires a JVM, and runs slower than Java in benchmarks. It's your thread and project, just my two cents.

Give 4 threads a go one day on sgminer. One card is randomly selected to display roughly double the rate it should be getting (even though the 5s and avg show otherwise) on my 3 card rig.

3 threads and xI 673 give the highest hash rate from my testing.

I will definitely try 3 threads on Linux later. Thanks! As for Clojure, the only performance issue I see is the slow startup time, which should not affect the overall performance of the miner. Also, I am way more productive in this language and its cross-platform support for multi-threading is excellent. The obscurity of the language and JVM requirement are real drawbacks, but they are trade-offs I am willing to make. In any case, switching to Clojure is a long shot, and it will be a sub-project to work on in my spare time, if there is any, for the foreseeable future.

You can't, linux multithreading is one awful thing...2 threads are more than enough. Optimize rounds for ethash and it will decrease power usage.

[zawawa]

Hmm, interesting... I am traveling around right now, but I should be able to access RX 470 tonight.
I really should get a laptop with an AMD APU..

[UnclWish]

It's too quiet here...   

[zawawa]

I was quite busy as I was traveling abroad to visit my family.
I finally got a few day window to work on the miner.
Let's see what I can do with my father's old computer...

[anatolikostis]

I was quite busy as I was traveling abroad to visit my family.
I finally got a few day window to work on the miner.
Let's see what I can do with my father's old computer...

v0.1.3-pre6b - any news?

[nikisev]

hey guys, what are the best/recommended polaris cards settings for different algos for this miner?
what is it developed on, stock?

[bytewise]

I was quite busy as I was traveling abroad to visit my family.
I finally got a few day window to work on the miner.
Let's see what I can do with my father's old computer...

v0.1.3-pre6b - any news?

Looks like an updated port on the XMR config from https://github.com/zawawawa/gatelessgate/commits/master

[marvykkio]

I was not able to do this from miner to eth,
if some kind soul could post the string of his bat file I copy
 

[Ursul0]

I was not able to do this from miner to eth,
if some kind soul could post the string of his bat file I copy
 

use this config - https://github.com/zawawawa/gatelessgate/blob/master/Core/gatelessgate-Ethereum.conf

[frostminer]

Any solutions for pools with different stratums?

When log says:

[18:10:55] Stratum Error: {
   "code": -3,
   "message": "Method not found"
}
[18:10:55] Closing main socket
[18:10:55] Stratum Error: {
   "code": -3,
   "message": "Method not found"
}
[18:10:55] Closing socket for stratum main

[Ursul0]

                               .
I just need more people to mine with the fee on.
At the current pace, I would be making an impressive $3/month. LOL

https://ethermine.org/miners/91fa32e00b0f365d629fb625182a83fed61f0642/payouts

you wish you would!
I'd say this looks shameful.

[nerdralph]

                               .
I just need more people to mine with the fee on.
At the current pace, I would be making an impressive $3/month. LOL

https://ethermine.org/miners/91fa32e00b0f365d629fb625182a83fed61f0642/payouts

you wish you would!
I'd say this looks shameful.

The problem is performance.  On Linux with my Rx470 I get 29.3Mh with sgminer-gm 5.5.5 (xI:1155) while gg does 29.0 (with xI:2304 to keep HW errors under control).  Claymore gets ~30.3, so we know there is room to improve ethash-new.

[marvykkio]

last miner claymore 9.1 too unstable, loses power from 28.9 down to 21, 19 mh / s, virtually cagae ago, I tried all settings, nothing to do, I do not know whether to blame the miner or the nanoopol,, the fact is that there is more stability in miners, perhaps better than clayore stops doble miner and focuses only on single

claymore no offense, maybe after a while you get to a point that you can no longer go on, then focus only on stability and keep off speed

[Ursul0]

                               .
I just need more people to mine with the fee on.
At the current pace, I would be making an impressive $3/month. LOL

https://ethermine.org/miners/91fa32e00b0f365d629fb625182a83fed61f0642/payouts

you wish you would!
I'd say this looks shameful.

The problem is performance.  On Linux with my Rx470 I get 29.3Mh with sgminer-gm 5.5.5 (xI:1155) while gg does 29.0 (with xI:2304 to keep HW errors under control).  Claymore gets ~30.3, so we know there is room to improve ethash-new.

yep, it is. the dual mine feature is even bigger factor though.

[Ursul0]

last miner claymore 9.1 too unstable, loses power from 28.9 down to 21, 19 mh / s, virtually cagae ago, I tried all settings, nothing to do, I do not know whether to blame the miner or the nanoopol,, the fact is that there is more stability in miners, perhaps better than clayore stops doble miner and focuses only on single

claymore no offense, maybe after a while you get to a point that you can no longer go on, then focus only on stability and keep off speed

not sure if you should try talking to claymore here:)

I had (hopefully that is) various issues with v9 rate at the pool side. For me the switch to a newer version also coincided with moving some of my hash through proxy which made things even more complicated. however now with 9.1 after getting rid of mem errors(it never was an issue for me earlier) completely and adding some more juice to the cards it seems to be doing fine. And well obviously the DCR hash is almost doubled.