[JCE]Fast & stable CN/v8/Heavy/Tube/XHV miner, CPU+GPU, Vega56 1800+ RX580 1200+

[Larvitar]

You are doing much progress with GPU. wow man

Sorry if it's boring, but I have to discuss Awesome Miner again.

Well, I tried today to run JCE "inside" Awesome miner. As you said some pages ago, JCE Miner has some similiarities with CPUminer-OPT API, so I tried to "camuflate" JCE as CPUminer-opt. Not success here

But reading log, I found the solution (or just a piece): The problem to run is the option "--api-remote -b 0.0.0.0:4034". I can set -mport easily, but Awesome don't alow to put off the --api-remote option.

To solve this, just ignore the --api-remote option or accept it as --mport (with IP:port form). Could you do this in the next release?

Thanks a lot man! Your miner is I M P R E S S I V E!

[1714837828]

1714837828

[1714837828]

1714837828

[1714837828]

1714837828

[JCE-Miner]

no problem, i'll ignore the parameter --api-remote. do you need i ignore -b too, with a dotted ip after ?
i think i'll release a last cpu version using the gpu version, but with gpu disabled. then a alpha with gpu+cpu

[whotheff]

Xmr Stack = 130h/s
JCE = 74h/s

huge pages on, avx-aes

on AMD A10-5800K

same pool, same time of day, waited atleast 15 min
to confirm hashrate.

[JCE-Miner]

Nope, JCE is not 50% slower than stak. Maybe not that faster, but not 50% slower.
And that score is very low for a A10, i can get 115 with a core2 quad. I guess you use one thread only, and stak enabled double-hash. It can be enabled on JCE with manual config too.

At least give the exact Stak configuration you used, and i can tell you how to do the same on JCE.

[whotheff]

Nope, JCE is not 50% slower than stak. Maybe not that faster, but not 50% slower.
And that score is very low for a A10, i can get 115 with a core2 quad. I guess you use one thread only, and stak enabled double-hash. It can be enabled on JCE with manual config too.

At least give the exact Stak configuration you used, and i can tell you how to do the same on JCE.

Noticed that XMR stack uses only two cores
(so it can supply them with 2Mb cache each=4MB total for that CPU).
Then disabled two of them in JCE and it jumped to 130h/s.

How to enable double hash? What does it do?

[Larvitar]

no problem, i'll ignore the parameter --api-remote. do you need i ignore -b too, with a dotted ip after ?
i think i'll release a last cpu version using the gpu version, but with gpu disabled. then a alpha with gpu+cpu

Yup!

[JCE-Miner]

thanks for the test, it seems my autoconfig is bad on the A10, it allocates too many threads. I'll fix it, thanks.

The multihash (double hash is the 2-case, you can set from 1- to 6- ) is also called low-power in stak IIRC. It's about using, on one CPU core, twice the register and twice the cache to get sometimes twice the speed. The trick is that it let the other cores free, so it consume less power and allows the Turbo to enable, for CPU with turbo.

Technically, it's good, sure when you want to save power, but also when you run out of cores and not of cache. If you have a CPU with 2 cores but 8M cache, normal config would give only 2x2M = 4M cache used.

you may enable double-hash to use 2x2x2M cache = 8M of cache, and get some extra perf.
It works more or less depending on the CPU. It's very efficient on Ryzen, and not at all on Core2.

I looked closer at the A10, and yeah that's a little APU with little cache.
I give you an experimental config that could let you get some extra perf, but not sure, i cannot test, i've no A10.

"cpu_threads_conf" : 
[ 
     { "cpu_architecture" : "auto", "affine_to_cpu" : 0, "use_cache" : true },     
     { "cpu_architecture" : "auto", "affine_to_cpu" : 1, "use_cache" : false },
     { "cpu_architecture" : "auto", "affine_to_cpu" : 2, "use_cache" : true },
     { "cpu_architecture" : "auto", "affine_to_cpu" : 3, "use_cache" : false },
]

now finishing 0.29c, the last CPU-only version, with some updates.

[Iamtutut]

What would the 0.29C update be ?

For the CPU / GPU miner, will it be possible to split CPU and GPU mining ? I don't mine CN heavy algos at all with the CPU (4MB cache), but I sometimes mine CN heavy with the GPUs.
 So if possible I'd like to mine CN V7 / lite with the CPU separated from any other CN algo with the GPUs.

[JCE-Miner]

good idea. It won't be possible in the earlier versions, but i love the idea.

0.29c updates :

Use of the GPU version frontend, but GPU disabled
Support of Masari fork --variation 11
Support of Haven fork --variation 12
Stellite defaults to XTL variation
ETN defaults to v7
other coins that have forked will default to their respective fork
Coin list in alphabetical order
new coin : BXB
new coin : WOWnero
-b parameter ignored for compatibility with AwsomeMiner
ditto with parameter --api-remote

[whotheff]

thanks for the test, it seems my autoconfig is bad on the A10, it allocates too many threads. I'll fix it, thanks.

The multihash (double hash is the 2-case, you can set from 1- to 6- ) is also called low-power in stak IIRC. It's about using, on one CPU core, twice the register and twice the cache to get sometimes twice the speed. The trick is that it let the other cores free, so it consume less power and allows the Turbo to enable, for CPU with turbo.

Technically, it's good, sure when you want to save power, but also when you run out of cores and not of cache. If you have a CPU with 2 cores but 8M cache, normal config would give only 2x2M = 4M cache used.

you may enable double-hash to use 2x2x2M cache = 8M of cache, and get some extra perf.
It works more or less depending on the CPU. It's very efficient on Ryzen, and not at all on Core2.

I looked closer at the A10, and yeah that's a little APU with little cache.
I give you an experimental config that could let you get some extra perf, but not sure, i cannot test, i've no A10.

"cpu_threads_conf" : 
[ 
     { "cpu_architecture" : "auto", "affine_to_cpu" : 0, "use_cache" : true },     
     { "cpu_architecture" : "auto", "affine_to_cpu" : 1, "use_cache" : false },
     { "cpu_architecture" : "auto", "affine_to_cpu" : 2, "use_cache" : true },
     { "cpu_architecture" : "auto", "affine_to_cpu" : 3, "use_cache" : false },
]

now finishing 0.29c, the last CPU-only version, with some updates.

12:52:22 | Hashrate Thread 0: 59.71 h/s
12:52:22 | Hashrate Thread 1: 2.57 h/s
12:52:22 | Hashrate Thread 2: 54.01 h/s
12:52:22 | Hashrate Thread 3: 2.44 h/s
12:52:22 | Total: 118.71 h/s
12:52:23 | Pool changes Difficulty to 2634.


limited to two threads with -t 2:

12:47:53 | Pool changes Difficulty to 2376.
12:47:59 | Hashrate Thread 0: 62.49 h/s
12:47:59 | Hashrate Thread 1: 60.26 h/s
12:47:59 | Total: 122.74 h/s

[JCE-Miner]

so the no-cache is bad on the A10. It's very good on modern Intel, decent on Ryzen and bad on Core2. And so, bad on A10 too.
So keep the config you had found before, with two normal threads.
I won't have time to fix the A10 autoconfig in next version, will be for 0.30

edit: i'm polishing the GPU version.
I see very strange results on my Pitcairn rig (five different Pitcairns : 7870 and 7850, 1G and 2G...)

On my 2G, when I use all the memory available, I can reach 526h/s on my 7870 2G, and 495 on CN-v7 on my 7850 2G, which is far above what Claymore 9.7 gave on CN-Classic. It's even the best score ever I reached, period. Maybe i didn't configure SRB well, but so far i didn't reach such an hashrate.

But on my little 7850 1G, that's terrible, i peak at 343 where Claymore gave a whooping 450 !
It looks like my code is faster when using the double-memory, but a lot slower otherwise. And double-memory is barely possible on 1G cards. It works on Bonaire 1G because it has a little GPU that can be saturated with two threads using ~512M each, but my Pitcairn 1G are underused by an order of magniture...

The total hashrate of the rig, counting the 2G that mine faster and the 1G that mine slower, is 10% lower that claymore 11.3, but with 0 bad share, where Claymore, as I said previously, gives ~10% bad shares on old GPU when configured at max, or is stable at 10% lower hashrate.

0.29c online

Use of the GPU version frontend, but GPU disabled
Support of Masari fork --variation 11
Support of Haven fork --variation 12
Stellite defaults to XTL variation
ETN defaults to v7
other coins that have forked will default to their respective fork
Coin list in alphabetical order
new coin : BXB
new coin : WOWnero
-b parameter ignored for compatibility with AwsomeMiner
ditto with parameter --api-remote

[4ward]

--variation 12
returns bad shares with multi_hash higher than 2
while --variation 5 works fine
tested on Core i5 7600k (6MB cache)

[Iamtutut]

It means Haven has not forked yet.

[JCE-Miner]

it should have forked already, and Masari should fork in a few hours.
probably my code is broken in multihash, i've so many assemblies now that it's almost impossible to test all in decent time.

i'll release an emergency fix asap, thanks for the report !

[gvb]

thanks for the work on this.

I just notice one problem on the machine I'm currently using.

It only uses 3 of the 4 cores.

Adding the -t 4 right behind the --auto parameter in the start batch file seems to solve the 'problem'.

CPU is an Intel i5-4590 3.3GHz

[raske]

It means Haven has not forked yet.

of course it is forked, the new algo is CN Haven, asic&NH proof

[JCE-Miner]

In most case such an error comes from a fork I enabled in advance, like XTL but here that's a real bug, Heavy algorithm is used in place of Haven when multihash is used. gracefully multihash is rarely used with Heavy/Haven. a hotfix will come soon.

[Larvitar]

so the no-cache is bad on the A10. It's very good on modern Intel, decent on Ryzen and bad on Core2. And so, bad on A10 too.
So keep the config you had found before, with two normal threads.
I won't have time to fix the A10 autoconfig in next version, will be for 0.30

edit: i'm polishing the GPU version.
I see very strange results on my Pitcairn rig (five different Pitcairns : 7870 and 7850, 1G and 2G...)

On my 2G, when I use all the memory available, I can reach 526h/s on my 7870 2G, and 495 on CN-v7 on my 7850 2G, which is far above what Claymore 9.7 gave on CN-Classic. It's even the best score ever I reached, period. Maybe i didn't configure SRB well, but so far i didn't reach such an hashrate.

But on my little 7850 1G, that's terrible, i peak at 343 where Claymore gave a whooping 450 !
It looks like my code is faster when using the double-memory, but a lot slower otherwise. And double-memory is barely possible on 1G cards. It works on Bonaire 1G because it has a little GPU that can be saturated with two threads using ~512M each, but my Pitcairn 1G are underused by an order of magniture...

The total hashrate of the rig, counting the 2G that mine faster and the 1G that mine slower, is 10% lower that claymore 11.3, but with 0 bad share, where Claymore, as I said previously, gives ~10% bad shares on old GPU when configured at max, or is stable at 10% lower hashrate.

0.29c online

Use of the GPU version frontend, but GPU disabled
Support of Masari fork --variation 11
Support of Haven fork --variation 12
Stellite defaults to XTL variation
ETN defaults to v7
other coins that have forked will default to their respective fork
Coin list in alphabetical order
new coin : BXB
new coin : WOWnero
-b parameter ignored for compatibility with AwsomeMiner
ditto with parameter --api-remote

I have to news, a good and a bad one:
The good: JCE is lauching normally by Awesome as CPUMiner-Opt.

The bad: API isn't working with Awesome:

18/06/2018 17:26:35.007 [022] [E]Failed to process API request (time: 1001 ms): summary
Nenhuma conexão pôde ser feita porque a máquina de destino as recusou ativamente 127.0.0.1:4034
18/06/2018 17:26:41.239 [023] [E]Failed to process API request (time: 1001 ms): summary
Nenhuma conexão pôde ser feita porque a máquina de destino as recusou ativamente 127.0.0.1:4034
18/06/2018 17:26:47.440 [027] [E]Failed to process API request (time: 1000 ms): summary
Nenhuma conexão pôde ser feita porque a máquina de destino as recusou ativamente 127.0.0.1:4034
18/06/2018 17:26:53.655 [041] [E]Failed to process API request (time: 1000 ms): summary
Nenhuma conexão pôde ser feita porque a máquina de destino as recusou ativamente 127.0.0.1:4034

[JCE-Miner]

sounds logical, jce api is either its own, or xmr-stak compatible, but not cpumineropt-compatible.
Maybe ask Awsomeminer to support jce, it should be simple since it has command line similar to cpuminerot (and to claymore) and api just like stak.

i got bad report about 0.29c algo selection, haven is broken, bixbite seems too, need to hot release a fix.

[gvb]

JCE,

is it possible to add a max hashspeed storing for the current session and display this in the 'r' report and the light blue/cyan line that reports hash speeds every now and then?

it's easier to know the max speed this way.