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

[pornello]

hi again, alloy is on cryptonight algo right? same as bytecoin for example, its not heavy right? while mining turtle i get around 5% improvement with jce when compared to xmrig but now as i started mining alloy my hashrate is 50% less than bytecoin on xmrig, i get best results on 4 threads. my hardware is intel-I7 64bit aes supported L2,L3 is 2 and 6 mb, can someone help me to choose manual config to see if ill get better results maybe? also i need example how to allocate manual config folder into start.bat, i dont get it what i type after -c? i typed -c C:\users\Neno\Desktop\cryptos\jce miner\config> but it didnt work, obviously i made mistake, sorry for being dumb but im kinda new into mining and completely new user of jce miner

also i need advice whether what config should i use for example how many threads, should i use cache or not and should i use multihash on some threads?

[JCE-Miner]

hi again, alloy is on cryptonight algo right?

On a fork: Cryptonight-Alloy. That's a variation that's a lot slower than normal cryptonight (the same as Electroneum or Bytecoin for example)
It's normal to get lower perf than with Bytecoin, so

as i started mining alloy my hashrate is 50% less than bytecoin on xmrig

on a AES CPU, Alloy is twice slower, so yes that's normal. On non-AES, the loss is lower (about -40%)

Now let's go for manual config.

1. In the .bat, replace --auto by -c config.txt
2. Create a file config.txt in the same folder as the .bat
3. In that new file, paste

Code:

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

4. Some tweak: since you didn't give your exact CPU model, i don't know how many CPU cores you have, so i wrote 0,1,2,3
If you have more that 4 logical cores, you may better set 0,2,4,6

You can also play with multi_hash, but with 6M of L3 cache, i guess it will be bad.
Please give the exact model of your CPU so i can tune it as optimal

[pornello]

hi again, alloy is on cryptonight algo right?

On a fork: Cryptonight-Alloy. That's a variation that's a lot slower than normal cryptonight (the same as Electroneum or Bytecoin for example)
It's normal to get lower perf than with Bytecoin, so

as i started mining alloy my hashrate is 50% less than bytecoin on xmrig

on a AES CPU, Alloy is twice slower, so yes that's normal. On non-AES, the loss is lower (about -40%)

Now let's go for manual config.

1. In the .bat, replace --auto by -c config.txt
2. Create a file config.txt in the same folder as the .bat
3. In that new file, paste

Code:

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

4. Some tweak: since you didn't give your exact CPU model, i don't know how many CPU cores you have, so i wrote 0,1,2,3
If you have more that 4 logical cores, you may better set 0,2,4,6

You can also play with multi_hash, but with 6M of L3 cache, i guess it will be bad.
Please give the exact model of your CPU so i can tune it as optimal

ahh now i understand, so it doesnt matter if the algo is same, forks can make same algorytm coins give different hashrates, didnt figured that out, im still a noob miner . yes i created config folder but i understood wrong how to put it into bat folder, i tried config but i didnt add .txt (what a dumbass) so of course bat couldnt find matching file it fully makes sense. processor is 4 cores, very old i7.
https://ark.intel.com/products/52219/Intel-Core-i7-2630QM-Processor-6M-Cache-up-to-2_90-GHz
i will try recommended config right now, on a link above you have all the necessary info so that you can play with tuning, thank you for help, you responded very fast, thumbs up for you

[pornello]

i get error "invalid config file syntax in config.txt

[s0ftcorn]

i get error "invalid config file syntax in config.txt

post your config


also:

i get extremely low hashrate for cryptonight-heavy on my Ryzen 2700X 150H/s to 200H/s at best. is this normal? since cryptonight heavy uses 4M blocks i should get better results with multi hash, right?

[jibskg]

hello
not able to mine solace coin with jce miner crashing application in the beginning

[Buyingallcoinsz]

That's a Piledriver, i haven't one but have an Excavator, i could use it as a test base for that archi. Current optimized assembly is for Ryzen only. And i never tested Intel Aes yet, but got tester reports of mitigate improvement, very close to xmrig.


Piledriver

[pornello]

i get error "invalid config file syntax in config.txt

post your config


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

i used config that admin suggested, pasted this into config.txt file

[JCE-Miner]

i get error "invalid config file syntax in config.txt

post your config


also:

i get extremely low hashrate for cryptonight-heavy on my Ryzen 2700X 150H/s to 200H/s at best. is this normal? since cryptonight heavy uses 4M blocks i should get better results with multi hash, right?

@pornello: did you copy that exact text, nothing more, nothing less, into config.txt ? And config.txt put in same folder as start.bat ?
It looks valid JSON...

cryptonight-heavy : nope, multi-hash add still more cache consumption, leading to worse perf. That's rather for cryptolight.
for your 2700X here's the best config:

Code:

"cpu_threads_conf" :  
[  
     { "cpu_architecture" : "ryzen", "affine_to_cpu" : 1, "use_cache" : true },
 
     { "cpu_architecture" : "ryzen", "affine_to_cpu" : 2, "use_cache" : false },
     { "cpu_architecture" : "ryzen", "affine_to_cpu" : 3, "use_cache" : false },
 
     { "cpu_architecture" : "ryzen", "affine_to_cpu" : 5, "use_cache" : true },
 
     { "cpu_architecture" : "ryzen", "affine_to_cpu" : 6, "use_cache" : false },
     { "cpu_architecture" : "ryzen", "affine_to_cpu" : 7, "use_cache" : false },
 
     { "cpu_architecture" : "ryzen", "affine_to_cpu" : 9, "use_cache" : true },
 
     { "cpu_architecture" : "ryzen", "affine_to_cpu" : 10, "use_cache" : false },
     { "cpu_architecture" : "ryzen", "affine_to_cpu" : 11, "use_cache" : false },
 
     { "cpu_architecture" : "ryzen", "affine_to_cpu" : 13, "use_cache" : true },
 
     { "cpu_architecture" : "ryzen", "affine_to_cpu" : 14, "use_cache" : false },    
     { "cpu_architecture" : "ryzen", "affine_to_cpu" : 15, "use_cache" : false },    
]

Solace crash : i admit i didn't test Solace explicitely, dozen of coins times hundreds of assembly are not fully testable. I'll do emergency test asap.
Other problem may be use of an unsupported CPU instruction, maybe try with --auto --archi generic, you'll get bad perf, but at least you can check if it runs.
Please copy the first lines of the log (where it detects CPU and build the mining threads) so i can know what assembly causes trouble.

[pornello]

@pornello: did you copy that exact text, nothing more, nothing less, into config.txt ? And config.txt put in same folder as start.bat ?
It looks valid JSON...

yup i copied exactly what you posted, put into empty config.txt which i already put in same folder as start.bat

[JCE-Miner]

So you did things right, but i tested exact same text, and works.
I bet you missed a character when copy-pasted.
If i remove last character ] i get the same error message than you.
All characters count!

re-testing CN-Heavy mining right now...

[pornello]

@pornello: did you copy that exact text, nothing more, nothing less, into config.txt ? And config.txt put in same folder as start.bat ?
It looks valid JSON...

yup i copied exactly what you posted, put into empty config.txt which i already put in same folder as start.bat

i dunno how but it started working after 5 attempts

[JCE-Miner]

No problem bro! Happy to read this.

[pornello]

No problem bro! Happy to read this.

i find best setting to use auto config and set t-8 which is max for me i think, it gives me 10-12% more hashrate than any other configuration

[JCE-Miner]

you're exploring news seas, the config rules of Cryptonight may not apply to Alloy, it's a very new algo. weak cpus like your Mobile i7 often have best perfs when mining with all cores.

you may also try with some multi_hash >1 in the config file, with either 4 or 8 threads, as an experiment

[sergneo]

JCE-Miner
I don't understand how you could get 116 h/s on a 2.66 GHz Core2Quad. I have a Core2Quad Q9400 2.66 GHz, on your miner I get about 100H / s. On what processor model did you test, probably Q9450?
(Win7 x64, MoneroV7)

miner defines architecture as Kentsfield, Core2Quad Q9400 (Yorkfield-6M )

[JCE-Miner]

that's on Cryptonight, on V7 it drops to 113
are you sure you have the Large Pages enabled ? 100 is the score i get with Large pages missing.
and that's on the 64 bits version, 32 bits gives me 92

but yes, maybe force config with -t 4 --archi core2
my score is real, and core2 is the most optimized on my assemblies

to be precise, my chipset is G31 with ddr2 800, if you run your quad on a i965 it may explain the bad perf.

edit: hooooo i get it
i've a Q9450 with 12M cache, and yours is a Q9400 with 6M cache. that's normal. i'll update the bench to be more precise, i didn't notice there were Quad with only 6M

[sergneo]

Yes, 8Mb cache Kentsfield (Q6xxx), Yorkfield (Q9x50) 12Mb , Yorkfield-6M (Q8x00, Q9x00) 6Mb.
the auto config on this processor does not work correctly.

Code:

For Windows 64-bits
Analyzing Processors topology...
Intel(R) Core(TM)2 Quad CPU Q9400 @ 2.66GHz
Architecture codename: Kentsfield
  SSE2          : Yes
  SSE3          : Yes
  SSE4          : Yes
  AES           : No
  AVX           : No
  AVX2          : No

Auto-configuration, selected CPUs will be highlighted...
Found CPU 0, with:
  L1 Cache:    32 KB
  L2 Cache:  3072 KB, shared with CPU 1
Found CPU 1, with:
  L1 Cache:    32 KB
  L2 Cache:  3072 KB, shared with CPU 0
Found CPU 2, with:
  L1 Cache:    32 KB
  L2 Cache:  3072 KB, shared with CPU 3
Found CPU 3, with:
  L1 Cache:    32 KB
  L2 Cache:  3072 KB, shared with CPU 2

Preparing 4 Mining Threads...

+-- Thread 0 config -----------------------------+
| Run on CPU:             0                      |
| Use cache:              yes                    |
| Multi-hash:             no                     |
| Assembly module:        generic_sse4           |
+------------------------------------------------+

+-- Thread 1 config -----------------------------+
| Run on CPU:             1                      |
| Use cache:              yes                    |
| Multi-hash:             no                     |
| Assembly module:        generic_sse4           |
+------------------------------------------------+

+-- Thread 2 config -----------------------------+
| Run on CPU:             2                      |
| Use cache:              yes                    |
| Multi-hash:             no                     |
| Assembly module:        generic_sse4           |
+------------------------------------------------+

+-- Thread 3 config -----------------------------+
| Run on CPU:             3                      |
| Use cache:              yes                    |
| Multi-hash:             no                     |
| Assembly module:        generic_sse4           |
+------------------------------------------------+

used by the kernel generic_sse4 , with 6M L2 cache uses all 4 cores Use cache: yes . as much as 8Mb of 6 ? Is that supposed to be?
what is the difference generic_sse4 and core2_sse4 ?
large pages of course included

MB Gigabyte GA-EP45-UD3LR 4Gb DDR3-1333

update:
if I turn on one core cache, then the rate drops to 81 H/s

Code:

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

[sergneo]

I also have a lot of Xeon(R) CPU E5440 @ 2.83GHz (core Harpertown)  overclocked bus 370Mhz = 3145MHz, auto config gives me 135H/s.
as a result, it is used 2+2+2+2 =8MB cache

Code:

For Windows 64-bits
Analyzing Processors topology...
Intel(R) Xeon(R) CPU E5440 @ 2.83GHz
Architecture codename: Harpertown
  SSE2          : Yes
  SSE3          : Yes
  SSE4          : Yes
  AES           : No
  AVX           : No
  AVX2          : No

Auto-configuration, selected CPUs will be highlighted...
Found CPU 0, with:
  L1 Cache:    32 KB
  L2 Cache:  6144 KB, shared with CPU 1
Found CPU 1, with:
  L1 Cache:    32 KB
  L2 Cache:  6144 KB, shared with CPU 0
Found CPU 2, with:
  L1 Cache:    32 KB
  L2 Cache:  6144 KB, shared with CPU 3
Found CPU 3, with:
  L1 Cache:    32 KB
  L2 Cache:  6144 KB, shared with CPU 2

Preparing 4 Mining Threads...

+-- Thread 0 config -----------------------------+
| Run on CPU:             0                      |
| Use cache:              yes                    |
| Multi-hash:             no                     |
| Assembly module:        core2_sse4             |
+------------------------------------------------+

+-- Thread 1 config -----------------------------+
| Run on CPU:             1                      |
| Use cache:              yes                    |
| Multi-hash:             no                     |
| Assembly module:        core2_sse4             |
+------------------------------------------------+

+-- Thread 2 config -----------------------------+
| Run on CPU:             2                      |
| Use cache:              yes                    |
| Multi-hash:             no                     |
| Assembly module:        core2_sse4             |
+------------------------------------------------+

+-- Thread 3 config -----------------------------+
| Run on CPU:             3                      |
| Use cache:              yes                    |
| Multi-hash:             no                     |
| Assembly module:        core2_sse4             |
+------------------------------------------------+

manual adjustment, gives +2 H/s

Code:

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

2+4+2+4 =12Mb -> Getting 137H/s

[JCE-Miner]

hi all,

i see one bug : the bad core2 detect.
autoconfig on non-aes core2 uses all cores, even with low cache, because it's faster. at least that's what i observed on mine. on a 6M naive config would be 3 threads, but i observed 4 was faster.

also sse4 instructions that later core2 have is well detected and used.

about the Architecture codename, it's misleading, i see too many people with bad info, or thinking it can be overriden by --archi. since it's cosmetical, i'll replace it with the Assembly codename, which is more useful (and this is what's overridable)

a better param would have been --assembly instead of --archi, my bad, but that's too late now.

linux version on the way.

edit: i found the core2 detect bug, next version will display assembly Codename instead of Architecture, and for all Core2/Core2Xeon, it will be core2 or core2_sse4

The difference between core2_sse4 and generic_sse4 is a few core2 specific ASM optims, mostly focused on filling well the pipeline. The perf difference is usually ~2 h/s but 2 is better than Nothing