Re: [ANN] lolMiner 0.42, Update: Aug 17th: Better AMD performance for Equi 144/5
Hello everyone, did some testing today with old cards (hawaii and tahiti).
Modded bios on every card (mem timings), slightly overclocked, some cards underperform, due to the lack of good memory timings.
60s avarage speeds
hawaii devices (0,1,3,5,7,8)
tahiti devices (2,4,6,9)
So overall performance is not bad, but i think there is some potential for improvement.
Comparing the old equihash 200,9 hashrates from those cards to a nvidia 1080ti with zhash hashrates (only comparison i can get first hand) we get 400Sol/s from a 390X vs 780Sol/s from a 1080ti. When mining Zhash, we get: 27 Sol/s vs 70 Sol/s for now, mining ethhash would be like 30Mh/s vs 52Mh/s. So the difference between the two cards would be somewhere like < 1:2. I think there are some other algo's with a similar factor between the two cards, so if there is not more potential for like 5-10Sol/s i would be sad.
But for now great work Lolliedieb!
Thats my feedback for old AMD cards (they have still juice
), running longterm test with these 10 cards now.
Modded bios on every card (mem timings), slightly overclocked, some cards underperform, due to the lack of good memory timings.
60s avarage speeds
hawaii devices (0,1,3,5,7,8)
tahiti devices (2,4,6,9)
Quote
device 0: AMD Radeon R9 200 Series
290X Sapphire 4GB (1040 MHz / 1400 MHz)
usage: core 100% / mem controller 30%
24.7 - 25.9 Sol/s ~115W
device 1: AMD Radeon (TM) R9 390 Series
390X Asus 8GB (1040 MHz / 1500 MHz)
usage: core 100% / mem controller 30%
25.9 - 27.6 Sol/s ~115W
device 2: AMD Radeon R9 200 / HD 7900 Series
HD 7970 Gigabyte 3GB (1050 MHz / 1400 MHz)
usage: core 99% / mem controller ? (that's normal for tahiti cards gpuz readings)
18.6 - 19.6 Sol/s ?
device 3: AMD Radeon R9 200 Series
290X Sapphire 8GB (1040 MHz / 1500 MHz)
usage: core 100% / mem controller 30%
22.8 - 23.2 Sol/s ~185W
device 4: AMD Radeon R9 200 / HD 7900 Series
HD 7970 Gigabyte 3GB (1050 MHz / 1400 MHz)
usage: core 99% / mem controller ?
18.4 - 19.4 Sol/s ?
device 5: AMD Radeon R9 200 Series
290X Sapphire 4GB (1080 MHz / 1400 MHz)
usage: core 100% / mem controller 30%
24.0 - 24.7 Sol/s ~135W
device 6: AMD Radeon R9 200 Series
280X HiS 3GB (1050 MHz / 1500 MHz)
usage: core 98% / mem controller ?
19.1 - 19.6 Sol/s ?
device 7: AMD Radeon R9 200 Series
290X MSI 4GB (1030 MHz / 1400 MHz)
usage: core 100% / mem controller 30%
21.1 - 22.5 Sol/s ~160W
device 8: AMD Radeon (TM) R9 390 Series
390X Sapphire 8GB (1080 MHz / 1500 MHz)
usage: core 100% / mem controller 30%
26.0 - 27.5 Sol/s ~170W
device 9: AMD Radeon R9 200 / HD 7900 Series
HD 7970 Gigabyte 3GB (1050 MHz / 1400 MHz)
usage: core 99% / mem controller ?
18.5 - 19.7 Sol/s ?
290X Sapphire 4GB (1040 MHz / 1400 MHz)
usage: core 100% / mem controller 30%
24.7 - 25.9 Sol/s ~115W
device 1: AMD Radeon (TM) R9 390 Series
390X Asus 8GB (1040 MHz / 1500 MHz)
usage: core 100% / mem controller 30%
25.9 - 27.6 Sol/s ~115W
device 2: AMD Radeon R9 200 / HD 7900 Series
HD 7970 Gigabyte 3GB (1050 MHz / 1400 MHz)
usage: core 99% / mem controller ? (that's normal for tahiti cards gpuz readings)
18.6 - 19.6 Sol/s ?
device 3: AMD Radeon R9 200 Series
290X Sapphire 8GB (1040 MHz / 1500 MHz)
usage: core 100% / mem controller 30%
22.8 - 23.2 Sol/s ~185W
device 4: AMD Radeon R9 200 / HD 7900 Series
HD 7970 Gigabyte 3GB (1050 MHz / 1400 MHz)
usage: core 99% / mem controller ?
18.4 - 19.4 Sol/s ?
device 5: AMD Radeon R9 200 Series
290X Sapphire 4GB (1080 MHz / 1400 MHz)
usage: core 100% / mem controller 30%
24.0 - 24.7 Sol/s ~135W
device 6: AMD Radeon R9 200 Series
280X HiS 3GB (1050 MHz / 1500 MHz)
usage: core 98% / mem controller ?
19.1 - 19.6 Sol/s ?
device 7: AMD Radeon R9 200 Series
290X MSI 4GB (1030 MHz / 1400 MHz)
usage: core 100% / mem controller 30%
21.1 - 22.5 Sol/s ~160W
device 8: AMD Radeon (TM) R9 390 Series
390X Sapphire 8GB (1080 MHz / 1500 MHz)
usage: core 100% / mem controller 30%
26.0 - 27.5 Sol/s ~170W
device 9: AMD Radeon R9 200 / HD 7900 Series
HD 7970 Gigabyte 3GB (1050 MHz / 1400 MHz)
usage: core 99% / mem controller ?
18.5 - 19.7 Sol/s ?
So overall performance is not bad, but i think there is some potential for improvement.
Comparing the old equihash 200,9 hashrates from those cards to a nvidia 1080ti with zhash hashrates (only comparison i can get first hand) we get 400Sol/s from a 390X vs 780Sol/s from a 1080ti. When mining Zhash, we get: 27 Sol/s vs 70 Sol/s for now, mining ethhash would be like 30Mh/s vs 52Mh/s. So the difference between the two cards would be somewhere like < 1:2. I think there are some other algo's with a similar factor between the two cards, so if there is not more potential for like 5-10Sol/s i would be sad.
But for now great work Lolliedieb!
Thats my feedback for old AMD cards (they have still juice
), running longterm test with these 10 cards now.