This repository has been archived by the owner on Oct 21, 2019. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 25
/
RELEASE_ANNOUNCEMENT
105 lines (89 loc) · 4.21 KB
/
RELEASE_ANNOUNCEMENT
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
Release 3.4.6 of cpuminer-opt is available for download with support for
Windows.
cpuminer-opt now supports 38 algorithms on CPUs with at least SSE2
capabilities including Intel Core2, Nehalem and AMD equivalent. See the
performance chart below for details.
In addition 19 algorithms have optimizations to take advantage of
CPUs with AES_NI for even greater performance, including the Intel
Westbridge and newer and AMD equivalent. See the performance
comparison below.
New in 3.4.6
For users:
- cryptolight algo is now supported with AES otimizations
- display format changed for share submissions
- colour keyed "Accepted" or "Rejected" status.
- reject count and rate displayed when share is rejected.
For developers:
- code restructuring for detecting new work
- cleaned up detection and handling of new work
- removed call to stratum_gen_work from niner_thread.
- eliminated gen_work_now gate function.
- renamed gate function init_nonce to get_new_work.
- renamed gate function alloc_scratchbuf to miner_thread_init,
removed all scracthbuf references from miner_thread and moved
implementation to the local algo files of those algos that need it.
- moved most gate targets from algo-gate.c to cpu-miner.c removing
most mining related code from algo-gate-api.c.
Users with non-SSE2 CPUs or who want to mine algos not supported by
cpuminer-opt may find cpuminer-multi by TPruvot useful.
Chart out of date, will be removed.
The performance chart below is for an Intel i7-6700K @ 4 GHz, 16 GB mem.
Normalization rates have been added to the chart to help with profit
switching pools. Reference algo x11 = 1.
Due to the peculiarities of some algorithms their performance on other CPU
architectures may not scale equally. Their normalizations rates will also
differ from those listed below. YMMV.
Normalized profitability = algo profitability * norm rate
AES-AVX SSE2(1) norm rate(5)
------- ------- ---------
x11 780 K 525 K 1
x13 392 298 0.50
x14 370 271 0.48
x15 341 270 0.45
x17 317 248 0.43
x11gost 562 392 0.72
x11evo 590 387 0.78
quark 1195 924 1.61
qubit 1182 765 1.45
nist5 2000 1592 3.37
zr5 850 650 1.15
c11 784 475 0.99
myr-gr 1572 1560 2.12
hmq1725 214 161 0.29
m7m 121 77.4 0.155
lyra2re 1380 900 1.76
lyra2rev2 1350 980 1.73
cryptonight 290 H 165 H 0.00039
cryptolight 685 ? 0.00093
hodl 600 200 0.00081
lbry (4) 2620 3.53
neoscrypt (4) 32 K 0.043
argon2 (4) 33.7 0.045
groestl (4) 931 1.26
skein (4) 5747 7.77
skein2 (4) 8675 11.7
pentablake (4) 3960 5.35
keccak (4) 7790 10.5
scrypt (4) 113 0.153
sha256d (4) 62.5 0.084
veltor (4) 1017 1.30
blake (4) 22.4 M 30.4
blake2s (4) 19.0 25.7
vanilla (4) 33.0 44.6
blakecoin (4) 33.9 45.8
decred (4) 22.6 30.5
axiom (4) 72 H 0.000098
yescrypt (4) 3760 0.0051
scryptjane (4) 250 0.00034
pluck(2) (4) 1925 0.0026
drop(2) (4) 934 K 1.26
fresh(2) (4) 528 0.71
whirlpool(2) (4) 1290 1.74
whirlpoolx(2) (4) 5110 6.9
Footnotes:
(1) SSE2 rates are simulated in software (-march=core2) on an i7.
(2) Benchmark tested only
(3) CPU architecture not supported for algo. It won't work.
(4) AES_NI Optimization not available for CPU artchitecture. Uses SSE2, slower.
(5) Normalised profitability = algo profitability * norm rate, x11 = 1
(6) Not supported on Windows