Mining hardware comparison

Below are some statistics about the mining performance of various hardware used in a mining rig.

The table shows (mostly) stock clock numbers. 10-20% performance improvement can be achieved via overclocking.

Note: Clock refers to the Shader clock only with nVidia cards (not Core or Memory). With AMD card the shader clock is not separate, but is part of the GPU clock.

Graphics cards

AMD

To get the maximum performance use the 2.1 release of the ATI Stream SDK. 2.2 wastes CPU time, and 2.3/2.4 drop mining performance by 5-10%.

Model	Mhash/s	Mhash/J	W	Clock	SP	SDK	Slot	Miner	Notes
3XXX								OpenCL Not Supported
42XX							PCI-E 2.0 x16		OpenCL Not Supported (intergrated/mobile GPU)
4350	6.93	0.346	20	575	80		PCI-E 2.0 x16	poclbm	-w 32, don't use vectors
4550	7.23	0.289	25	600	80		PCI-E 2.0 x16	poclbm	-w 32, don't use vectors
4570M	8.02	0.297	25	680	80		PCI-E 2.0 x16	Diablo/Windows7	-w 64, SDK 2.4, Cat 11.4. Model: Sony Vaio NW
4570M	9.6	0.331	29	800	80		PCI-E 2.0 x16	Diablo/Windows7	-w 64, SDK 2.4, Cat 11.4. Model: Sony Vaio NW. O/C core 680->825. U/C memory 800->500.
4650	31.33	0.653	48	650	320		PCI-E 2.0 x16	poclbm	-w 32, don't use vectors
4670	36.14	0.613	59	750	320		PCI-E 2.0 x16	poclbm	-w 32, don't use vectors
4670	40.11		59	800	320		AGP x8	poclbm/Ubuntu 10.10 w/ H == 0 mod	-w 32 -f 0, don't use vectors source
4730	72.29	0.657	110	750	640		PCI-E 2.0 x16
4770	72.29	0.904	80	750	640		PCI-E 2.0 x16
4830	55.42	0.583	95	575	640		PCI-E 2.0 x16
4850	75.30	0.685	110	625	800		PCI-E 2.0 x16
4850	101			817	500	2.1	PCI-E 2.0 x16	guiminer	-f 0, core 817mhz(default 625)@1.123v , mem downclock to 500mhz (default 993) source
4860	67.47	0.519	130	700	640		PCI-E 2.0 x16
4870	90.36	0.602	150	750	800	2.1	PCI-E 2.0 x16	clmine
4870	78						PCI-E 2.0 x16	m0mchil's OpenCL/Vista 64bit	source
4890	102.41	0.539	190	850	800		PCI-E 2.0 x16
4890	108.3		190	975	800		PCI-E 2.0 x16
4850X2	150.60	0.602	250	625	1600		PCI-E 2.0 x16
4870X2	180.72	0.632	286	750	1600		PCI-E 2.0 x16
4870X2	180.6		286	800	1600		PCI-E 2.0 x16
5450	11.99	0.631	19	650	80		PCI-E 2.1 x16
5550	40.59	1.041	39	550	320		PCI-E 2.1 x16
5570	59.96	1.538	39	650	400		PCI-E 2.1 x16
5570	64	1.641	39	650			PCI-E 2.1 x16	m0mchil's OpenCL/WinXP	source
5650	48	1.37	35 ?				PCI-E 2.1 x16	m0mchil's OpenCL/Win7-64	source source -- not TDP but load
5670	71.49	1.117	64	775	400		PCI-E 2.1 x16
5670	72	1.64	44	850			PCI-E 2.0 x16	poclbm-mod (Win7-64)	Sapphire 100287VGAL card is low power
5670	85			900	400		PCI-E 2.1 x16	poclbm	-v -f 0 -w 128
5750	116.24	1.352	86	700	720		PCI-E 2.1 x16
5750	137			710			PCI-E 1.1 x16	poclbm	-v -f 0 -w 128 using stock memory and GPU clocks
5750	154.56	1.45	106	830	720		PCI-E 2.1 x16		-v -w128 -f20 underclocked memory clocks from 1150MHz to 300MHz
5750	170			870			PCI-E 1.1 x16	poclbm	-v -f 0 -w 128 patched BIOS to underclock memory to 300MHz
5750	173			875			PCI-E 1.1 x16	phoenix	875/300 w/MSI Afterburner, BFI_INT VECTORS AGGRESSION=12 FASTLOOP=FALSE -k phatk
5770	156.83	1.452	108	850	800	2.1	PCI-E 2.1 x16	clmine
5770	180	1.406	128[*]	950	800	2.4RC1	PCI-E 2.1 x16	poclbm	-v -w 128 -f 30 (*Overclocked wattage calculated here )
5770	205.58			935	300	2.1	PCI-E 2.1 x16	poclbm 2011-04-28	-v -w 256
5770	223		100+	1050	300	2.4	PCI-E 2.1 x16	guiminer	-v -w256 source
5830	206.64	1.181	175	800	1120		PCI-E 2.1 x16
5830	241	1.377	175	1006		2.3	PCI-E 2.1 x16	poclbm 2011-03-11 / Win7 64
5830	245	1.4	175	880	900	2.4	PCI-E 2.1 x16	poclbm (guiminer)	-v -w 256
5830	248	1.42	175	880	500	2.4	PCI-E 2.1 x16	phoenix 1.46	-k poclbm VECTORS BFI_INT AGGRESSION=11
5830	285			960		2.4	PCI-E 2.1 x16	phoenix 1.47 / poclbm / Win7 64	VECTORS BFI_INT FASTLOOP=false AGGRESSION=12
5830	290			996		2.4	PCI-E 2.1 x16	phoenix 1.4 / poclbm / Win7 64	VECTORS AGGRESSION=12 BFI_INT
5830	307			996		2.4	PCI-E 2.1 x16	phoenix 1.48 / phatk / Win7 64	VECTORS AGGRESSION=12 BFI_INT
5850	240.77	1.595	151	725	1440		PCI-E 2.1 x16
5850	250.26	1.657	151	725	1440		PCI-E 2.1 x16	opencl client	source
5850	252	1.575	160	765	1440	2.3	PCI-E 2.1 x16	poclbm 2011-01-25
5850	255.3	1.593	160	765	1440	2.2	PCI-E 2.1 x16	poclbm 2011-01-25
5850	264	1.748	151	725	1440	2.4	PCI-E 1.0 x8	guiminer-20110501	-v -w128
5850	280	1.75	160	765	1440	2.2	PCI-E 2.1 x16	Phoenix 1.3	with BFI_INT
5850	292	1.825	160	765	1440	2.1	PCI-E 2.1 x16	poclbm 2011-04-28 -v -f 1
5850	298	1.8620	160	765	1440	2.1	PCI-E 2.1 x16	poclbm 2011-04-28 -v -f 1 -w 64
5850	300	1.5460	194	925 (OC)	1440		PCI-E 2.1 x16	m0mchil's OpenCL	source
5850	314	1.8362	171	820 (OC)	1440	2.4	PCI-E 1.0 x8	poclbm	-v -w128
5850	335	1.8611	180	890 (OC)	1440	2.4	PCI-E 2.0 x8	poclbm	-v -w128, Memory downclocked to 300Mhz
5850	344	1.8594	185	890 (OC)	1440		PCI-E 2.0 x16	phoenix
5850	382	1.8454	207	995 (OC)	1440		PCI-E 2.0 x8	phoenix	OC 1.177v on core
5870M	116.5	?	?	?	?	2.1	PCI-E 2.1 x16	gui-miner (win-7)
5870	313.65	1.668	188	850	1600	2.1	PCI-E 2.1 x16	clmine
5870	313	1.665	188	900?	1600	2.3	PCI-E 2.1 x16	Diablo/Linux
5870	343	1.824	188	900?	1600	2.1	PCI-E 2.1 x16	Diablo/Linux
5870	355	1.888	188	900?	1600	2.1	PCI-E 2.1 x16	poclbm/Linux
5870	340	1.809	188	850	1600		PCI-E 2.1 x16	m0mchil's OpenCL	source
5870	360	1.6822	214	970	700		PCI-E 2.0 x16	m0mchil's OpenCL w7-64	-f 0 -v -w 128 source
5870	408	1.8888	216	980	1600	2.1	PCI-E 2.0 x16	poclbm/Win7x64	-f 20 -v -w 128, BIT_ALIGN, BFI_INT
5870	410	1.9523	210	950	1600	2.4	PCI-E 2.0 x16	Phoenix 1.48	-k phatk VECTORS BFI_INT AGGRESSION=11 WORKSIZE=128
5870	400	1.9047	210	950	1600	2.4	PCI-E 2.0 x8	Phoenix 1.3	-k poclbm VECTORS AGGRESSION=7 FASTLOOP BFI_INT
5870	421	1.9581	215	975	1600	2.4	PCI-E 2.0 x16	Phoenix 1.47	VECTORS BFI_INT AGGRESSION=13 -k phatk
5870	440	2.0000	220	995	1600	2.4	PCI-E 2.0 x16	Phoenix
5870	438	1.9819	221	1000 (OC)	1600	2.1	PCI-E 2.0 x16	poclbm/Linux	-v -w 64 -f 1
5970	530	1.803	294	725	3200	2.4.595.10	PCI-E 2.1 x16	poclbm/201103.beta3/Windows	-v -w128 -f60
5970	535.06	1.820	294	725	3200	2.1	PCI-E 2.1 x16	clmine
5970	560	1.905	294	725	3200		PCI-E 2.1 x16	Diablo
5970	565	1.922	294	725	3200	2.1	PCI-E 2.1 x16	clmine2
5970	604	2.054	294	725	3200	2.1	PCI-E 2.1 x16	clmine
5970	645	1.875	344	850	3200	2.1	PCI-E 2.1 x16	m0mchil/poclbm 03-07-11	-f1, Debian 6, fglrx-driver 10.9.3
5970	740	2.1511	344	850	3200	2.1	PCI-E 2.1 x16	phoenix/1.3	AGGRESSION=12 WORKSIZE=128 VECTORS BFI_INT, memory clock @ 300MHz
5970	755	2.2076	342	848	3200	2.1	PCI-E 2.1 x16	poclbm 28-04-11	-f 1 -w 256 -v, Ubuntu 10.10, fglrx 11.4, memory clock @ 300MHz
5970	802	2.31	347	850	3200	2.4	PCI-E 2.1 x16	phoenix poclbm Ubuntu 11.04	-q 6 -k AGGRESSION=19 WORKSIZE=128 BFI_INT VECTORS, fglrx 11.4, mem clock @ 1000MHz
5870x2	620			850	3200	2.1	PCI-E 2.1 x16	guiminer	-v -w 128
5870x2	826		1100	950	3200	2.1	PCI-E 2.1 x16	phoenix 1.4 w/ Phtak	AGGRESSION=13 WORKSIZE=128 VECTORS BFI_INT Memory @ 300MHz
5870x2	878		1500	1000	3200	2.1	PCI-E 2.1 x16	phoenix 1.7 w/ Phtak	AGGRESSION=13 WORKSIZE=128 VECTORS BFI_INT Memory @ 300MHz
6310M	9.821	0.545	18	500	80	2.4	Integrated/APU	poclbm-gui -v -w128	HP DM1z 18W TDP is shared with the CPU
6470M	24.1	?	?	?	?	2.1	Integrated/APU	guiminer
6490M	16.289	0.708	23				PCI-E 2.1 x16	poclbm-mod (Mac OS X 10.6.7)	MacBook Pro early 2011
6490M	15.21							DiabloMiner GUI v2011-05-22 / Mac OS X 10.6.7	Model: MacBookPro8,2 VRAM: 256MB
6750	167.59			870			PCI-E 2.1 x16	guiminer	-v -f30 -w128
6850	171.59	1.351	127	775	960	2.1	PCI-E 2.1 x16	clmine
6850	196			850	960		PCI-E 2.1 x16	poclbm	-v -w 128 -f 0
6850	234.8			940	960	2.4	PCI-E 2.1 x16	Phoenix 1.3	AGGRESSION=8 WORKSIZE=128 VECTORS BFI_INT
6850	220	1.236	178[*]	1000	960	2.4RC1	PCI-E 2.1 x16	poclbm	-v -w 128 -f 30 (*Overclocked wattage calculated here )
6870	232.47	1.540	151	900	1120		PCI-E 2.1 x16	poclbm -v -w 128
6870	277.47			940	1120		PCI-E 2.1 x16	Phoenix 1.3	AGGRESSION=8 WORKSIZE=128 VECTORS BFI_INT
6870	260.1	1.611	175	1001	1150	2.2	PCI-E 2.1 x16	poclbm-gui windows7x86 cat 11.3	-v -w 128 -f 1
6870	272	1.802 (?)	151 (?)	900	1120	2.4	PCI-E 2.1 x16	hashkill-0.2.5 alpha cat 11.4 sdk 2.4	-G2 -D stock clocks, debian linux 64bit source
6870	300.06	1.830	164	1020	1120	2.4	PCI-E 2.1 x16	Phoenix 1.4 cat 11.4	VECTORS BFI_INT AGGRESSION=8 fastloop mem clock 344
6870	302	1.726 (?)	175 (?)	1000	1120	2.4	PCI-E 2.1 x16	hashkill-0.2.5 alpha cat 11.4 sdk 2.4	-G2 -D overclocked to 1000, debian linux 64bit source
6870	300	1.72	174	1038	1120	2.4	PCI-E 2.1 x16	poclbm windows7x86 cat 11.4	-v -w128 -f1 mem clock 360, fan 100% temp 73C
6870	322	1.882	170	1050	1120	2.4	PCI-E 2.1 x16	phoenix 1.47 phatk 11.5 sdk 2.4	BFI_INT VECTORS FASTLOOP=false AGGRESSION=12 WORKSIZE=256, mem clock @ 400MHz
6950	295		160 (?)	810	1536	2.4.595.0	PCI-E 2.1 x16	m0mchil/poclbm 03-07-11	unlocked shaders, default mem 1250
6950	314		(?)	800	1408	2.4.595.0	PCI-E 2.1 x16	guiminer/poclbm 05-01-11	-v -f30 -w128 default mem 1250 Stock Sapphire 6950 1gb
6950	325		200 (?)	885	1536	2.4.595.0	PCI-E 2.1 x16	m0mchil/poclbm 03-07-11	unlocked shaders, default mem 1250
6950	340		(?)	800	1536	(?)	PCI-E 2.1 x16	guiminer/poclbm 05-21-11	-v -f30 -w128 default mem 1250,Sapphire 6950 1gb, unlocked shaders
6950	360		200 (?)	970	1536	2.4.595.0	PCI-E 2.1 x16	m0mchil/poclbm 03-07-11	unlocked shaders, default mem 1250
6950	360		(?)	870	1375	(?)	PCI-E 2.1 x16	guiminer/poclbm	flashed with 6970 BIOS, -v -f30 -w128
6950	417		+200?	975	1536	2.4.595.0	PCI-E 2.1 x16	guiminer/poclbm	unlocked shaders, +20% overdrive, mem @ 1250, fan @ 77%, temp 64C, -v -w128, gpu core @ 1145mV
6970	323	1.468	220	880	1536	2.3	PCI-E 2.1 x16	poclbm	-w 64, SDK 2.1 not supported on 69xx.
6970	372		220 (?)	900	1536	2.4	PCI-E 2.1 x16	phoenix (svn trunk 05.14.2011)	clock 900/1375, VECTORS BFI_INT AGGRESSION=11 WORKSIZE=128, fan 45% (auto) / temp 90C
6970	385	1.964	196	900	1536	2.4	PCI-E 2.1 x16	guiminer/poclbm	-v -w128, memory downclocked to 684Mhz, fan @ 70%/temp 81C
6970	403		(?)	MAX	1536	2.4	PCI-E 2.1 x16	guiminer/poclbm	clock Max/50%+stock, -v -w128, fan= 100%/temp 72C (Note: Overclocked using CCC utility under Windows 7, Clock set to Max value w/ 50% increase to memory and 5% increase to voltage)
6990	670	1.94	346	830	3072	2.4	PCI-E 2.1 x16	poclbm	Catalyst 11.4 source.
6990	708	2.05	346	830	3072	2.3	PCI-E 2.1 x16	mrb's hdminer	Catalyst 11.1 source.
6990	746	1.82	410	880	3072	2.3	PCI-E 2.1 x16	mrb's hdminer	source. SDK 2.1 not supported on 69xx.
6990	758			880		2.4		Phoenix with Poclbm	Catalyst 11.4, OC switch ON, VECTORS AGGRESSION=12 BFI_INT WORKSIZE=128 FASTLOOP=false
6990	771	1.8804	410	880	3072	2.3	PCI-E 2.1 x16	mrb's hdminer	SDK 2.1 not supported on 69xx. Memory clock @ 1280Mhz.
6990	772	1.8380	420	900	3072	2.3	PCI-E 2.1 x16	Phoenix & Poclbm	OC switch ON + overclocked to 900 mhz, VECTORS AGGRESSION=12 BFI_INT WORKSIZE=128
FirePro V8700	84.8			750	800			poclbm-mod.03.24.2011
FirePro M5800	61.4			650				poclbm-mod.03.24.2011	Ubuntu

Nvidia

Model	Mhash/s	Mhash/J	W	Clock	SP	Comment
8200 mGPU	1.2			1200	16	128 MB shared memory, "poclbm -w 128 -f 0"
8400 GS	2.3					"poclbm -w 128"
8400 GS	1.6		128	1238		DiabloMiner -w 128 -f 1 source
8600M GT	4.93				32
8600GT	5.66			1188	32
8600GT OC	7.3			1602	32	poclbm -w 128 source
8800GT	25	0.24	105	1300
8800GT	24.5	0.23	105	1300		source
8800GT	31.1		105	1855		Overclocked 715 GPU / 1030 Memory / 1855 Shader; GuiMiner v2011-05-21
8800GTS	16.8	0.109	154			source source
8800 GTS	18.7	0.124	150	1200		poclbm -w 64 no vectors
9300GE	1.57			1300	8
9300GS	1.69			1400	8
9300/nForce 730i	2.15			1200	16
9400GT	3.37	0.067	50	1400	16
9400M (MacBook)	1.90	0.32	6	700		poclbm -f 8 no vectors
9500GT	6.75	0.135	50	1400	32
9600GSO	19.88	0.237	84	1375	96
9600GSO512	11.75	0.131	90	1625	48
9600GT	15.66	0.165	95	1625	64
9600GT Zotac	15			1650	64
9600GT OC	18.8	<0.198	>95	1981	64	poclbm -w 128 -f 10 source source
9800GT	30.36	0.289	105	1800	112
9800GTX	32.54	0.232	140	1688	128
9800GTX+	35.39	0.251	141	1836	128
9800GTX+	37.23	0.266	140	1890	128	win7x64 275.27 phoenix(1.48) -k poclbm AGGRESSION=3 WORKSIZE=64
9800GX2	57.83	0.294	197		2x128
9800GX2	28	0.142	197		2x128	source
G210	3.38	0.111	30.5	1402	16
G210	3.79	0.124	30.5	1402	16	poclbm -f 1
GT220	10.8		128	1360		DiabloMiner -w 128 -f 1000 source
GT240	19.37	0.281	69	1340	96
GT240	21.24				96	poclbm-mod -f 0 -v source
GTS250	35.39	0.244	145	1836	128
GTS250	35.2	0.243	145	1836	128
GTS250 OC	37	0.255	145	2047	128
GTX260	35.91	0.178	202	1242	192
GTX260	44	0.242	182	1242	216	source
GTX260c216	40.40	0.236	171	1242	216
GTX260c216	50.00			1050	216	m0mchil GPU client, Windows 7 64-bit, x6 @ 3.5ghz
GTX260c216 OC	52.0			1461	216	"poclbm -w 256 -f 1"
GTX260c216 OC	58.9			756/400	216	Core overclocked, Memory underclocked, Ubuntu 11.04 Classic Session(No Effects) "poclbm -w 256 -f 1"
GTX275	50.75	0.232	219	1404	240
GTX275	58			729/1458	240	poclbm -f 0 -w 256
GTX280	46.84	0.198	236	1296	240
GTX285	53.35	0.262	204	1476	240
GTX295	89.78	0.311	289	1242	480
GT 320M (MacBook Air)	6.12			1212	48
GT 330M	7.97					Model: MacBookPro6,2 VRAM: 512MB Software: DiabloMiner GUI v2011-05-22 / Mac OS X 10.6.7
GT 330M (Sony Vaio Z)	7.8	0.71 ( 0.3 total)	11 (26w total)	1045	48
GT430	20.24	0.413	49	1400	96
GTS450	45.28	0.427	106	1566	192
GTX460SE	56.39	0.376	150	1300	288
GTX460	68.31	0.427	160	1350	336
GTX460 (2 cards)	102	0.319?	320?	1350		source
GTX460 (2 cards) OC	127	0.374	340	1620	2x 336	rpcminer-cuda -gpugrid=128 -gputhreads=128 ver.20110227
GTX465	64.41	0.322	200	1215	352
GTX470	81.98	0.381	215	1215	448
GTX470	94.7			1414
GTX470	111.9		215	1650	448
GTX470	115			1616		phoenix 1.46 poclbm VECTORS BFI_INT FASTLOOP AGGRESSION=6
GTX480	101.28	0.405	250	1401	480
GTX560 Ti	67.7	0.39	170	1700	384	standard EVGA 560, no overclock
GTX560 Ti	74.8	0.41	180	1700	384	gigabyte 900Mhz clock; -w 512 -v
GTX560 OC	86.7	<0.51	>170	1800	384	rpcminer-cuda source
GTX570	105.83	0.483	219	1464	480
GTX580	140.05	---	244	1544	512
Quadro FX 580	5.7	0.14	40	1125	4	rpcminer-cuda, Win, -gpugrid=128 -gputhreads=512
Quadro FX 770M	5.75			500	32	DiabloMiner-Windows.exe BFI_INT FASTLOOP AGGRESSION=6, Win
Quadro FX 1600M	6	0.12	50	625	32	rpcminer-cuda, Win
Quadro NVS 135M	1.05	0.1	10	800	1
Quadro NVS 3100M	3.6	0.257	14	600	16	rpcminer-cuda, Win, CUDA 3.1.1
Tesla M2050	79.8					DiabloMiner
Tesla M2050	100	0	0	-	512	Amazon AWS CUDA RPC Miner & CUDA 2.0

CPUs/APUs

A lot of nice data can be pulled from this thread to seed this section. Also from this page on the old wiki.

Might you wish to throttle CPU load a bit, so it doesn't reach 100⁰C, it can be done with Battle Encoder Shirase to a margin, like, -15% of available CPU cycles.

I found the best combination to be is to throttle CPU miner to -5% _AND_ set it to least of priority levels, so as not to choke GPU miner up.

AMD

Model	nprocs	Mhash/s	Mhash/J	TDP [W]	CPU Clock	Mhash/s CPU	Mhash/s GPU	GPU	GPU Clock	SP	Software	Notes
4x Opteron 6174	48	115	0.36	320W	2.2 GHz	2.4					cpuminer v0.8.1	--alg 4way --threads 48
2x Opteron 6128	16	32.4	0.141	230W	2 GHz	32.4					0.3.19	-4way
Athlon XP 2000+	2	0.62	0.009	70W	1.67 GHz	0.62					0.3.18/Ubuntu	source source
Athlon 64 3500+	1	1.18	0.013	89W	2.54 GHz	1.18					ufasoft v0.4	overclocked (originally 2.2 GHz)
Athlon 64 X2 3800+	2	1.73	0.03	65 W	2.00 GHz	1.73					cpuminer (v0.8.1-1-g69529c3)	-algo=4way
Athlon 64 X2 4000+	2	1.9	0.02	65W	2.1 GHz	1.9					rpc-miner
Athlon 64 X2 4400+		2.09	0.032	65W	2.3GHz	2.09					0.3.19/Win x64	source
Athlon 64 X2 6000+	2	2.81	0.02	125W	3 GHz	2.81						source
Athlon 64 X2 6400+ Black Edition	2	2.9	0.023	125W	3.2 GHz	2.9					0.3.20.2 BETA/Win 7 x64	-4way
Athlon II X2 240e	2	2.71	0.06	45W	2.81 GHz	2.71					bitcoind	source
Athlon II X4 630	4	10.7	0.11	95W	2.8 GHz	10.7					bitcoin-miner 0.4
Phenom II X3 720	3	3.8	0.04	95W	2.8 GHz	3.8					0.3.1x/WinXP	source
Phenom II X3 720	3	7.2	0.08	95W	2.8 GHz	7.2					cpu-miner 0.2.1/WinXP	source
Phenom II X4 955	4	11	0.09	125W	3.2 GHz	11					rpcminer-4way
Phenom II X4 965	4	12	0.09	140W	3.4 GHz	11					rpcminer-4way
Phenom II X6 1055T	6	15.84	0.13	125W	2.82 GHz	15.84					bitcoind	source
Phenom II X6 1090T	6	18		141W	3.50 GHz	3					minerd
Phenom II X6 1100T	6	22	0.176	125W	3.82 GHz	22					bitcoin-miner	Aciid#bitcoin-dev
Zacate E-350	2	11	0.615	18W	1.6 Ghz	1.231	9.831	HD6310M	492 Mhz	80	poclbm -v -w128 + Ufasoft 0.7	stock speeds
Zacate E-350	2	12			1.648 Ghz	1.252	10.87	HD6310M	492 Mhz	80	poclbm -v -w 256 -f 1	"autotune" speed with ASUS EFI BIOS

ARM

Model	p/t	Mhash/s	Mhash/J	ACP [W]	Clock	Version	Comment
ARM1136EJ-S	1	0.11	?	?	0.528 GHz	cpuminer git	T-Mobile Pulse, inside Debian chroot, 'c' algo
Cortex-A8	1	0.125	0.36	0.35 W	0.6 GHz	cpuminer git (2011-03-26)	Nokia N900: 'cryptopp'
Cortex-A8	1	0.2	0.57	0.35 W	0.6 GHz	cpuminer git (2011-03-26)	Nokia N900: 'c' algo
2x Cortex-A9	2	0.57	1.14	0.5 W	1.0 GHz	cpuminer git (2011-03-24)	Toshiba AC100, ubuntu, 'c' algo

Intel

Model	p/t	Mhash/s	Mhash/J	ACP [W]	Clock	Version	Comment
Pentium III mobile ?	1	0.3	0.014	21W	1.07 GHz	0.3.1x/Win2K	source source
Pentium M 1.6ghz	1	.4			1.6 GHZ	minerd (jgarzik's 1.0)	cryppto, Windows 7
Pentium M 1.73ghz	1	.5			1.73 GHZ	minerd (jgarzik's 1.0)	cryppto, Windows 7
Old Xeon 512k (Dual)	2x1/2	2.0	?	?	3.06GHz	cpuminer (v0.8.1-1-g69529c3)	HT disabled, algo=4way (twice as fast as the 2nd best algo)
Pentium 4 2.0A	1	0.85			2.0 GHz	ufasoft-0.4/WinXP	-g no -t 2
Pentium Dual-Core E5400	2/2	2.27	0.03	65W	2.7 GHz	bitcoind	source
Celeron E330	2/2	2.2	0.03	65W	2.5 GHz	0.3.19/Ubuntu10.04	source
Core 2 Quad Q6600	4/4	11.0		105W	2.40 GHz	bitcoin-miner 0.10	-a 60 -g no -t 4
Core 2 Quad Q9400	4/4	11	0.046	95W	2.66 GHz	bitcoin-miner 0.10	-a 60 -g no -t 4
Core 2 Quad Q9650	4/4	18.67		95W	4.00 GHz	bitcoin-miner 0.10	-a 60 -g no -t 4
Core i3 530	2/4	8.31	0.10	80W	3.66 GHz	Ufasoft 0.7	-v -a5 -g no -t 4
Core i3 M350	2/4	1.48	0.04	35W	2.27 GHz	bitcoind	source
Core i5 M450	2/4	1.8	0.05	35W	1.2 GHz	0.3.17/Win7-54	source
Core i5-650	2/4	5.1	0.04 ?		3.2 GHz	cpuminer-0.7	-4way
Core i5 ?	4/?	6.5				client from svn	source
Core i5-2400	4/4	4.5	0.05	95W	3.1 GHz	cpuminer git (2011-01-22)	cryptopp_asm32
Core i5-2400	4/4	14	0.15	95W	3.1 GHz	cpuminer git (2011-03-26)	sse2_64
Core i7 620M	2/4	1.888			2.66 GHz	RPCMiner GUI v2011-05-22 / Mac OS X 10.6.7	Model: MacBookPro6,2 RAM: 4GB
Core i7 920	4/8	19.2	0.10	195	4.0 GHz (x21)	ufasoft	-a 5
Core i7 950	4/8	5.88	0.039	150W	3.83 GHz (x23)	bitcoin-0.3.20.2 Win7-64
Core i7 950	4/8	18.9	0.126	150W	3.83 GHz (x23)	ufasoft v0.4
Core i7 980x	6/12	19.2	0.15	130	4.4 GHz (x33)	cpuminer/Win7-64
Core i7 980x	6/12	8.7			3.9 GHz (x27)	0.3.17/Win7-64
Core i7 2635QM	4/8	2.93			2.00 GHz	RPCMiner GUI v2011-05-22 / Mac OS X 10.6.7	Model: MacBookPro8,2 RAM: 4GB
Core i7 620M	2/4	6.3	0.18	35	2.66 GHz	ufasoft v0.4
Core 2 Duo U7600	1	1.1			1.2 GHZ	minerd (jgarzik's 1.0)	cryppto, Windows 7
Core 2 Duo E5200	2/2	6.2	0.086	72W	2.76 Ghz	Ufasoft/Win7-64	TDP is 65W + 7W overclock
Core 2 Duo E6550	1/2	2.45	?	?	2.33 GHz	cpuminer 0.7.1 (Linux)	--algo=sse2_64
Core 2 Duo E6850	2/2	6.75	0.10	65W	3.0 Ghz	ufasoft-0.3
Core 2 Duo E7300	2/2	7.76	0.11	70W	3.33 GHz (o/c?)	ufasoft-0.3	miner optimized for Intel Core
Core 2 Duo E7300	2/2	2.52	0.04	65W	2.66 GHz	bitcoind	source
Core 2 Duo E8400	2/2	6.9	?	?	3.0 GHz	ufasoft v0.10	windows 7/64bit (-t 2)
Core 2 Duo T6400	2/2	4.2	0.12	35 W	2.00 GHz	cpuminer git (Linux)	--algo=sse2_64
Xeon 3060	1/2	2.03	0.03	65W	2.4GHz	cpuminer 0.8.1 (Linux)	--algo=cryptopp_asm32
Xeon Prestonia 2.4 (dual)	2x1/2	2.16	0.017	130W	2.4 GHz	ufasoft v0.4
Xeon X5355 (dual)	2x4/4	10.13	0.16	120W	2.66GHz	bitcoind	Roughly the same speed as the "c" algo in cpuminer
Xeon X5355 (dual)	2x4/4	22.76	0.09	120W	2.66GHz	cpuminer (v0.8.1-1-g69529c3)	-O2 -march=core2, algo=sse2_64
Xeon E5440	4/8	7.3	?	80W	2.66 GHz	Kiv's poclbm-gui	FIXME: Either wrong model # or wrong threads/speed info
Xeon E5520	4/8	6.5	0.08	80W	2.27 GHz	bitcoind	source
Xeon E5530	4/8	7.14	0.09	80W	2.4 GHz	bitcoind	source
Xeon E5630 (dual)	2x4/8	8	0.1	80W	2.53 GHz	0.3.17/Win7-64	source
Xeon E6520 (dual)	2x4/8	24.7	?	?	2.53 GHz	ufasoft v0.10	windows 2008 R2 64bit (-t 16)
Xeon E7520 (dual)	2x4/16	18	?	95W	1.87 GHz	ufasoft v0.10	windows 2008 R2 64bit (-t 16)
Atom N270	2	1.19	0.24	5W	1.6 GHz	ufasoft v0.10

Other

Model	p/t	Mhash/s	Mhash/J	ACP [W]	Clock	Version	Comment
Cell	6	21	0.35	60W	3.2 GHz	cellminer	Sony Playstation 3 (FAT): Needs custom firmware (no links to it, potential issues with Sony assholes.) source source Total power usage of PS3 fat is 120W but only the Cell cpu is active.
Cell	6	21	0.7	35W	3.2 GHz	cellminer	Sony Playstation 3 (SLIM): Total power usage of PS3 slim is 70W but only the Cell cpu is active.

See Also

• Mining rig
• Generation Calculator
• Power Calc
• OpenCL miner
• ATI Stream vs. NVIDIA CUDA - GPGPU computing battle royale