Author

Topic: [XMR] JCE Miner Cryptonight/forks, now with GPU! - page 114. (Read 90841 times)

member
Activity: 350
Merit: 22
Thanks ! Cheesy
The best is GNU Assembly, it's high level assembly and very flexible, you can let compiler choose the registers, or not, reorder or not... excellent assembly.

Ok for mining rentals, I add it too. I'll make a 0.22 with just new coins and pools, no perf change, no dualshare yet.
jr. member
Activity: 196
Merit: 1
Thanks man. I wasn't kidding about the speed, it's really all written in assembly, instruction after instruction. For every Cryptonight variation, for 32 and 64 bits, with or without AES, with or withour dual-mining... Dozens of assembly codes, hence the large binary size.

The hardest to beat was the Original Cryptonight + AES : it's so simple, a big loop that does AES and 64-bits multiply, that there were no room for improvement. I beat a compiler by a small 1% Sad The new, more complex variations let JCE have a real gain over C-based miners.

I'm making 0.18 right now, with support of TRTL version Cryptolight-v7. If you need support of more coins and/or pools, just ask Wink

I know i'm a new member here, but no danger, the miner is not a virus not a trojan, and all hashrate values are real.

Thanks for your sharing mate! Honestly speaking i LOVE assembly language too! Was quite good at it while most people in my class are having trouble with it.

Yes very low level but fastest as no junk code Smiley

Good job ya. Currently using XMRIG (From XMR-STAK)

Will try using yours
newbie
Activity: 35
Merit: 0
Account of what "pool" ?
I already support a few of them like Nicehash or miningpoolhub.

I don't ask account, just the "pool" or "hashrate Marketplace" you want to use.
Pool - miningrigrentals.com
Login type: name.number
member
Activity: 350
Merit: 22
Account of what "pool" ?
I already support a few of them like Nicehash or miningpoolhub.

I don't ask account, just the "pool" or "hashrate Marketplace" you want to use.
newbie
Activity: 35
Merit: 0
Does not work if the Wallet parameter is the account login.
"Wallet does not match any of the supported currencies."
member
Activity: 350
Merit: 22
Ryzen did not notice the increase.
1600 @ 3850 XMR Stack = 600+ h / s.

Redid test on a fresh newly installed Win10 pro, all Large Pages enabled, Ryzen 1600@stock, mining Monero (v7), best config (8 threads):
Code:
xmr-stak 2.4.2 e10e8e67
HASHRATE REPORT - CPU
| ID |    10s |    60s |    15m | ID |    10s |    60s |    15m |
|  0 |   55.8 |   (na) |   (na) |  1 |   67.0 |   (na) |   (na) |
|  2 |   67.0 |   (na) |   (na) |  3 |   56.0 |   (na) |   (na) |
|  4 |   56.0 |   (na) |   (na) |  5 |   66.8 |   (na) |   (na) |
|  6 |   66.9 |   (na) |   (na) |  7 |   56.0 |   (na) |   (na) |
Totals (CPU):   491.5    0.0    0.0 H/s
-----------------------------------------------------------------
Totals (ALL):    491.5    0.0    0.0 H/s
Code:
JCE 0.21 x64
10:18:32 | Hashrate Thread 0: 59.16 h/s
10:18:32 | Hashrate Thread 1: 59.69 h/s
10:18:32 | Hashrate Thread 2: 66.99 h/s
10:18:32 | Hashrate Thread 3: 66.99 h/s
10:18:32 | Hashrate Thread 4: 58.91 h/s
10:18:32 | Hashrate Thread 5: 58.91 h/s
10:18:32 | Hashrate Thread 6: 66.67 h/s
10:18:32 | Hashrate Thread 7: 66.69 h/s
10:18:32 | Total: 503.99 h/s

JCE is almost 3% faster on Ryzen. Got same results against XMrig
member
Activity: 350
Merit: 22
https://ark.intel.com/products/83428/Intel-Pentium-Processor-G3460-3M-Cache-3_50-GHz

Easy, just add -t 2 to the command line.
I'll improve the autoconfig to better handle the 3M cache cpu, for them 2 threads is always better.
Parameter --low for low intensity
full member
Activity: 1120
Merit: 131
Mining with a Pentium G3460 since around 12hs.
One core / thread is used, it let the miner in auto config.

Results:
- pool is Stellite miner rocks
- coin in Stellite (XTL)
- average hashing speed from pool is 30,88H/s
- power draw is 11/12w at the wall.

How could I make the miner use both cores ? With low intensity please.
member
Activity: 350
Merit: 22
hello.. Im trying to mine other Cryptonight coins... Crep coin, BCN coin, Plura coin
But wallet does not match any supported currencies /?
I note, those coins will be added in next version. I want to check coin for diagnostic and auto fork detection. Otherwise 99% of defect report are because of a typo in wallet or bad fork selected.
newbie
Activity: 7
Merit: 0
hello.. Im trying to mine other Cryptonight coins... Crep coin, BCN coin, Plura coin
But wallet does not match any supported currencies /?
member
Activity: 350
Merit: 22


Quote
02:10:32 | Pool pool.minexmr.com:4444
02:10:32 | Currency Monero (XMR)
02:10:32 | Current pool Difficulty 20000
02:10:32 | Valid Shares found 56
02:10:32 | Total hashes 1120000
02:10:32 | Miner uptime 2:10:50
02:10:32 | All time average 142.49 H/s

Ok, that will be the net (fees deduced) effective hashrate, the same as reported by pool, including the good/bad luck factor. Will be equal to physical hashrate on the long term, mathematically after an infinite time Wink Good idea, will be in next version Smiley
member
Activity: 564
Merit: 19
@JCE-Miner

It will be nice to see a total hashes / uptime when I press R key.

So far so good, a nice addition to my GPU rigs. Thanks
Is it a request? Jce already gives such info when pressing R.
Btw, thanks Smiley

Yea, I want something like this please:

Quote
02:10:32 | Pool pool.minexmr.com:4444
02:10:32 | Currency Monero (XMR)
02:10:32 | Current pool Difficulty 20000
02:10:32 | Valid Shares found 56
02:10:32 | Total hashes 1120000
02:10:32 | Miner uptime 2:10:50
02:10:32 | All time average 142.49 H/s
member
Activity: 350
Merit: 22
@JCE-Miner

It will be nice to see a total hashes / uptime when I press R key.

So far so good, a nice addition to my GPU rigs. Thanks
Is it a request? Jce already gives such info when pressing R.
Btw, thanks Smiley

Edit : I should detect L4 (not tested real life, my best cpu is Ryzen with L3) but yeah the perf gap seems to come from that dualshare mode. I'm currently writing the assembly, starting with AES 64 mode. Not sure I'll do a 32 bits code. Not enough registers Sad
member
Activity: 350
Merit: 22
Jce now has a limit of 64 threads, from version 0.21
I have no dual cpu machine with so many cores so I couldn't test real life, but the 32 limit is gone.
-t won't let you spawn more threads than cpu detected, on purpose. To go above, use config file.

A copy of the first lines logged, where it detects your cpu, could help.

I've had bad reports about perfs on i7 and other big Intel cpu. I don't own one so I cannot assembly optimize for now. I'm surprised of such a difference, but yeah it may be the lack of dualshare mode.

On ryzen I'm better but the difference is low, about 1% on CN, on v7 I'm about 2.5% faster. I reach 502 on ryzen 1600 while xmrig and stak max at 489
newbie
Activity: 79
Merit: 0
@Mind163
your Broadwell CPU not only has L3 cache, but even L4 cache  Shocked Thats why XMR-Stak switched to double hash mode (there called "low_power_mode"). This feature JCE is missing so far, but the developer is just working on, if you have read the thread Smiley
I apologize. My native language is russian. It's hard for me to read the whole topic. So I decided to unsubscribe and send a configuration to help the author.
newbie
Activity: 1
Merit: 0
Hi!

I have a 4 x e7-4870 (4 x 10 core + hyperthreading, 30mb cache per cpu) machine. It won't let me set more than 32 threads with the -t command. I've had the machine up to 1400h/s on xmrig, but I just can't seem to get past 830 with this. Please help Smiley

Straight from download, I get ~300h/s

If I set 32 threads, I get about 820h/s.
jr. member
Activity: 37
Merit: 5
@Mind163
your Broadwell CPU not only has L3 cache, but even L4 cache  Shocked Thats why XMR-Stak switched to double hash mode (there called "low_power_mode"). This feature JCE is missing so far, but the developer is just working on, if you have read the thread Smiley
newbie
Activity: 43
Merit: 0
Sure no problem, 10 min run BRB.

EDIT:

PhenomII X6 Dualmine 128H/s

newbie
Activity: 79
Merit: 0
Processor i7 5775c Broadwell on XRM STACK gives ~ 380 h / s. On your miner is ~ 260 h / s Cryptonight V7. Here is the configuration:

[
    { "low_power_mode" : true, "no_prefetch" : true, "affine_to_cpu" : false },
    { "low_power_mode" : true, "no_prefetch" : true, "affine_to_cpu" : false },
    { "low_power_mode" : true, "no_prefetch" : true, "affine_to_cpu" : false },
    { "low_power_mode" : true, "no_prefetch" : true, "affine_to_cpu" : false },
    { "low_power_mode" : true, "no_prefetch" : true, "affine_to_cpu" : false },
    { "low_power_mode" : true, "no_prefetch" : true, "affine_to_cpu" : false },
    { "low_power_mode" : true, "no_prefetch" : true, "affine_to_cpu" : false },
    { "low_power_mode" : true, "no_prefetch" : true, "affine_to_cpu" : false },
    

],


Also your miner, in contrast, for example XMRIG is tied to coins from the list. It is not comfortable.
On the XMRIG, the result is about 370 h / s.
Remember, these processors generally have an L3 cache.
Ryzen did not notice the increase.
1600 @ 3850 XMR Stack = 600+ h / s.
jr. member
Activity: 37
Merit: 5
Still PhenomII X6 have better performance with 6T than 5T+helper, 153 vs 141.
Could you please try the following config for your X6; just for interest Smiley

"cpu_threads_conf" :
[
     { "cpu_architecture" : "thuban", "affine_to_cpu" : 0, "use_cache" : true },
     { "cpu_architecture" : "thuban", "affine_to_cpu" : 1, "use_cache" : true },
     { "cpu_architecture" : "thuban", "affine_to_cpu" : 2, "use_cache" : true },
     { "cpu_architecture" : "thuban", "affine_to_cpu" : 3, "use_cache" : true, "dual_mine_with": 2 },
     { "cpu_architecture" : "thuban", "affine_to_cpu" : 4, "use_cache" : true },
     { "cpu_architecture" : "thuban", "affine_to_cpu" : 5, "use_cache" : true, "dual_mine_with": 4 },
]
THEORETICALLY, this must be the fastest config for Phenom II X6

Quote
What clock is your X4 running?
Phenom II X4 905e runs at 2500 MHz.
Jump to: