Closed tevador closed 5 years ago
ARM Cortex-A53, 1.2 GHz, 1 GB LPDDR2, Ubuntu 16.04 armv7l (32-bit), compiled with GCC 8.1.0
> bin/randomx --verify --softAes
Using software AES.
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 8.55579 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Performance: 578.734 ms per hash
Clocked at 3.35 GHz (Turbo disabled), 16 GB of DDR4-2933 (dual channel), Ubuntu 16.04 x86_64, compiled with GCC 8.1.0
> bin/randomx --verify --largePages
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 0.6282 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Performance: 55.5542 ms per hash
> bin/randomx --mine --largePages --init 16 --threads 8 --nonces 1000000
RandomX - mining mode
Initializing (16 threads) ...
Dataset (4 GiB) initialized in 9.40938 s
Initializing 8 virtual machine(s) ...
Running benchmark (1000000 nonces) ...
Calculated result: 9979734c3f8a2f665810851050aaf8a63d6bc9810cfecee9d55a3706a81266fd
Performance: 4124.02 hashes per second
Boosts to ~3.5 GHz at the beginning, then throttles down to ~2.4 GHz (mining mode). 16 GB of DDR4-2400 (dual channel), Windows 10 64-bit, compiled with Visual Studio 2017 15.8.3
>randomx.exe --verify --largePages
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 0.576263 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Performance: 28.8373 ms per hash
>randomx.exe --mine --largePages --init 8 --threads 4 --nonces 100000
RandomX - mining mode
Initializing (8 threads) ...
Dataset (4 GiB) initialized in 14.2439 s
Initializing 4 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: abe4b79b0439e54e2896e0134cc89c8bb1b7a378192095efc690f8f85699a225
Performance: 1678.74 hashes per second
16GB of DDR4. Base speed 2.8GHz. Boosts to ~3.5GHz in testing. L3 cache 6MB. Windows 10 64-bit. Used precompiled binaries.
./randomx.exe --verify RandomX - verification mode Initializing ... Cache (256 MiB) initialized in 0.906537 s Initializing 1 virtual machine(s) ... Running benchmark (1000 nonces) ... Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851 Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851 Performance: 49.6185 ms per hash
./randomx.exe --mine --init 8 --threads 3 --nonces 100000 RandomX - mining mode Initializing (8 threads) ... Dataset (4 GiB) initialized in 20.02 s Initializing 3 virtual machine(s) ... Running benchmark (100000 nonces) ... Calculated result: abe4b79b0439e54e2896e0134cc89c8bb1b7a378192095efc690f8f85699a225 Performance: 753.452 hashes per second
./randomx.exe --mine --init 4 --threads 3 --nonces 100000 RandomX - mining mode Initializing (4 threads) ... Dataset (4 GiB) initialized in 30.4974 s Initializing 3 virtual machine(s) ... Running benchmark (100000 nonces) ... Calculated result: abe4b79b0439e54e2896e0134cc89c8bb1b7a378192095efc690f8f85699a225 Performance: 782.408 hashes per second
@SamsungGalaxyPlayer Can you try --largePages
? With 16 GB, you should have no issues with contiguous memory.
Intel Core i5-8250U CPU @ 1.60GHz 8GB DDR4, L3 Cache 6 MB
$ ./randomx --verify
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 0.769388 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Performance: 42.6374 ms per hash
./randomx --mine --init 8 --threads 6 --nonces 100000
RandomX - mining mode
Initializing (8 threads) ...
Dataset (4 GiB) initialized in 20.5339 s
Initializing 6 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: abe4b79b0439e54e2896e0134cc89c8bb1b7a378192095efc690f8f85699a225
Performance: 1021.56 hashes per second
Compiled latest commit 18ca8b5
, Arch Linux, gcc 8.2.1 20181127
Intel(R) Xeon(R) CPU E5506 @ 2.13GHz
./randomx --verify --softAes Using software AES. RandomX - verification mode Initializing ... Cache (256 MiB) initialized in 1.7613 s Initializing 1 virtual machine(s) ... Running benchmark (1000 nonces) ... Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851 Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851 Performance: 132.288 ms per hash
AMD Ryzen Mobile Pro 2700U @ 2.20GHz, 32GB DDR4-2400 Ubuntu 18.10, gcc (Ubuntu 8.2.0-7ubuntu1) 8.2.0
vielle:/home/software/RandomX> bin/randomx --verify --largePages
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 0.616493 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Performance: 60.0637 ms per hash
vielle:/home/software/RandomX> bin/randomx --mine --init 8 --threads 4 --nonces 100000 --largePages
RandomX - mining mode
Initializing (8 threads) ...
Dataset (4 GiB) initialized in 14.8269 s
Initializing 4 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: abe4b79b0439e54e2896e0134cc89c8bb1b7a378192095efc690f8f85699a225
Performance: 1372.41 hashes per second
Geekbox (Rockchip RK3368, 8xARM Cortex-A53 @ 1.5GHz, 2GB RAM) Debian Jessie, gcc (Debian/Linaro 4.9.2-10) 4.9.2
geekbox@geekbox:/home/software/RandomX$ bin/randomx --verify --softAes
Using software AES.
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 4.05107 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Performance: 447.505 ms per hash
I'm a bit overwhelmed with work at the moment, but I can get ARM AES working after my next conference is over in a couple weeks. Or someone else can beat me to it if they like.
Xeon E5-2670v2 @ 2.50Ghz (2 processors), 128GB ECC DD3 Windows Server 2012 R2
C:\Users\Administrator\Desktop\randomx>randomx --verify
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 2.08985 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Performance: 88.0998 ms per hash
C:\Users\Administrator\Desktop\randomx>randomx --verify --largePages
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 1.50766 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Performance: 81.2674 ms per hash
C:\Users\Administrator\Desktop\randomx>randomx --mine --largePages --init 20 --threads 12 --nonces 100000
RandomX - mining mode
Initializing (20 threads) ...
Dataset (4 GiB) initialized in 11.2982 s
Initializing 12 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: abe4b79b0439e54e2896e0134cc89c8bb1b7a378192095efc690f8f85699a225
Performance: 3556.08 hashes per second
C:\Users\Administrator\Desktop\randomx>randomx --mine --largePages --init 40 --threads 25 --nonces 100000
RandomX - mining mode
Initializing (40 threads) ...
Dataset (4 GiB) initialized in 7.52452 s
Initializing 25 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: abe4b79b0439e54e2896e0134cc89c8bb1b7a378192095efc690f8f85699a225
Performance: 4986.93 hashes per second
Try mining mode with only 10 threads. It might run faster than 12 threads because of L2 cache size.
You mean L3? It is 25Mb per CPU (plus 2.5Mb L2). So 12 threads should be ok.
C:\Users\Administrator\Desktop\randomx>randomx --mine --largePages --init 20 --threads 10 --nonces 100000
RandomX - mining mode
Initializing (20 threads) ...
Dataset (4 GiB) initialized in 11.0515 s
Initializing 10 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: abe4b79b0439e54e2896e0134cc89c8bb1b7a378192095efc690f8f85699a225
Performance: 3164.38 hashes per second
OK. Someone tested an 8 core Xeon with 20 MB of cache in https://github.com/tevador/RandomX/issues/22 and got better performance with 8 threads than 10. So your case is different I guess. In any case, since your system has NUMA, you should run 2 instances of RandomX (one for each CPU) for a total of ~7000 H/s.
Note that the only system where the binary worked was an ubuntu 18 box. Everything else i had to compile
gcc version 5.4.0 20160609 (Ubuntu 5.4.0-6ubuntu1~16.04.11)
model name : Intel(R) Core(TM) i7-2700K CPU @ 3.50GHz
stepping : 7
microcode : 0x28
cpu MHz : 1651.425
cache size : 8192 KB
physical id : 0
siblings : 8
core id : 0
cpu cores : 4
mondaemon@jdc-user:~/randomx_test_20190220/RandomX/bin$ ./randomx --verify
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 0.896662 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Performance: 52.4011 ms per hash
mondaemon@jdc-user:~/randomx_test_20190220/RandomX/bin$ time ./randomx --mine --init 4 --threads 4 --nonces 40000
RandomX - mining mode
Initializing (4 threads) ...
Dataset (4 GiB) initialized in 25.8137 s
Initializing 4 virtual machine(s) ...
Running benchmark (40000 nonces) ...
Calculated result: c7aa337763ce1e317fa85f76bc48fd698fbbc90c3811625d5d98080251e148e6
Performance: 1221.14 hashes per second
real 0m58.683s
user 3m49.432s
sys 0m1.640s
gcc version 5.4.0 20160609 (Ubuntu 5.4.0-6ubuntu1~16.04.11)
model name : Intel(R) Core(TM) i7-7700K CPU @ 4.20GHz
stepping : 9
microcode : 0x8e
cpu MHz : 800.205
cache size : 8192 KB
physical id : 0
siblings : 8
core id : 0
cpu cores : 4
fastpc@fastpc-desktop:~/randomx_test_20190220/RandomX/bin$ ./randomx --verify
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 0.539419 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Performance: 32.1323 ms per hash
fastpc@fastpc-desktop:~/randomx_test_20190220/RandomX/bin$ time ./randomx --mine --init 4 --threads 7 --nonces 50000
RandomX - mining mode
Initializing (4 threads) ...
Dataset (4 GiB) initialized in 25.4099 s
Initializing 7 virtual machine(s) ...
Running benchmark (50000 nonces) ...
Calculated result: 2ef104fc3715a3d6a5f1c3da808061ea7d70d3b9e3b5bf671c5dea9c07acf86a
Performance: 1509.77 hashes per second
Now here's an interesting case. This is a 4 core virtual machine on top of a .... some kind of opteron. 6138 or something. nproc spits out 4, but /proc/cpuinfo shows 1.
gcc version 5.4.0 20160609 (Ubuntu 5.4.0-6ubuntu1~16.04.11)
model name : AMD Opteron 23xx (Gen 3 Class Opteron)
stepping : 1
microcode : 0x1000065
cpu MHz : 2100.040
cache size : 512 KB
physical id : 0
siblings : 1
core id : 0
cpu cores : 1
main12@main12-kvm:~/randomx_test_20190220/RandomX/bin$ ./randomx --verify --softAes
Using software AES.
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 2.9437 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: 1913e746e9cdeb21ccc2130cbe11b053a41248b6aa9b2947b00eb7512932830e
Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Performance: 183.997 ms per hash
main12@main12-kvm:~/randomx_test_20190220/RandomX/bin$ time ./randomx --mine --init 4 --threads 4 --nonces 200 --softAes
Using software AES.
RandomX - mining mode
Initializing (4 threads) ...
Dataset (4 GiB) initialized in 148.164 s
Initializing 4 virtual machine(s) ...
Running benchmark (200 nonces) ...
Calculated result: ac2827fcdd2c8f42a2808bb1bdff41d0fe17f8493c373d8de696a863b554ec1c
Performance: 63.5885 hashes per second
gcc version 5.4.0 20160609 (Ubuntu 5.4.0-6ubuntu1~16.04.10)
model name : Intel(R) Core(TM) i5-3450 CPU @ 3.10GHz
stepping : 9
microcode : 0x20
cpu MHz : 3278.613
cache size : 6144 KB
physical id : 0
siblings : 4
core id : 0
cpu cores : 4
gingerale@gingerale-Inspiron-1545:~/randomx_20190220_testing/RandomX/bin$ ./randomx --verify
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 0.83772 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Performance: 61.319 ms per hash
gingerale@gingerale-Inspiron-1545:~/randomx_20190220_testing/RandomX/bin$ time ./randomx --mine --init 4 --threads 5 --nonces 50000
RandomX - mining mode
Initializing (4 threads) ...
Dataset (4 GiB) initialized in 26.5661 s
Initializing 5 virtual machine(s) ...
Running benchmark (50000 nonces) ...
Calculated result: 2ef104fc3715a3d6a5f1c3da808061ea7d70d3b9e3b5bf671c5dea9c07acf86a
Performance: 1083.29 hashes per second
binary worked
2 gigs ram, mining didn't work
model name : AMD C-50 Processor
stepping : 0
microcode : 0x5000029
cpu MHz : 800.216
cache size : 512 KB
physical id : 0
siblings : 2
core id : 0
cpu cores : 2
user@aspire1:~/randomx_test_20190220$ ./randomx --verify --softAes
Using software AES.
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 5.75274 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Performance: 381.602 ms per hash
user@aspire1:~/randomx_test_20190220$ ./randomx --mine --init 1 --threads 1 --nonces 2 --softAes
Using software AES.
RandomX - mining mode
Initializing (1 thread) ...
Dataset (4 GiB) initialized in 531.865 s
Initializing 1 virtual machine(s) ...
Running benchmark (2 nonces) ...
Calculated result: 5cd74da718d6d36db7a5b5d139d177b3b24264dc21cf41dff892b34fa2cee332
Performance: 0.00825968 hashes per second
And here's where I grumble about this 4 GB dataset again.
ubuntu 18 so the bins work
model name : Intel(R) Celeron(R) CPU G3900 @ 2.80GHz
stepping : 3
microcode : 0xc6
cpu MHz : 800.136
cache size : 2048 KB
physical id : 0
siblings : 2
core id : 0
cpu cores : 2
monrig@monrig-MS-7977:~/randomx_test_20190220$ free -h
total used free shared buff/cache available
Mem: 3.8G 348M 3.1G 32K 351M 3.2G
Swap: 7.0G 731M 6.2G
monrig@monrig-MS-7977:~/randomx_test_20190220$ ./randomx --verify
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 0.799833 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Performance: 42.342 ms per hash
monrig@monrig-MS-7977:~/randomx_test_20190220$ ./randomx --mine --init 1 --threads 1 --nonces 1
RandomX - mining mode
Initializing (1 thread) ...
Dataset (4 GiB) initialized in 107.559 s
Initializing 1 virtual machine(s) ...
Running benchmark (1 nonces) ...
Calculated result: 007dd95406aa001eef124467fd37cea696ac84f5bc81dea789a5a22d8fb7cf76
Performance: 1.54605 hashes per second
I had to increase swap on the last one just to get it to mine. Is it going to require 8 GB of ram even if the box is just GPU mining? I mean, I'm definitely not pro GPU focused mining, but making every rig have a minimum of 8GB .... basically GPU miners try to get the things that run the GPUs as cheap as possible. Check the octominer. This thing maxes out at 8 GB ram. grumble grumble.
@Gingeropolous
model name : AMD Opteron 23xx (Gen 3 Class Opteron)
...
Calculated result: 1913e746e9cdeb21ccc2130cbe11b053a41248b6aa9b2947b00eb7512932830e
Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Can you please do:
g++ -maes -march=native -O3 -DTRACE -c src/main.cpp -o obj/main.o
make
bin/randomx --verify --softAes
and pastebin the result?
And here's where I grumble about this 4 GB dataset again.
Even if those CPUs had enough RAM, I'm estimating the following hashrates:
Perhaps the Celeron G3900 would be worth it if you could find a cheap 4 GB stick to add to the system. But all these CPUs will have a very low performance per watt. My laptop can do ~1670 H/s at 30 W, so probably 5-10x the efficiency of these machines.
Trying to mine using the swap file is a pointless effort. If you insist to mine on a low-end system, you can mine in verification mode and at least get 5-25 H/s using just 256 MB of RAM.
4x Xeon E5-4640 @ 2.5 GHz 128 GB ECC DDR3 (32 GB quad channel per node) Arch Linux, gcc 8.2.1
verification mode:
$ ./randomx --verify --largePages
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 1.17563 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Performance: 69.3351 ms per hash
mining mode:
$ seq 0 3 | xargs -P 0 -I node numactl -N node ./randomx --mine --largePages --nonces 1000000 --init 16 --threads 8
RandomX - mining mode
Initializing (16 threads) ...
RandomX - mining mode
Initializing (16 threads) ...
RandomX - mining mode
Initializing (16 threads) ...
RandomX - mining mode
Initializing (16 threads) ...
Dataset (4 GiB) initialized in 12.6244 s
Initializing 8 virtual machine(s) ...
Running benchmark (1000000 nonces) ...
Dataset (4 GiB) initialized in 13.1586 s
Initializing 8 virtual machine(s) ...
Running benchmark (1000000 nonces) ...
Dataset (4 GiB) initialized in 13.1965 s
Initializing 8 virtual machine(s) ...
Running benchmark (1000000 nonces) ...
Dataset (4 GiB) initialized in 13.3002 s
Initializing 8 virtual machine(s) ...
Running benchmark (1000000 nonces) ...
Calculated result: 9979734c3f8a2f665810851050aaf8a63d6bc9810cfecee9d55a3706a81266fd
Performance: 2554.73 hashes per second
Calculated result: 9979734c3f8a2f665810851050aaf8a63d6bc9810cfecee9d55a3706a81266fd
Performance: 2551.39 hashes per second
Calculated result: 9979734c3f8a2f665810851050aaf8a63d6bc9810cfecee9d55a3706a81266fd
Performance: 2543.19 hashes per second
Calculated result: 9979734c3f8a2f665810851050aaf8a63d6bc9810cfecee9d55a3706a81266fd
Performance: 2517 hashes per second
total hashrate: 10166.31 hashes per second
best result with a single process:
$ numactl -i all ./randomx --mine --largePages --nonces 1000000 --init 64 --threads 32
RandomX - mining mode
Initializing (64 threads) ...
Dataset (4 GiB) initialized in 9.44417 s
Initializing 32 virtual machine(s) ...
Running benchmark (1000000 nonces) ...
Calculated result: 9979734c3f8a2f665810851050aaf8a63d6bc9810cfecee9d55a3706a81266fd
Performance: 6940.36 hashes per second
2x Xeon E5-2450L @ 2 GHz 6 GB ECC DDR3 (3 GB, triple channel per node) Arch Linux, gcc 8.2.1
$ ./randomx --verify --largePages
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 1.25896 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Performance: 80.4889 ms per hash
$ numactl -i all ./randomx --mine --largePages --nonces 100000 --init 32 --threads 16
RandomX - mining mode
Initializing (32 threads) ...
Dataset (4 GiB) initialized in 9.72395 s
Initializing 16 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: abe4b79b0439e54e2896e0134cc89c8bb1b7a378192095efc690f8f85699a225
Performance: 3961.66 hashes per second
@tevador , it was just grumbling.
Here's the results... i had to use pastebin
@Gingeropolous Sorry, I missed some options on the command line.
Try again this:
g++ -std=c++11 -maes -march=native -O3 -DTRACE -flto -c src/main.cpp -o obj/main.o
make
bin/randomx --verify --softAes
@tevador , i fixed it on that VM. When i changed the makefile earlier today i didn't make clean, so yeah. I
main12@main12-kvm:~/randomx_test_20190220/RandomX$ bin/randomx --verify --softAes
Using software AES.
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 3.03045 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851
Performance: 193.019 ms per hash
AMD Ryzen 5 2600 (6 cores), DDR4 2666 MHz (all stock), Windows 10.
Precompiled binary:
RandomX - mining mode
Initializing (12 threads) ...
Dataset (4 GiB) initialized in 10.3716 s
Initializing 6 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: abe4b79b0439e54e2896e0134cc89c8bb1b7a378192095efc690f8f85699a225
Performance: 3966.78 hashes per second
Binary compiled with GCC 8.2.0 on msys64:
RandomX - mining mode
Initializing (12 threads) ...
Dataset (4 GiB) initialized in 10.4535 s
Initializing 6 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: abe4b79b0439e54e2896e0134cc89c8bb1b7a378192095efc690f8f85699a225
Performance: 3967.49 hashes per second
I have pushed a new release with a different form of division instruction (issue #26). Slight performance increase can be expected (1-5%).
i5-4310U 12GB (hp840G1 mint19.1)
numactl -i all ./randomx --mine --nonces 100000 --init 1 --threads 1
RandomX - mining mode
Initializing (1 thread) ...
Dataset (4 GiB) initialized in 108.793 s
Initializing 1 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 5ae89ddb121558190540372d3b93189b490ba44435dee8c72ebde7cd976d8a5e
Performance: 304.541 hashes per second
numactl -i all ./randomx --mine --nonces 100000 --init 1 --threads 2
RandomX - mining mode
Initializing (1 thread) ...
Dataset (4 GiB) initialized in 112.114 s
Initializing 2 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 5ae89ddb121558190540372d3b93189b490ba44435dee8c72ebde7cd976d8a5e
Performance: 513.354 hashes per second
numactl -i all ./randomx --mine --nonces 100000 --init 2 --threads 1
RandomX - mining mode
Initializing (2 threads) ...
Dataset (4 GiB) initialized in 60.5406 s
Initializing 1 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 5ae89ddb121558190540372d3b93189b490ba44435dee8c72ebde7cd976d8a5e
Performance: 298.663 hashes per second
numactl -i all ./randomx --mine --nonces 100000 --init 2 --threads 2
RandomX - mining mode
Initializing (2 threads) ...
Dataset (4 GiB) initialized in 60.4542 s
Initializing 2 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 5ae89ddb121558190540372d3b93189b490ba44435dee8c72ebde7cd976d8a5e
Performance: 502.217 hashes per second
Intel Xeon E5-2651 v2 1.80GHz (2 processors) 8Gb Memory
RandomX - verification mode Initializing ... Cache (256 MiB) initialized in 1.23143 s Initializing 1 virtual machine(s) ... Running benchmark (1000 nonces) ... Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26 851 Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26 851 Performance: 85.3171 ms per hash
RandomX - mining mode Initializing (12 threads) ... Dataset (4 GiB) initialized in 21.3293 s Initializing 24 virtual machine(s) ... Running benchmark (100000 nonces) ... Calculated result: abe4b79b0439e54e2896e0134cc89c8bb1b7a378192095efc690f8f85699a 225 Performance: 3717.61 hashes per second
i5-4310U 12GB (hp840G1 mint19.1)
randomx --verify
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 0.890826 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: d3ae5a9365196ed48bb98ebfc3316498e29443ea7f056ecbd272f749c6af7730
Reference result: d3ae5a9365196ed48bb98ebfc3316498e29443ea7f056ecbd272f749c6af7730
Performance: 46.1011 ms per hash
randomx --verify --softAes
Using software AES.
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 0.918776 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: d3ae5a9365196ed48bb98ebfc3316498e29443ea7f056ecbd272f749c6af7730
Reference result: d3ae5a9365196ed48bb98ebfc3316498e29443ea7f056ecbd272f749c6af7730
Performance: 45.9135 ms per hash
i5-8500 8GB (Windows10beta1803)
randomx --verify
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 0.643638 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: d3ae5a9365196ed48bb98ebfc3316498e29443ea7f056ecbd272f749c6af7730
Reference result: d3ae5a9365196ed48bb98ebfc3316498e29443ea7f056ecbd272f749c6af7730
Performance: 31.3919 ms per hash
randomx --verify --softAes
Using software AES.
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 0.640813 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: d3ae5a9365196ed48bb98ebfc3316498e29443ea7f056ecbd272f749c6af7730
Reference result: d3ae5a9365196ed48bb98ebfc3316498e29443ea7f056ecbd272f749c6af7730
Performance: 33.0118 ms per hash
randomx --mine --nonces 100000 --init 1 --threads 4
RandomX - mining mode
Initializing (1 thread) ...
Dataset (4 GiB) initialized in 82.6982 s
Initializing 4 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 5ae89ddb121558190540372d3b93189b490ba44435dee8c72ebde7cd976d8a5e
Performance: 1498.43 hashes per second
randomx --mine --nonces 100000 --init 4 --threads 4
RandomX - mining mode
Initializing (4 threads) ...
Dataset (4 GiB) initialized in 23.553 s
Initializing 4 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 5ae89ddb121558190540372d3b93189b490ba44435dee8c72ebde7cd976d8a5e
Performance: 1491.41 hashes per second
randomx --mine --nonces 100000 --init 6 --threads 6
RandomX - mining mode
Initializing (6 threads) ...
Dataset (4 GiB) initialized in 17.3481 s
Initializing 6 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 5ae89ddb121558190540372d3b93189b490ba44435dee8c72ebde7cd976d8a5e
Performance: 1706.28 hashes per second
randomx --mine --nonces 100000 --init 6 --threads 5
RandomX - mining mode
Initializing (6 threads) ...
Dataset (4 GiB) initialized in 17.3604 s
Initializing 5 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 5ae89ddb121558190540372d3b93189b490ba44435dee8c72ebde7cd976d8a5e
Performance: 1648.45 hashes per second
randomx --mine --nonces 100000 --init 6 --threads 8
RandomX - mining mode
Initializing (6 threads) ...
Dataset (4 GiB) initialized in 17.2848 s
Initializing 8 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 5ae89ddb121558190540372d3b93189b490ba44435dee8c72ebde7cd976d8a5e
Performance: 1700.84 hashes per second
AMD Ryzen 6 9850 (128 cores) 256Gb Memory
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 0.27183 s
Initializing 128 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: fs3leefd7ed1cecl73e87c06b4b66b38e58b23ab255e8f9c6b62745e43a26851
Reference result: fs3leefd7ed1cecl73e87c06b4b66b38e58b23ab255e8f9c6b62745e43a26851
Performance: 1.3281 ms per hash
HP DL580 G7: 4x E7-4830 (24M Cache, 2.13 GHz, 8 core) 64 GB DDR3-1066 (16 modules x 4GB, 16GB quad channel per node) Ubuntu Server 18.04
verification mode:
$ ./randomx --verify --largePages
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 1.6193 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: d3ae5a9365196ed48bb98ebfc3316498e29443ea7f056ecbd272f749c6af7730
Reference result: d3ae5a9365196ed48bb98ebfc3316498e29443ea7f056ecbd272f749c6af7730
Performance: 113.487 ms per hash
mining mode (Power Reading at iLO3 = 726 Watts; Idle = 323 Watts)
$ seq 0 3 | xargs -P 0 -I node numactl -N node ./randomx --mine --largePages --nonces 1000000 --init 16 --threads 8
RandomX - mining mode
Initializing (16 threads) ...
RandomX - mining mode
Initializing (16 threads) ...
RandomX - mining mode
Initializing (16 threads) ...
RandomX - mining mode
Initializing (16 threads) ...
Dataset (4 GiB) initialized in 23.0033 s
Initializing 8 virtual machine(s) ...
Running benchmark (1000000 nonces) ...
Dataset (4 GiB) initialized in 23.0544 s
Initializing 8 virtual machine(s) ...
Running benchmark (1000000 nonces) ...
Dataset (4 GiB) initialized in 23.1518 s
Initializing 8 virtual machine(s) ...
Running benchmark (1000000 nonces) ...
Dataset (4 GiB) initialized in 23.2302 s
Initializing 8 virtual machine(s) ...
Running benchmark (1000000 nonces) ...
Calculated result: b13f5038492d3abfd14da6671f92c28456d96c29a5ffb120edac924abe5aabe9
Performance: 1968.31 hashes per second
Calculated result: b13f5038492d3abfd14da6671f92c28456d96c29a5ffb120edac924abe5aabe9
Performance: 1966.13 hashes per second
Calculated result: b13f5038492d3abfd14da6671f92c28456d96c29a5ffb120edac924abe5aabe9
Performance: 1961.65 hashes per second
Calculated result: b13f5038492d3abfd14da6671f92c28456d96c29a5ffb120edac924abe5aabe9
Performance: 1951.87 hashes per second
Single process mining (Power Reading at iLO3 = 704 Watts):
$ numactl -i all ./randomx --mine --largePages --nonces 1000000 --init 64 --threads 32
RandomX - mining mode
Initializing (64 threads) ...
Dataset (4 GiB) initialized in 8.9006 s
Initializing 32 virtual machine(s) ...
Running benchmark (1000000 nonces) ...
Calculated result: b13f5038492d3abfd14da6671f92c28456d96c29a5ffb120edac924abe5aabe9
Performance: 6774.47 hashes per second
@kdovijak
AMD Ryzen 6 9850 (128 cores)
AFAIK there is no such CPU and also your result doesn't match any version of RandomX. Can you please fix your comment and include the whole command line?
Intel i7-4770 @3.4Ghz
RandomX - verification mode Initializing ... Cache (256 MiB) initialized in 0.6833 s Initializing 1 virtual machine(s) ... Running benchmark (1000 nonces) ... Calculated result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851 Reference result: fe31e8fd7ed1cec773e87c0684b66b38e58b23ab255e8f9c6b62745e43a26851 Performance: 31.9503 ms per hash
RandomX - mining mode Initializing (8 threads) ... Dataset (4 GiB) initialized in 16.6505 s Initializing 4 virtual machine(s) ... Running benchmark (100000 nonces) ... Calculated result: abe4b79b0439e54e2896e0134cc89c8bb1b7a378192095efc690f8f85699a225 Performance: 1821.36 hashes per second
I am sorry for that. Can you remove it please? It do not belong here, My mistake. I've never been here
@kdovijak AFAIK there is no such CPU and also your result doesn't match any version of RandomX. Can you please fix your comment and include the whole command line?
I am sorry for that. Can you remove it please? It do not belong here, My mistake. I've never been here
@kdovijak AFAIK there is no such CPU and also your result doesn't match any version of RandomX. Can you please fix your comment and include the whole command line?
What a curious thing to generate and post here
i7 3770K (cores boosted to 3.7-3.9 GHz during the test) ram is 16GB (4x4) dual channel DDR3 1866MHz 11-11-11-32
(randomx.exe --verify --largePages) RandomX - verification mode Initializing ... Cache (256 MiB) initialized in 0.695975 s Initializing 1 virtual machine(s) ... Running benchmark (1000 nonces) ... Calculated result: d3ae5a9365196ed48bb98ebfc3316498e29443ea7f056ecbd272f749c6af7730 Reference result: d3ae5a9365196ed48bb98ebfc3316498e29443ea7f056ecbd272f749c6af7730 Performance: 51.3777 ms per hash
(randomx.exe --mine --init 4 --threads 4 --nonces 10000 --largePages) RandomX - mining mode Initializing (4 threads) ... ERROR: allocLargePagesMemory - VirtualAlloc: Insufficient system resources exist to complete the requested service.
Anyone know what the problem is? Task manager shows 5GB of 16GB RAM used.
edit tried again after PC reboot and now results in BSOD
MEMORY_MANAGEMENT (1a)
Arguments: Arg1: 0000000000061941, The subtype of the bugcheck. Arg2: 000001a080000000 Arg3: 000000000000000f Arg4: fffff98644f87b00
KEY_VALUES_STRING: 1
STACKHASH_ANALYSIS: 1
TIMELINE_ANALYSIS: 1
DUMP_CLASS: 1
DUMP_QUALIFIER: 400
BUILD_VERSION_STRING: 17763.1.amd64fre.rs5_release.180914-1434
DUMP_TYPE: 2
BUGCHECK_P1: 61941
BUGCHECK_P2: 1a080000000
BUGCHECK_P3: f
BUGCHECK_P4: fffff98644f87b00
BUGCHECK_STR: 0x1a_61941
CPU_COUNT: 8
CPU_MHZ: db6
CPU_VENDOR: GenuineIntel
CPU_FAMILY: 6
CPU_MODEL: 3a
CPU_STEPPING: 9
CUSTOMER_CRASH_COUNT: 1
DEFAULT_BUCKET_ID: WIN8_DRIVER_FAULT
PROCESS_NAME: randomx.exe
CURRENT_IRQL: 0
ANALYSIS_SESSION_HOST: DESKTOP-UDEN2FS
ANALYSIS_SESSION_TIME: 02-23-2019 19:22:26.0824
ANALYSIS_VERSION: 10.0.17763.132 x86fre
LAST_CONTROL_TRANSFER: from fffff8011604ae94 to fffff80115fb4440
STACK_TEXT:
fffff98644f87958 fffff801
1604ae94 : 000000000000001a 00000000
00061941 000001a080000000 00000000
0000000f : nt!KeBugCheckEx
fffff98644f87960 fffff801
15fc1f83 : ffffad0e654ac080 ffffad0e
65918080 0000000000000000 ffffad0e
65e42d60 : nt!MmAccessFault+0x142864
fffff98644f87b00 00007ff6
f7d173cb : 0000000000000000 00000000
00000000 0000000000000000 00000000
00000000 : nt!KiPageFault+0x343
000000289e5ff870 00000000
00000000 : 0000000000000000 00000000
00000000 0000000000000000 00000000
00000000 : 0x00007ff6`f7d173cb
THREAD_SHA1_HASH_MOD_FUNC: 1feb98d5c6601cfab1575099bc1303f202ec7478
THREAD_SHA1_HASH_MOD_FUNC_OFFSET: 37bfc2bc4e861c90c544e608469bfd3b3b1f7f1d
THREAD_SHA1_HASH_MOD: 2a7ca9d3ab5386d53fea7498e1d81b9c4a4c036b
SYMBOL_NAME: ANALYSIS_INCONCLUSIVE
FOLLOWUP_NAME: MachineOwner
MODULE_NAME: Unknown_Module
IMAGE_NAME: Unknown_Image
DEBUG_FLR_IMAGE_TIMESTAMP: 0
STACK_COMMAND: .thread ; .cxr ; kb
FAILURE_BUCKET_ID: 0x1a_61941_ANALYSIS_INCONCLUSIVE!unknown_function
BUCKET_ID: 0x1a_61941_ANALYSIS_INCONCLUSIVE!unknown_function
PRIMARY_PROBLEM_CLASS: 0x1a_61941_ANALYSIS_INCONCLUSIVE!unknown_function
TARGET_TIME: 2019-02-23T17:04:54.000Z
OSBUILD: 17763
OSSERVICEPACK: 316
SERVICEPACK_NUMBER: 0
OS_REVISION: 0
SUITE_MASK: 272
PRODUCT_TYPE: 1
OSPLATFORM_TYPE: x64
OSNAME: Windows 10
OSEDITION: Windows 10 WinNt TerminalServer SingleUserTS
OS_LOCALE:
USER_LCID: 0
OSBUILD_TIMESTAMP: unknown_date
BUILDDATESTAMP_STR: 180914-1434
BUILDLAB_STR: rs5_release
BUILDOSVER_STR: 10.0.17763.1.amd64fre.rs5_release.180914-1434
ANALYSIS_SESSION_ELAPSED_TIME: 822
ANALYSIS_SOURCE: KM
FAILURE_ID_HASH_STRING: km:0x1a_61941_analysis_inconclusive!unknown_function
FAILURE_ID_HASH: {5deef4f7-7ffd-da2a-fbcb-ad563ba13c90}
@JustFranz Try running memtest86. RandomX definitely cannot cause BSOD. It looks like defective memory.
Just finsished memtest86, 4 passes 0 errors. My BIOS is latest for mobo. My ram is in 4x4 GB configuration. Windows is up to date. Correction to RAM speed 1866 MHz 11-11-11-32
Just tested if BSOD was fluke, its reproducible 2/2.
Tried again in safe mode to rule out anything with windows, BSOD
have not used linux in my life (technically a lie), but going to give it a shot. What is a good USB bootable version?
What is a good USB bootable version?
Ubuntu 18.04 LTS.
The cause of the BSOD can be either:
./randomx --mine --init 4 --threads 4 --nonces 400000 --largePages RandomX - mining mode Initializing (4 threads) ... ERROR: allocLargePagesMemory - mmap failed
./randomx --mine --init 4 --threads 4 --nonces 400000 RandomX - mining mode Initializing (4 threads) ... Dataset (4 GiB) initialized in 23.9515 s Initializing 4 virtual machine(s) ... Running benchmark (400000 nonces) ... Calculated result: 655eec9f61d57e27b435ea1b6aed50c56e90f586e1fa786349ce4758c2cfc318 Performance: 1340.1 hashes per second
Large pages is failing for some reason, I was under the impression that it was enabled by default for linux. What is the command to enter into terminal to properly enable it?
result in windows without --largePages RandomX - mining mode Initializing (4 threads) ... Dataset (4 GiB) initialized in 23.9145 s Initializing 4 virtual machine(s) ... Running benchmark (400000 nonces) ... Calculated result: 655eec9f61d57e27b435ea1b6aed50c56e90f586e1fa786349ce4758c2cfc318 Performance: 1348.07 hashes per second
I have largePages enabled, lock pages in memory has Administrators, MyUserName and *S-1-5-32-500 Why the BSOD? Any ideas?
sfc /scannow found 1 corrupt file and repaired it, still BSOD
Large pages is failing for some reason, I was under the impression that it was enabled by default for linux. What is the command to enter into terminal to properly enable it?
https://github.com/fireice-uk/xmr-stak/blob/master/doc/FAQ.md#error-memory-alloc-failed-mmap-failed
note that RandomX will require larger numbers than xmr-stak (2304 for vm.nr_hugepages
and 4718592 for ulimit -l
should be enough).
AMD Ryzen 1800X 3.6GHs, 8GB Single 2400MHz, Windows 10 64bit
D:\RandomX-windows-x64-v0.3.1-beta>randomx.exe --verify RandomX - verification mode Initializing ... Cache (256 MiB) initialized in 0.754375 s Initializing 1 virtual machine(s) ... Running benchmark (1000 nonces) ... Calculated result: d3ae5a9365196ed48bb98ebfc3316498e29443ea7f056ecbd272f749c6af7730 Reference result: d3ae5a9365196ed48bb98ebfc3316498e29443ea7f056ecbd272f749c6af7730 Performance: 56.095 ms per hash
TR1950X @3.9Ghz RandomX - verification mode Initializing ... Cache (256 MiB) initialized in 0.679985 s Initializing 1 virtual machine(s) ... Running benchmark (1000 nonces) ... Calculated result: d3ae5a9365196ed48bb98ebfc3316498e29443ea7f056ecbd272f749c6af7730 Reference result: d3ae5a9365196ed48bb98ebfc3316498e29443ea7f056ecbd272f749c6af7730 Performance: 53.2547 ms per hash
init 32 --threads 16 --nonces 1000000 --largePages RandomX - mining mode Initializing (32 threads) ... Dataset (4 GiB) initialized in 5.08343 s Initializing 16 virtual machine(s) ... Running benchmark (1000000 nonces) ... Calculated result: b13f5038492d3abfd14da6671f92c28456d96c29a5ffb120edac924abe5aabe9 Performance: 8917.85 hashes per second
@seizahub you should try running two mining processes at once with 8 threads each, one process on each numa node.
I have pushed a new release which fixes some portability issues of the interpreter.
No impact on performance.
Intel Core i5-8250U CPU @ 1.60GHz 8GB DDR4, L3 Cache 6 MB
$ ./randomx --verify
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 0.78788 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006
Reference result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006
Performance: 41.3737 ms per hash
./randomx --mine --init 8 --threads 6 --nonces 100000
RandomX - mining mode
Initializing (8 threads) ...
Dataset (4 GiB) initialized in 20.712 s
Initializing 6 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 6b75c9504202c15bc58d3579bb37f029c3fa711bad4a8182d7284ff612dc08f2
Performance: 1031.32 hashes per second
Compiled commit d9bc6cf
(tag/0.3.2-beta), Arch Linux, gcc 8.2.1 20181127
RandomX v0.3.2 release binaries
boosts to 4 GHz when verifying
>randomx.exe --verify --largePages
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 0.580229 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006
Reference result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006
Performance: 27.93 ms per hash
boosts to 2.6 GHz when mining
>randomx.exe --mine --largePages --init 8 --threads 4 --nonces 100000
RandomX - mining mode
Initializing (8 threads) ...
Dataset (4 GiB) initialized in 14.2677 s
Initializing 4 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 6b75c9504202c15bc58d3579bb37f029c3fa711bad4a8182d7284ff612dc08f2
Performance: 1718.28 hashes per second
3.35 GHz
> bin/randomx --mine --largePages --init 16 --threads 8 --nonces 1000000
RandomX - mining mode
Initializing (16 threads) ...
Dataset (4 GiB) initialized in 9.46747 s
Initializing 8 virtual machine(s) ...
Running benchmark (1000000 nonces) ...
Calculated result: 38f1097eb1f1a77973559d8142273db7de8e679303b6cdb6c46ea9e6420c5e8d
Performance: 4194.87 hashes per second
Ubuntu 16.04 LTS
erk@erk-PORTEGE-R830:~/.grin/main$ lscpu
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Byte Order: Little Endian
CPU(s): 4
On-line CPU(s) list: 0-3
Thread(s) per core: 2
Core(s) per socket: 2
Socket(s): 1
NUMA node(s): 1
Vendor ID: GenuineIntel
CPU family: 6
Model: 42
Model name: Intel(R) Core(TM) i7-2620M CPU @ 2.70GHz
Stepping: 7
CPU MHz: 1074.740
CPU max MHz: 3400.0000
CPU min MHz: 800.0000
BogoMIPS: 5382.61
Virtualization: VT-x
L1d cache: 32K
L1i cache: 32K
L2 cache: 256K
L3 cache: 4096K
NUMA node0 CPU(s): 0-3
erk@erk-PORTEGE-R830:~/Brain/RandomX/RandomX/bin$ ./randomx --mine --init 4 --threads 2 --nonces 100000 --largepages
RandomX - mining mode
Initializing (4 threads) ...
Dataset (4 GiB) initialized in 34.483 s
Initializing 2 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 6b75c9504202c15bc58d3579bb37f029c3fa711bad4a8182d7284ff612dc08f2
Performance: 505.33 hashes per second
It's broken on powerpc big endian, little endian seems to be working now.
Added "|| define(powerpc)" to trigger the ifdef for rotr in softAes.cpp
Powermac G5 (PPC970 BE)
jribeiro@powermagi ~/Development/RandomX/bin $ ./randomx --verify --softAes --largePages Using software AES. RandomX - verification mode Initializing ... Cache (256 MiB) initialized in 4.43579 s Initializing 1 virtual machine(s) ... Running benchmark (1000 nonces) ... Calculated result: 883c628e4f5045c4cfcdfead388640e2dfe3e81400f8566fd7b3d3e1d1f98626 Reference result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006 Performance: 513.212 ms per hash
IBM 821LC (POWER8 LE)
Using software AES. RandomX - verification mode Initializing ... Cache (256 MiB) initialized in 1.2094 s Initializing 1 virtual machine(s) ... Running benchmark (1000 nonces) ... Calculated result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006 Reference result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006 Performance: 208.994 ms per hash
And power8 running 80 threads in 10 cores:
Using software AES. RandomX - verification mode Initializing ... Cache (256 MiB) initialized in 1.04643 s Initializing 80 virtual machine(s) ... Running benchmark (1000 nonces) ... Calculated result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006 Reference result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006 Performance: 6.64076 ms per hash
@nioroso-x3 Thanks. I will prepare a diagnostic version of RandomX, so we can find the cause of the issue on big endian systems.
Tevador,
Looks like moving to W10 helps. and the sweet spot is 19 Threads. 2x Xeons E5-2651 v2 1.80Ghz server
RandomX - verification mode Initializing ... Cache (256 MiB) initialized in 1.35356 s Initializing 1 virtual machine(s) ... Running benchmark (1000 nonces) ... Calculated result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006 Reference result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006 Performance: 93.7652 ms per hash
RandomX - mining mode Initializing (24 threads) ... Dataset (4 GiB) initialized in 18.8612 s Initializing 19 virtual machine(s) ... Running benchmark (100000 nonces) ... Calculated result: 6b75c9504202c15bc58d3579bb37f029c3fa711bad4a8182d7284ff612dc08f2 Performance: 5005.93 hashes per second
Here are my test results. All binaries compiled by myself using latest commit (d9bc6cf - v0.3.2-beta).
CPU: Intel(R) Core(TM) i7-7500U CPU @ 2.70GHz, 400 MHz - 3.5 GHz RAM: 2x8 GB DDR4 2133 MHz OS: Ubuntu 16.04 GCC version 5.4.0 20160609 (Ubuntu 5.4.0-6ubuntu1~16.04.11)
./randomx --verify
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 0.658071 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006
Reference result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006
Performance: 33.8129 ms per hash
./randomx --verify --largePages
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 0.550309 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006
Reference result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006
Performance: 30.0763 ms per hash
./randomx --mine --largePages --threads 2 --init 4 --nonces 100000
RandomX - mining mode
Initializing (4 threads) ...
Dataset (4 GiB) initialized in 25.2313 s
Initializing 2 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 6b75c9504202c15bc58d3579bb37f029c3fa711bad4a8182d7284ff612dc08f2
Performance: 1020.34 hashes per second
./randomx --mine --largePages --threads 4 --init 4 --nonces 100000
RandomX - mining mode
Initializing (4 threads) ...
Dataset (4 GiB) initialized in 25.2198 s
Initializing 4 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 6b75c9504202c15bc58d3579bb37f029c3fa711bad4a8182d7284ff612dc08f2
Performance: 1064.74 hashes per second
./randomx --mine --threads 2 --init 4 --nonces 100000
RandomX - mining mode
Initializing (4 threads) ...
Dataset (4 GiB) initialized in 27.1011 s
Initializing 2 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 6b75c9504202c15bc58d3579bb37f029c3fa711bad4a8182d7284ff612dc08f2
Performance: 680.794 hashes per second
./randomx --mine --threads 4 --init 4 --nonces 100000
RandomX - mining mode
Initializing (4 threads) ...
Dataset (4 GiB) initialized in 27.1022 s
Initializing 4 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 6b75c9504202c15bc58d3579bb37f029c3fa711bad4a8182d7284ff612dc08f2
Performance: 909.257 hashes per second
CPU: Intel(R) Core(TM) i3 CPU M 370 @ 2.40GHz, 933 MHz - 2.4 GHz RAM: 2x4 GB DDR3 1066MHz OS: Ubuntu 16.04 GCC version 5.4.0 20160609 (Ubuntu 5.4.0-6ubuntu1~16.04.11)
./randomx --verify --softAes
Using software AES.
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 1.58066 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006
Reference result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006
Performance: 117.928 ms per hash
./randomx --verify --softAes --largePages
Using software AES.
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 1.42848 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006
Reference result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006
Performance: 112.494 ms per hash
./randomx --mine --init 4 --threads 2 --nonces 100000 --softAes
Using software AES.
RandomX - mining mode
Initializing (4 threads) ...
Dataset (4 GiB) initialized in 76.4672 s
Initializing 2 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 6b75c9504202c15bc58d3579bb37f029c3fa711bad4a8182d7284ff612dc08f2
Performance: 221.167 hashes per second
./randomx --mine --init 4 --threads 2 --nonces 100000 --softAes --largePages
Using software AES.
RandomX - mining mode
Initializing (4 threads) ...
Dataset (4 GiB) initialized in 72.725 s
Initializing 2 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 6b75c9504202c15bc58d3579bb37f029c3fa711bad4a8182d7284ff612dc08f2
Performance: 257.854 hashes per second
./randomx --mine --init 4 --threads 4 --nonces 100000 --softAes
Using software AES.
RandomX - mining mode
Initializing (4 threads) ...
Dataset (4 GiB) initialized in 76.8335 s
Initializing 4 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 6b75c9504202c15bc58d3579bb37f029c3fa711bad4a8182d7284ff612dc08f2
Performance: 271.51 hashes per second
./randomx --mine --init 4 --threads 4 --nonces 100000 --softAes --largePages
Using software AES.
RandomX - mining mode
Initializing (4 threads) ...
Dataset (4 GiB) initialized in 72.7733 s
Initializing 4 virtual machine(s) ...
Running benchmark (100000 nonces) ...
Calculated result: 6b75c9504202c15bc58d3579bb37f029c3fa711bad4a8182d7284ff612dc08f2
Performance: 309.146 hashes per second
NanoPi M3 CPU: Samsung S5P6818 Octa-Core Cortex-A53, 400M Hz - 1.4G Hz RAM: 1GB DDR3 OS: Armbian Bionic 18.04 GCC version 7.3.0 (Ubuntu/Linaro 7.3.0-27ubuntu1~18.04)
./randomx --verify --softAes
Using software AES.
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 3.79737 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006
Reference result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006
Performance: 444.42 ms per hash
./randomx --verify --softAes --largePages
Using software AES.
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 3.68037 s
Initializing 1 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006
Reference result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006
Performance: 403.769 ms per hash
@nioroso-x3
I found one case when endianness was not properly handled. Could you retest latest master on big endian?
If the result still doesn't match, please run
./randomx --genAsm --nonces 42
and compare the output with file src/program.inc. It should be the same code.
I ran a longer verification test (~40 hours) on Raspberry Pi 3 (ARMv7+Neon). Result matches x86.
bin/randomx --verify --softAes --threads 4 --nonces 1000000
Using software AES.
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 8.67455 s
Initializing 4 virtual machine(s) ...
Running benchmark (1000000 nonces) ...
Calculated result: 38f1097eb1f1a77973559d8142273db7de8e679303b6cdb6c46ea9e6420c5e8d
Performance: 145.724 ms per hash
Raspberry Pi 3 running 64-bit Arch Linux
$ ./randomx --verify --softAes --largePages --threads 4
Using software AES.
RandomX - verification mode
Initializing ...
Cache (256 MiB) initialized in 6.91513 s
Initializing 4 virtual machine(s) ...
Running benchmark (1000 nonces) ...
Calculated result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006
Reference result: e1b4144293ff9ab5aa4c98f2389bb18950d8c3fd874891ac64628e028a286006
Performance: 231.564 ms per hash
Please post:
Precompiled binaries are available for Windows 64-bit and Ubuntu/Debian x86-64. Download the latest release here: https://github.com/tevador/RandomX/releases
Verification mode
Run as:
--jit
option to get 2-3 times faster verification.--softAes
option to get a small increase in performance.--largePages
option to get a small increase in performance.Mining mode
Mining mode is currently only supported on 64-bit x86 CPUs. Requires at least 2.25 GiB of RAM. Run as:
Q
(number of initialization threads) equal to the number of hardware threads of your CPU.T
(number of mining threads) which produces the highest hashrate. Starting point should be 1 thread per 2 MB of L3 cache, but some CPUs can benefit from running more threads, while some CPUs cannot run more than 1 thread per core efficiently depending on other factors such as L1/L2 cache sizes.N
(number of nonces) equal to10000
,100000
or1000000
depending on the performance of your system. Aim for at least a 60-second mining period for accurate results.--softAes
option. Your mining performance will be about 40% lower.--largePages
option to get a significant increase in performance (up to 25%).