Benchmarks

[davidBAM]

[davidBAM]

Not the fastest CPU so almost certainly explains the drop from the table in the post following. Actual power consumption is well below 130W per card though - fluctuates but maybe about 15% lower

GenuineIntel Intel(R) Xeon(R) CPU E5-2609 v3 @ 1.90GHz [Family 6 Model 63 Stepping 2] (6 processors) - one cpu per gpu

Run Time	CPU Time	Credit	Per day/proc	Credit per kWH	Application
497.97	466.90	6,144	1,066,011	341,670	Distributed.net Client v1.05 (opencl_nvidia_101)
501.42	457.11	6,144	1,058,677	339,319	Distributed.net Client v1.05 (opencl_nvidia_101)
503.98	462.58	6,144	1,053,299	337,596	Distributed.net Client v1.05 (opencl_nvidia_101)
502.90	466.05	6,144	1,055,561	338,321	Distributed.net Client v1.05 (opencl_nvidia_101)
495.39	458.34	6,144	1,071,563	343,450	Distributed.net Client v1.05 (opencl_nvidia_101)
507.61	466.25	6,144	1,045,767	335,182	Distributed.net Client v1.05 (opencl_nvidia_101)
498.16	466.32	6,144	1,065,605	341,540	Distributed.net Client v1.05 (opencl_nvidia_101)
510.32	461.55	6,144	1,040,213	333,402	Distributed.net Client v1.05 (opencl_nvidia_101)
505.45	466.75	6,144	1,050,236	336,614	Distributed.net Client v1.05 (opencl_nvidia_101)
504.73	474.44	6,144	1,051,734	337,094	Distributed.net Client v1.05 (opencl_nvidia_101)
521.69	493.01	6,512	1,078,489	345,669	Distributed.net Client v1.05 (opencl_nvidia_101)
501.65	464.12	6,144	1,058,191	339,164	Distributed.net Client v1.05 (opencl_nvidia_101)
498.03	460.27	6,144	1,065,883	341,629	Distributed.net Client v1.05 (opencl_nvidia_101)
499.55	469.05	6,144	1,062,640	340,590	Distributed.net Client v1.05 (opencl_nvidia_101)
508.11	472.27	6,144	1,044,738	334,852	Distributed.net Client v1.05 (opencl_nvidia_101)
509.16	465.56	6,144	1,042,583	334,161	Distributed.net Client v1.05 (opencl_nvidia_101)
501.35	466.57	6,144	1,058,824	339,367	Distributed.net Client v1.05 (opencl_nvidia_101)
507.27	468.19	6,144	1,046,468	335,406	Distributed.net Client v1.05 (opencl_nvidia_101)
491.19	466.54	6,144	1,080,726	346,386	Distributed.net Client v1.05 (opencl_nvidia_101)
499.13	464.32	6,144	1,063,534	340,876	Distributed.net Client v1.05 (opencl_nvidia_101)
503.25	466.81	6,162	1,058,037	339,114	Averages for nVidia GTX1660 Ti, 130W, 1 WU per proc

[davidBAM]

Not sure exactly which Xeon this is ... Genuine Intel(R) CPU @ 2.30GHz [Family 6 Model 62 Stepping 2] (24 processors)

Run Time	CPU Time	Credit	Per day/proc	Credit per kWH	Application
509.65	494.73	6,584	1,116,173	357,748	Distributed.net Client v1.05 (opencl_nvidia_101)
510.07	493.59	6,584	1,115,254	357,453	Distributed.net Client v1.05 (opencl_nvidia_101)
508.18	494.00	6,576	1,118,042	358,347	Distributed.net Client v1.05 (opencl_nvidia_101)
510.10	493.66	6,584	1,115,188	357,432	Distributed.net Client v1.05 (opencl_nvidia_101)
507.26	494.39	6,584	1,121,432	359,433	Distributed.net Client v1.05 (opencl_nvidia_101)
509.08	494.70	6,584	1,117,423	358,148	Distributed.net Client v1.05 (opencl_nvidia_101)
510.66	493.62	6,584	1,113,965	357,040	Distributed.net Client v1.05 (opencl_nvidia_101)
508.64	493.98	6,584	1,118,389	358,458	Distributed.net Client v1.05 (opencl_nvidia_101)
512.61	492.72	6,584	1,109,728	355,682	Distributed.net Client v1.05 (opencl_nvidia_101)
488.98	474.86	6,304	1,113,881	357,013	Distributed.net Client v1.05 (opencl_nvidia_101)
510.17	492.35	6,576	1,113,681	356,949	Distributed.net Client v1.05 (opencl_nvidia_101)
510.11	493.59	6,584	1,115,167	357,425	Distributed.net Client v1.05 (opencl_nvidia_101)
507.07	484.83	6,544	1,115,037	357,384	Distributed.net Client v1.05 (opencl_nvidia_101)
508.84	494.80	6,584	1,117,950	358,317	Distributed.net Client v1.05 (opencl_nvidia_101)
510.09	493.73	6,584	1,115,210	357,439	Distributed.net Client v1.05 (opencl_nvidia_101)
508.08	494.96	6,584	1,119,622	358,853	Distributed.net Client v1.05 (opencl_nvidia_101)
476.92	455.00	6,144	1,113,062	356,751	Distributed.net Client v1.05 (opencl_nvidia_101)
474.99	456.18	6,144	1,117,585	358,200	Distributed.net Client v1.05 (opencl_nvidia_101)
476.93	455.03	6,144	1,113,039	356,743	Distributed.net Client v1.05 (opencl_nvidia_101)
509.11	493.65	6,584	1,117,357	358,127	Distributed.net Client v1.05 (opencl_nvidia_101)
503.38	486.72	6,501	1,115,859	357,647	Averages for nVidia GTX1660 Ti, 130W, 1 WU per proc

[davidBAM]

GenuineIntel
Intel(R) Xeon(R) CPU E5-2690 v2 @ 3.00GHz [Family 6 Model 62 Stepping 4]
(20 processors)

Run Time	CPU Time	Credit	Per day/proc	Credit per kWH	Application
392.82	374.71	6,584	1,448,138	241,356	Distributed.net Client v1.05 (opencl_nvidia_101)
364.73	347.83	6,144	1,455,437	242,573	Distributed.net Client v1.05 (opencl_nvidia_101)
390.86	375.43	6,600	1,458,937	243,156	Distributed.net Client v1.05 (opencl_nvidia_101)
365.80	345.67	6,144	1,451,180	241,863	Distributed.net Client v1.05 (opencl_nvidia_101)
389.99	375.74	6,584	1,458,647	243,108	Distributed.net Client v1.05 (opencl_nvidia_101)
392.88	372.85	6,584	1,447,917	241,319	Distributed.net Client v1.05 (opencl_nvidia_101)
390.20	375.23	6,584	1,457,862	242,977	Distributed.net Client v1.05 (opencl_nvidia_101)
389.85	375.75	6,584	1,459,170	243,195	Distributed.net Client v1.05 (opencl_nvidia_101)
387.88	367.76	6,584	1,466,581	244,430	Distributed.net Client v1.05 (opencl_nvidia_101)
365.72	346.05	6,144	1,451,497	241,916	Distributed.net Client v1.05 (opencl_nvidia_101)
390.87	376.63	6,584	1,455,363	242,560	Distributed.net Client v1.05 (opencl_nvidia_101)
389.95	374.53	6,584	1,458,796	243,133	Distributed.net Client v1.05 (opencl_nvidia_101)
361.80	349.60	6,144	1,467,224	244,537	Distributed.net Client v1.05 (opencl_nvidia_101)
365.81	346.02	6,144	1,451,140	241,857	Distributed.net Client v1.05 (opencl_nvidia_101)
366.83	345.84	6,144	1,447,105	241,184	Distributed.net Client v1.05 (opencl_nvidia_101)
364.77	347.09	6,144	1,455,278	242,546	Distributed.net Client v1.05 (opencl_nvidia_101)
392.04	373.72	6,584	1,451,019	241,837	Distributed.net Client v1.05 (opencl_nvidia_101)
364.91	347.95	6,144	1,454,719	242,453	Distributed.net Client v1.05 (opencl_nvidia_101)
390.09	374.53	6,584	1,458,273	243,045	Distributed.net Client v1.05 (opencl_nvidia_101)
364.95	347.22	6,144	1,454,560	242,427	Distributed.net Client v1.05 (opencl_nvidia_101)
379.14	362.01	6,387	1,455,442	242,574	Averages for nVidia GTX1080 Ti GPU, 250W, 1 WU per proc

[davidBAM]

I say 'double duty' as this was from my main workstation so was running X-windows all day while I worked. I can't say I really noticed any significant degradation in use - it would even play BBC videos acceptably. Normally, I have to suspend any GPU project while I am working. Note that Moo refused to let me progress beyond cuda60 on this machine (perhaps due to driver version 390.11?). Might need to try it with a gtx1660ti as a 'permanent Moo machine

GenuineIntel Intel(R) Xeon(R) CPU E5-2696 v4 @ 2.20GHz [Family 6 Model 79 Stepping 1] (44 processors)

Run Time	CPU Time	Credit	Per day/proc	Credit per kWH	Application
503.31	89.22	6,584	1,130,233	188,372	Distributed.net Client v1.04 (cuda60)
498.11	86.40	6,584	1,142,032	190,339	Distributed.net Client v1.04 (cuda60)
505.14	88.97	6,592	1,127,507	187,918	Distributed.net Client v1.04 (cuda60)
501.38	88.71	6,584	1,134,584	189,097	Distributed.net Client v1.04 (cuda60)
470.99	84.09	6,144	1,127,076	187,846	Distributed.net Client v1.04 (cuda60)
475.14	86.83	6,144	1,117,232	186,205	Distributed.net Client v1.04 (cuda60)
477.89	87.57	6,144	1,110,803	185,134	Distributed.net Client v1.04 (cuda60)
462.93	81.06	6,144	1,146,700	191,117	Distributed.net Client v1.04 (cuda60)
473.77	83.19	6,144	1,120,463	186,744	Distributed.net Client v1.04 (cuda60)
466.99	81.44	6,144	1,136,730	189,455	Distributed.net Client v1.04 (cuda60)
493.06	89.14	6,400	1,121,486	186,914	Distributed.net Client v1.04 (cuda60)
492.18	85.87	6,520	1,144,557	190,759	Distributed.net Client v1.04 (cuda60)
482.59	92.50	6,144	1,099,985	183,331	Distributed.net Client v1.04 (cuda60)
474.59	85.00	6,144	1,118,527	186,421	Distributed.net Client v1.04 (cuda60)
475.45	85.88	6,144	1,116,504	186,084	Distributed.net Client v1.04 (cuda60)
467.08	82.11	6,144	1,136,511	189,419	Distributed.net Client v1.04 (cuda60)
474.20	84.62	6,144	1,119,447	186,574	Distributed.net Client v1.04 (cuda60)
466.28	79.93	6,144	1,138,461	189,744	Distributed.net Client v1.04 (cuda60)
476.75	86.92	6,144	1,113,459	185,577	Distributed.net Client v1.04 (cuda60)
469.01	82.04	6,144	1,131,834	188,639	Distributed.net Client v1.04 (cuda60)
480.34	85.57	6,264	1,126,706	187,784	Averages for nVidia GTX1080 Ti GPU, 250W, 1 WU per proc

[VietOZ]

from vietOz (thank you)

Pentium G4560

Run Time	CPU Time	Credit	Per day/proc	Credit per kWH	Application
496.48	443.20	6,144	1,069,210	342,696	Distributed.net Client v1.05 (opencl_nvidia_101)
543.62	479.78	6,512	1,034,982	331,725	Distributed.net Client v1.05 (opencl_nvidia_101)
527.33	465.06	6,368	1,043,360	334,410	Distributed.net Client v1.05 (opencl_nvidia_101)
506.69	450.64	6,144	1,047,665	335,790	Distributed.net Client v1.05 (opencl_nvidia_101)
509.82	448.50	6,144	1,041,233	333,729	Distributed.net Client v1.05 (opencl_nvidia_101)
497.66	437.78	6,144	1,066,675	341,883	Distributed.net Client v1.05 (opencl_nvidia_101)
505.65	450.58	6,144	1,049,820	336,481	Distributed.net Client v1.05 (opencl_nvidia_101)
504.74	450.48	6,144	1,051,713	337,087	Distributed.net Client v1.05 (opencl_nvidia_101)
508.79	450.55	6,144	1,043,341	334,404	Distributed.net Client v1.05 (opencl_nvidia_101)
505.76	450.33	6,144	1,049,592	336,408	Distributed.net Client v1.05 (opencl_nvidia_101)
513.83	458.80	6,144	1,033,107	331,124	Distributed.net Client v1.05 (opencl_nvidia_101)
503.77	449.81	6,144	1,053,738	337,737	Distributed.net Client v1.05 (opencl_nvidia_101)
514.89	459.06	6,144	1,030,981	330,442	Distributed.net Client v1.05 (opencl_nvidia_101)
513.93	461.63	6,144	1,032,906	331,060	Distributed.net Client v1.05 (opencl_nvidia_101)
551.81	494.13	6,616	1,035,904	332,021	Distributed.net Client v1.05 (opencl_nvidia_101)
527.20	464.38	6,288	1,030,507	330,291	Distributed.net Client v1.05 (opencl_nvidia_101)
495.45	440.75	6,144	1,071,433	343,408	Distributed.net Client v1.05 (opencl_nvidia_101)
505.60	451.25	6,144	1,049,924	336,514	Distributed.net Client v1.05 (opencl_nvidia_101)
510.93	459.16	6,144	1,038,971	333,004	Distributed.net Client v1.05 (opencl_nvidia_101)
533.14	476.34	6,400	1,037,176	332,428	Distributed.net Client v1.05 (opencl_nvidia_101)
513.85	457.11	6,217	1,045,612	335,132	Averages for nVidia GTX1660 Ti, 130W, 1 WU per proc

[VietOZ]

these results were with a power limit reduction of 10% on this rig (15% on the other one). So the power consumption is around 118W per card, measured with a kill-a-watt at the wall. This is the 4 cards setup built on a pentium G4650 (4 threads) using mining cable risers (pcie 1x). As you can see it doesn't matter much with pcie 16x or 1x on this project. However, Moo does require a full thread of cpu. Windows used some cpu resources, so I don't really have a full 4 threads for 4 cards. If I have a 6700k or 7700k (8 threads cpu) then the run time should be a bit better.

[davidBAM]

Thanks

I certainly agree about the heavy CPU usage on Moo - I was seeing load averages of up to 50-60% higher than expected (frequently over 9 on a 6 core / 6 thread processor - ordinarily I'd expect that to be a little over 6, 7 at most)

Did you notice the very LOW cpu usage on https://tsbt.co.uk/forum/viewtopic.php? ... 425#p72420 though

[VietOZ]

cpu: Ryzen 5 2400G @3.8ghz Win 10
Radeon settings:
voltage: 1010mV
Clock: 1750
Power limit: default
Mem clock: default
Fan speed: 100%
Power consumption: 191W

Run Time	CPU Time	Credit	Per day/proc	Credit per kWH	Application
340.66	21.80	6,584	1,669,869	364,282	Distributed.net Client v1.05 (opencl_ati_101)
340.81	19.66	6,584	1,669,134	364,122	Distributed.net Client v1.05 (opencl_ati_101)
339.83	22.06	6,584	1,673,948	365,172	Distributed.net Client v1.05 (opencl_ati_101)
352.03	21.17	6,584	1,615,935	352,516	Distributed.net Client v1.05 (opencl_ati_101)
352.00	20.61	6,584	1,616,073	352,546	Distributed.net Client v1.05 (opencl_ati_101)
339.82	22.02	6,584	1,673,997	365,183	Distributed.net Client v1.05 (opencl_ati_101)
340.88	20.86	6,584	1,668,791	364,047	Distributed.net Client v1.05 (opencl_ati_101)
328.94	20.44	6,144	1,613,795	352,049	Distributed.net Client v1.05 (opencl_ati_101)
317.70	20.22	6,144	1,670,890	364,505	Distributed.net Client v1.05 (opencl_ati_101)
339.92	21.94	6,584	1,673,504	365,075	Distributed.net Client v1.05 (opencl_ati_101)
322.63	19.59	6,144	1,645,357	358,935	Distributed.net Client v1.05 (opencl_ati_101)
339.75	19.83	6,584	1,674,342	365,258	Distributed.net Client v1.05 (opencl_ati_101)
318.66	19.63	6,144	1,665,856	363,407	Distributed.net Client v1.05 (opencl_ati_101)
337.92	21.36	6,568	1,679,318	366,343	Distributed.net Client v1.05 (opencl_ati_101)
340.86	20.97	6,584	1,668,889	364,068	Distributed.net Client v1.05 (opencl_ati_101)
341.86	21.09	6,584	1,664,007	363,003	Distributed.net Client v1.05 (opencl_ati_101)
341.97	21.84	6,640	1,677,621	365,973	Distributed.net Client v1.05 (opencl_ati_101)
337.80	20.66	6,512	1,665,591	363,349	Distributed.net Client v1.05 (opencl_ati_101)
337.84	20.13	6,512	1,665,394	363,306	Distributed.net Client v1.05 (opencl_ati_101)
327.97	18.70	6,328	1,667,040	363,665	Distributed.net Client v1.05 (opencl_ati_101)
336.99	20.73	6,478	1,660,968	362,340	Averages for AMD Radeon7 GPU, 191W, 1 WU per proc

[davidBAM]

Good results, thanks. I'd say the points per watt are about the same as the 1660ti as it certainly doesn't run the full 130W the whole time. Now I want a points per $ column but that is a bit too tricky

Vastly lower CPU times - you are going to have to swap over the cards and do it all again I have a feeling that a good CPU is a must no matter what GPU you are using for Moo

[VietOZ]

Moo use very little CPU time with AMD cards. The same Pentium G4650 4 threads can run 6x RX 580 with no trouble, but have a really hard time to run 4x 1660ti

[davidBAM]

from bluestang
i7-2600k running at 4Ghz.

Run Time	CPU Time	Credit	Per day/proc	Credit per kWH	Application
508.36	479.75	6,584	1,119,005	358,656	Distributed.net Client v1.05 (opencl_nvidia_101)
505.04	480.73	6,584	1,126,361	361,013	Distributed.net Client v1.05 (opencl_nvidia_101)
499.07	480.89	6,584	1,139,835	365,332	Distributed.net Client v1.05 (opencl_nvidia_101)
493.26	476.27	6,584	1,153,261	369,635	Distributed.net Client v1.05 (opencl_nvidia_101)
498.36	473.58	6,552	1,135,911	364,074	Distributed.net Client v1.05 (opencl_nvidia_101)
508.45	480.42	6,584	1,118,807	358,592	Distributed.net Client v1.05 (opencl_nvidia_101)
513.00	491.83	6,584	1,108,884	355,412	Distributed.net Client v1.05 (opencl_nvidia_101)
501.65	483.95	6,584	1,133,973	363,453	Distributed.net Client v1.05 (opencl_nvidia_101)
505.03	482.89	6,584	1,126,384	361,020	Distributed.net Client v1.05 (opencl_nvidia_101)
467.34	443.66	6,144	1,135,879	364,064	Distributed.net Client v1.05 (opencl_nvidia_101)
494.04	478.41	6,584	1,151,440	369,051	Distributed.net Client v1.05 (opencl_nvidia_101)
459.45	449.42	6,144	1,155,385	370,316	Distributed.net Client v1.05 (opencl_nvidia_101)
460.12	446.50	6,144	1,153,703	369,776	Distributed.net Client v1.05 (opencl_nvidia_101)
461.45	447.17	6,144	1,150,377	368,711	Distributed.net Client v1.05 (opencl_nvidia_101)
500.04	472.30	6,584	1,137,624	364,623	Distributed.net Client v1.05 (opencl_nvidia_101)
472.65	457.50	6,144	1,123,118	359,974	Distributed.net Client v1.05 (opencl_nvidia_101)
505.52	487.41	6,584	1,125,292	360,671	Distributed.net Client v1.05 (opencl_nvidia_101)
461.17	447.80	6,144	1,151,076	368,935	Distributed.net Client v1.05 (opencl_nvidia_101)
498.34	477.92	6,584	1,141,505	365,867	Distributed.net Client v1.05 (opencl_nvidia_101)
473.03	448.44	6,144	1,122,216	359,684	Distributed.net Client v1.05 (opencl_nvidia_101)
489.27	469.34	6,428	1,135,502	363,943	Averages for nVidia GTX1660 Ti, 130W, 1 WU per proc

[davidBAM]

Yup - about 10% better with the better CPU, 20% better on points per KwH

I've come to the conclusion that Moo is a 2 core per nVidia GPU app - certainly on slower cpus

[davidBAM]

Notice that this uses much less CPU than does the nVidia

Run Time	CPU Time	Credit	Per day/proc	Credit per kWH	Application
819.78	60.22	6,144	647,542	145,843	Distributed.net Client v1.05 (opencl_ati_101)
796.26	65.35	6,592	715,280	161,099	Distributed.net Client v1.05 (opencl_ati_101)
826.66	65.16	6,592	688,976	155,175	Distributed.net Client v1.05 (opencl_ati_101)
797.33	64.32	6,592	714,320	160,883	Distributed.net Client v1.05 (opencl_ati_101)
795.21	65.54	6,592	716,224	161,312	Distributed.net Client v1.05 (opencl_ati_101)
811.25	65.77	6,592	702,063	158,122	Distributed.net Client v1.05 (opencl_ati_101)
811.36	65.72	6,592	701,968	158,101	Distributed.net Client v1.05 (opencl_ati_101)
797.35	65.35	6,592	714,302	160,879	Distributed.net Client v1.05 (opencl_ati_101)
799.10	65.22	6,592	712,738	160,527	Distributed.net Client v1.05 (opencl_ati_101)
797.04	64.94	6,592	714,580	160,941	Distributed.net Client v1.05 (opencl_ati_101)
797.75	63.62	6,592	713,944	160,798	Distributed.net Client v1.05 (opencl_ati_101)
795.07	62.90	6,584	715,481	161,144	Distributed.net Client v1.05 (opencl_ati_101)
811.26	65.60	6,584	701,203	157,929	Distributed.net Client v1.05 (opencl_ati_101)
795.01	64.76	6,584	715,535	161,157	Distributed.net Client v1.05 (opencl_ati_101)
797.13	64.04	6,584	713,632	160,728	Distributed.net Client v1.05 (opencl_ati_101)
735.40	54.26	6,144	721,841	162,577	Distributed.net Client v1.05 (opencl_ati_101)
744.37	61.04	6,144	713,142	160,618	Distributed.net Client v1.05 (opencl_ati_101)
744.32	60.61	6,144	713,190	160,628	Distributed.net Client v1.05 (opencl_ati_101)
744.37	61.43	6,144	713,142	160,618	Distributed.net Client v1.05 (opencl_ati_101)
757.51	62.85	6,144	700,772	157,831	Distributed.net Client v1.05 (opencl_ati_101)
788.68	63.44	6,456	707,494	159,345	Averages for AMD RX580 GPU, 185W, 1 WU per proc

[davidBAM]

Run Time	CPU Time	Credit	Per day/proc	Credit per kWH	Application
294.88	279.16	6,160	1,804,883	406,505	Distributed.net Client v1.05 (opencl_nvidia_101)
313.13	299.40	6,592	1,818,889	409,660	Distributed.net Client v1.05 (opencl_nvidia_101)
291.92	279.59	6,144	1,818,449	409,561	Distributed.net Client v1.05 (opencl_nvidia_101)
289.97	280.87	6,144	1,830,678	412,315	Distributed.net Client v1.05 (opencl_nvidia_101)
288.00	277.58	6,144	1,843,200	415,135	Distributed.net Client v1.05 (opencl_nvidia_101)
292.09	279.63	6,144	1,817,391	409,322	Distributed.net Client v1.05 (opencl_nvidia_101)
289.92	279.60	6,144	1,830,993	412,386	Distributed.net Client v1.05 (opencl_nvidia_101)
292.07	279.61	6,144	1,817,515	409,350	Distributed.net Client v1.05 (opencl_nvidia_101)
289.95	280.70	6,144	1,830,804	412,343	Distributed.net Client v1.05 (opencl_nvidia_101)
301.08	291.46	6,144	1,763,125	397,100	Distributed.net Client v1.05 (opencl_nvidia_101)
288.84	277.56	6,144	1,837,840	413,928	Distributed.net Client v1.05 (opencl_nvidia_101)
290.09	279.82	6,144	1,829,920	412,144	Distributed.net Client v1.05 (opencl_nvidia_101)
311.98	300.43	6,600	1,827,809	411,669	Distributed.net Client v1.05 (opencl_nvidia_101)
312.05	296.43	6,528	1,807,464	407,087	Distributed.net Client v1.05 (opencl_nvidia_101)
312.06	300.39	6,584	1,822,911	410,566	Distributed.net Client v1.05 (opencl_nvidia_101)
311.09	301.49	6,584	1,828,595	411,846	Distributed.net Client v1.05 (opencl_nvidia_101)
307.79	297.56	6,584	1,848,200	416,261	Distributed.net Client v1.05 (opencl_nvidia_101)
290.00	279.95	6,144	1,830,488	412,272	Distributed.net Client v1.05 (opencl_nvidia_101)
310.96	300.51	6,584	1,829,359	412,018	Distributed.net Client v1.05 (opencl_nvidia_101)
311.08	301.46	6,584	1,828,654	411,859	Distributed.net Client v1.05 (opencl_nvidia_101)
299.45	288.16	6,319	1,823,358	410,666	Averages for nVidia RTX2070 super undervolt, 185W, 1 WU per proc

[VietOZ]

Run Time	CPU Time	Credit	Per day/proc	Credit per kWH	Application
457.28	436.85	6,144	1,160,868	326,821	Distributed.net Client v1.05 (opencl_nvidia_101)
454.51	439.74	6,144	1,167,943	328,813	Distributed.net Client v1.05 (opencl_nvidia_101)
478.43	464.04	6,584	1,189,009	334,744	Distributed.net Client v1.05 (opencl_nvidia_101)
480.36	465.47	6,592	1,185,671	333,804	Distributed.net Client v1.05 (opencl_nvidia_101)
458.36	443.81	6,144	1,158,132	326,051	Distributed.net Client v1.05 (opencl_nvidia_101)
458.35	444.52	6,144	1,158,158	326,058	Distributed.net Client v1.05 (opencl_nvidia_101)
495.52	481.33	6,648	1,159,160	326,340	Distributed.net Client v1.05 (opencl_nvidia_101)
450.25	434.80	6,144	1,178,993	331,924	Distributed.net Client v1.05 (opencl_nvidia_101)
480.32	465.60	6,584	1,184,330	333,426	Distributed.net Client v1.05 (opencl_nvidia_101)
486.50	471.14	6,592	1,170,707	329,591	Distributed.net Client v1.05 (opencl_nvidia_101)
460.37	443.75	6,144	1,153,076	324,627	Distributed.net Client v1.05 (opencl_nvidia_101)
454.41	443.71	6,144	1,168,200	328,885	Distributed.net Client v1.05 (opencl_nvidia_101)
454.20	435.79	6,144	1,168,740	329,037	Distributed.net Client v1.05 (opencl_nvidia_101)
490.51	474.40	6,584	1,159,727	326,500	Distributed.net Client v1.05 (opencl_nvidia_101)
483.37	464.51	6,648	1,188,297	334,543	Distributed.net Client v1.05 (opencl_nvidia_101)
488.36	472.61	6,608	1,169,079	329,132	Distributed.net Client v1.05 (opencl_nvidia_101)
480.44	464.49	6,440	1,158,138	326,052	Distributed.net Client v1.05 (opencl_nvidia_101)
445.14	430.85	6,144	1,192,527	335,734	Distributed.net Client v1.05 (opencl_nvidia_101)
463.26	450.83	6,376	1,189,152	334,784	Distributed.net Client v1.05 (opencl_nvidia_101)
449.20	436.71	6,144	1,181,749	332,700	Distributed.net Client v1.05 (opencl_nvidia_101)
468.46	453.25	6,355	1,172,083	329,978	Averages for evga rtx2060ko, 148W, 1 WU per proc

[SolidAir79]

I thought that the rtx 2060 ko might be better on something like asteriods as some of the compute capabilities have not been disabled have you tried it VietOZ ?

[davidBAM]

Over 10x variation in run times for essentially the same payment. Fecking madness

Run Time	CPU Time	Credit	Per day/proc	Credit per kWH	Application
8257.16	30.25	6,144	64,289	14,479	Distributed.net Client v1.05 (opencl_ati_101)
9347.82	24.62	6,144	56,788	12,790	Distributed.net Client v1.05 (opencl_ati_101)
10038.32	20.34	6,144	52,882	11,910	Distributed.net Client v1.05 (opencl_ati_101)
10333.46	24.21	6,584	55,050	12,399	Distributed.net Client v1.05 (opencl_ati_101)
9028.98	25.55	6,144	58,793	13,242	Distributed.net Client v1.05 (opencl_ati_101)
789.91	54.07	6,144	672,028	151,358	Distributed.net Client v1.05 (opencl_ati_101)
819.31	50.81	6,584	694,313	156,377	Distributed.net Client v1.05 (opencl_ati_101)
815.26	50.39	6,520	690,980	155,626	Distributed.net Client v1.05 (opencl_ati_101)
766.59	49.75	6,144	692,471	155,962	Distributed.net Client v1.05 (opencl_ati_101)
781.80	54.24	6,144	678,999	152,928	Distributed.net Client v1.05 (opencl_ati_101)
9352.34	17.25	6,144	56,760	12,784	Distributed.net Client v1.05 (opencl_ati_101)
1011.99	37.11	6,144	524,552	118,142	Distributed.net Client v1.05 (opencl_ati_101)
766.59	48.98	6,144	692,471	155,962	Distributed.net Client v1.05 (opencl_ati_101)
9596.39	16.69	6,144	55,317	12,459	Distributed.net Client v1.05 (opencl_ati_101)
836.55	58.05	6,584	680,004	153,154	Distributed.net Client v1.05 (opencl_ati_101)
761.53	48.17	6,144	697,072	156,998	Distributed.net Client v1.05 (opencl_ati_101)
759.50	46.66	6,144	698,936	157,418	Distributed.net Client v1.05 (opencl_ati_101)
815.35	56.88	6,440	682,426	153,700	Distributed.net Client v1.05 (opencl_ati_101)
4,159.94	39.67	6,255	433,563	97,649	Averages for AMD RX580 GPU, 185W, 1 WU per proc

[davidBAM]

Run Time	CPU Time	Credit	Per day/proc	Credit per kWH	Application
165.14	152.80	6,224	3,256,350	387,661	Distributed.net Client v1.05 (opencl_nvidia_101)
177.15	161.88	6,568	3,203,360	381,352	Distributed.net Client v1.05 (opencl_nvidia_101)
163.17	151.98	6,144	3,253,304	387,298	Distributed.net Client v1.05 (opencl_nvidia_101)
162.14	142.48	6,144	3,273,971	389,758	Distributed.net Client v1.05 (opencl_nvidia_101)
152.19	137.97	6,144	3,488,019	415,240	Distributed.net Client v1.05 (opencl_nvidia_101)
162.19	147.16	6,584	3,507,353	417,542	Distributed.net Client v1.05 (opencl_nvidia_101)
154.17	146.25	6,144	3,443,222	409,907	Distributed.net Client v1.05 (opencl_nvidia_101)
166.19	153.86	6,584	3,422,935	407,492	Distributed.net Client v1.05 (opencl_nvidia_101)
169.16	152.95	6,584	3,362,838	400,338	Distributed.net Client v1.05 (opencl_nvidia_101)
160.16	142.52	6,144	3,314,446	394,577	Distributed.net Client v1.05 (opencl_nvidia_101)
151.20	145.81	6,144	3,510,857	417,959	Distributed.net Client v1.05 (opencl_nvidia_101)
151.15	137.75	6,144	3,512,019	418,097	Distributed.net Client v1.05 (opencl_nvidia_101)
153.17	146.25	6,144	3,465,702	412,584	Distributed.net Client v1.05 (opencl_nvidia_101)
150.15	138.94	6,144	3,535,409	420,882	Distributed.net Client v1.05 (opencl_nvidia_101)
153.17	145.66	6,144	3,465,702	412,584	Distributed.net Client v1.05 (opencl_nvidia_101)
152.15	138.22	6,144	3,488,936	415,350	Distributed.net Client v1.05 (opencl_nvidia_101)
151.17	138.05	6,144	3,511,554	418,042	Distributed.net Client v1.05 (opencl_nvidia_101)
162.18	147.06	6,584	3,507,569	417,568	Distributed.net Client v1.05 (opencl_nvidia_101)
162.17	156.00	6,584	3,507,786	417,594	Distributed.net Client v1.05 (opencl_nvidia_101)
173.18	163.06	6,584	3,284,777	391,045	Distributed.net Client v1.05 (opencl_nvidia_101)
159.57	147.33	6,301	3,415,805	406,643	Averages for rtx3090, 350W, 1 WU per proc

[davidBAM]

Run Time	CPU Time	Credit	Per day/proc	Credit per kWH	Application
194.07	179.79	6,144	2,735,310	356,160	Distributed.net Client v1.05 (opencl_nvidia_101)
206.11	189.71	6,584	2,759,971	359,371	Distributed.net Client v1.05 (opencl_nvidia_101)
187.83	187.83	6,480	2,980,738	388,117	Distributed.net Client v1.05 (opencl_nvidia_101)
201.99	192.74	6,528	2,792,312	363,582	Distributed.net Client v1.05 (opencl_nvidia_101)
189.02	175.35	6,144	2,808,389	365,676	Distributed.net Client v1.05 (opencl_nvidia_101)
205.04	189.71	6,536	2,754,147	358,613	Distributed.net Client v1.05 (opencl_nvidia_101)
207.01	189.72	6,584	2,747,972	357,809	Distributed.net Client v1.05 (opencl_nvidia_101)
208.01	188.70	6,584	2,734,761	356,089	Distributed.net Client v1.05 (opencl_nvidia_101)
369.78	357.15	6,144	1,435,561	186,922	Distributed.net Client v1.05 (opencl_nvidia_101)
192.94	180.83	6,144	2,751,330	358,246	Distributed.net Client v1.05 (opencl_nvidia_101)
208.02	188.72	6,584	2,734,629	356,072	Distributed.net Client v1.05 (opencl_nvidia_101)
262.34	143.84	6,584	2,168,398	282,344	Distributed.net Client v1.05 (opencl_nvidia_101)
191.03	182.44	6,144	2,778,839	361,828	Distributed.net Client v1.05 (opencl_nvidia_101)
192.99	180.82	6,144	2,750,617	358,153	Distributed.net Client v1.05 (opencl_nvidia_101)
207.00	189.71	6,584	2,748,104	357,826	Distributed.net Client v1.05 (opencl_nvidia_101)
207.02	189.72	6,584	2,747,839	357,792	Distributed.net Client v1.05 (opencl_nvidia_101)
192.97	180.81	6,144	2,750,902	358,190	Distributed.net Client v1.05 (opencl_nvidia_101)
207.02	189.73	6,584	2,747,839	357,792	Distributed.net Client v1.05 (opencl_nvidia_101)
193.96	179.80	6,144	2,736,861	356,362	Distributed.net Client v1.05 (opencl_nvidia_101)
194.03	179.80	6,144	2,735,874	356,234	Distributed.net Client v1.05 (opencl_nvidia_101)
210.91	191.85	6,376	2,670,020	347,659	Averages for rtx3080, 320W, 1 WU per proc

[Miklos M]

Task name
click for details
Show IDs Work unit
click for details Sent Time reported
or deadline
explain Status Run time
(sec) CPU time
(sec) Credit Application
dnetc_r72_1606942977_12_768_0 105583136 2 Dec 2020, 21:47:13 UTC 6 Dec 2020, 0:58:54 UTC Completed and validated 212.51 184.08 6,144.00 Distributed.net Client v1.05 (opencl_nvidia_101)
windows_intelx86
Sorry I do not have a nice set up for this quote. It is my 2080Ti. They all take about 212-300 seconds per unit.

[davidBAM]

@Miklos M if you post me the URL showing validated results for one of your computers, I can do a table for you.

The link should look something like this https://moowrap.net/results.php?hostid= ... e=4&appid= (but with your hostid on it)

[Miklos M]

@Miklos M if you post me the URL showing validated results for one of your computers, I can do a table for you.

The link should look something like this https://moowrap.net/results.php?hostid= ... e=4&appid= (but with your hostid on it)

Thank you.
https://moowrap.net/results.php?hostid=1304502

[davidBAM]

Run Time	CPU Time	Credit	Per day/proc	Credit per kWH	Application
212.51	184.08	6,144	2,497,961	416,327	Distributed.net Client v1.05 (opencl_nvidia_101)
247.59	215.56	6,632	2,314,329	385,722	Distributed.net Client v1.05 (opencl_nvidia_101)
209.31	181.80	6,144	2,536,150	422,692	Distributed.net Client v1.05 (opencl_nvidia_101)
212.33	184.28	6,144	2,500,078	416,680	Distributed.net Client v1.05 (opencl_nvidia_101)
212.41	183.47	6,144	2,499,137	416,523	Distributed.net Client v1.05 (opencl_nvidia_101)
213.71	68.75	6,584	2,661,820	443,637	Distributed.net Client v1.05 (opencl_nvidia_101)
212.44	184.02	6,144	2,498,784	416,464	Distributed.net Client v1.05 (opencl_nvidia_101)
212.44	184.69	6,144	2,498,784	416,464	Distributed.net Client v1.05 (opencl_nvidia_101)
212.44	184.03	6,144	2,498,784	416,464	Distributed.net Client v1.05 (opencl_nvidia_101)
211.54	184.55	6,144	2,509,415	418,236	Distributed.net Client v1.05 (opencl_nvidia_101)
212.58	183.75	6,144	2,497,138	416,190	Distributed.net Client v1.05 (opencl_nvidia_101)
212.47	183.69	6,144	2,498,431	416,405	Distributed.net Client v1.05 (opencl_nvidia_101)
209.65	182.52	6,144	2,532,037	422,006	Distributed.net Client v1.05 (opencl_nvidia_101)
224.82	185.09	6,144	2,361,185	393,531	Distributed.net Client v1.05 (opencl_nvidia_101)
218.95	186.06	6,144	2,424,488	404,081	Distributed.net Client v1.05 (opencl_nvidia_101)
211.21	186.02	6,144	2,513,336	418,889	Distributed.net Client v1.05 (opencl_nvidia_101)
197.12	179.63	6,144	2,692,987	448,831	Distributed.net Client v1.05 (opencl_nvidia_101)
206.18	186.97	6,144	2,574,651	429,109	Distributed.net Client v1.05 (opencl_nvidia_101)
287.58	247.66	6,584	1,978,085	329,681	Distributed.net Client v1.05 (opencl_nvidia_101)
212.56	185.61	6,144	2,497,373	416,229	Distributed.net Client v1.05 (opencl_nvidia_101)
217.49	183.11	6,212	2,479,248	413,208	Averages for gtx2080ti, 250W, 1 WU per proc

[davidBAM]

There you go @Miklos M - the last line is the average of the 20 above it

[Miklos M]

There you go @Miklos M - the last line is the average of the 20 above it

It looks great, thank you.

[davidBAM]

do you have a measured wattage? That one is an optimistic one from nVidia

[Miklos M]

do you have a measured wattage? That one is an optimistic one from nVidia

Where would I see the wattage?

[Dirk Broer]

do you have a measured wattage? That one is an optimistic one from nVidia

Where would I see the wattage?

In the end: on your power bill

[davidBAM]

Lol

nvidia-smi on linux, not sure about windows

[Miklos M]

My kw used per hour for everything electrical is about 140-200 per day.

[SolidAir79]

Windows GPU-Z

[Miklos M]

Windows GPU-Z

What does that mean?

[SolidAir79]

Here you go mate link to the download on the page too https://www.techpowerup.com/gpuz/

[Miklos M]

Here you go mate link to the download on the page too https://www.techpowerup.com/gpuz/

Thank you, got it.

[Dirk Broer]

Windows GPU-Z

Or the more elaborate GPU-Caps