-
Notifications
You must be signed in to change notification settings - Fork 16
/
Copy pathRELEASE_ANNOUNCEMENT
124 lines (101 loc) · 4.04 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
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
Release 3.1.16 of cpuminer-opt is available for download.
All users are encouraged to upgrade.
cpuminer-opt now supports 32 algorithms on CPUs with at least SSE2
capabilities including Intel Core2 and AMD equivalent. See the
performance chart below for details.
In addition 13 algorithms have optimizations to take advantage of
CPUs with AES_NI for even greater performance, including the Intel
Sandy Bridge (Core-i 2xxx) and AMD equivalent.
New in v3.1.16
- added m7m algo
- new dependency on gmp
New in v3.1.15
- unified build procedure fixed
- build.sh now works for CPUs with and without AES_NI
- it is no longer necessary to add -DNO_AES_NI CFLAG to the
configure command when building for CPUs without AES_NI.
Recently added dependencies:
- gmp
- libboost-dev
- libboost-system-dev
- libboost-thread-dev
Recently added algos:
- hodl added with support for CPUs with and without AES_NI
- argon2
- decred (blake256r14)
- vanilla (blake256r8vnl)
- x14 added with AES_NI and SSE2 optimizations
- blake
- blake2s
- yescrypt
- scryptjane
- scrypt
- sha256d
Recently improved algos:
- x11gost (sib) optimised:
+116% AES_NI + SSE2
+63% SSE2
- nist5 optimised tor AES_NI & SSE2
+111% AES_NI + SSE2
+68% SSE2
- c11 +57% optimized for AES_NI & SSE2
- zr5 optimized for AES_NI and SSE2
+98% AES_NI + SSE2
+45% SSE2
- myr-gr optimized for AES_NI, marginal increase in hashrate
Users with non-SSE2 CPUs or whoi want to mine algos not supported by
cpuminer-opt may find cpuminer-multi by TPruvot useful.
The performance chart below is for an Intel i7-6700K @ 4 GHz, 16 GB mem.
Normalization rates have ben 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 * normalization rate
AES_NI SSE2(1) norm rate(5)
------ ---- ---------
x11 740 K 525 K 1
x13 373 298 0.50
x14 357 271 0.48
x15 333 270 0.45
x17(2) 317 248 0.43
quark 1195 924 1.61
qubit 1090 765 1.45
nist5 2000 1592 3.37
zr5 850 650 1.15
x11gost 540 404 0.73
c11 733 475 0.99
myr-gr 1572 1560 2.12
cryptonight 290 H 165 H 0.00039
hodl 275 200 H 0.00037
neoscrypt (4) 32 K 0.043
lyra2re (4) 930 1.25
lyra2rev2 (4) 495 0.65
argon2 (4) 33.7 0.045
m7m (4) 48.7 0.066
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
blake (4) 22.4 M 30.4
blake2s (4) 17.5 23.6
vanilla (4) 33.0 44.6
decred (4) 22.6 30.5
axiom (4) 72 H 0.098
yescrypt (4) 3760 0.0051
scryptjane (4) 250 0.00034
blakecoin(2) (4) 33.9 M 45.8
cryptolight(2) (4) 685 H 0.00093
pluck(2) (4) 1925 H 0.0026
drop(2) (4) 934 K 1.26
fresh(2) (4) 528 K 0.71
Footnotes:
(1) SSE2 rates are simulated in software.
(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
Unlisted algos are in an unknown state