fullzero (OP)
Legendary
 Offline
Activity: 1260
Merit: 1009
|
 |
March 16, 2019, 12:59:20 PM |
|
1600 series Nvidia GPUs are significantly better than everything else for mining. The 1660ti almost matches a 1070's performance with only 75 watts. Currently, most mining clients don’t support the 1660 or 1660ti. Linux drivers currently only have limited support for the 1660 and 1660ti. The 1670 or 1670ti will be the next GPU that will be optimal to buy for mining. They will likely have a 150-160 watt TDP, with an optimal mining powerlimit around 90-95 watts. plan accordingly  |
|
|
|
|
|
|
|
"This isn't the kind of software where we can leave so many unresolved bugs that we need a tracker for them." -- Satoshi
|
|
|
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
|
|
|
|
|
Xazax310
Member
Offline
Activity: 246
Merit: 24
|
|
March 16, 2019, 07:11:16 PM |
|
agreed that both the 1660 and 1660ti are excellent mining cards, at least for everything but Ethereum/XMR. They do OK in ethereum AMDs still hold XMR throne. I doubt we'll see anything better than 1660ti. I assumed something was in the works as nvidias "Mining GPU" however the crash left a sour note in there mouths. Doubt we'll ever see mining GPU from nvidia again.
|
|
|
|
|
|
Marvell2
|
|
March 16, 2019, 07:53:11 PM |
|
1600 series Nvidia GPUs are significantly better than everything else for mining. The 1660ti almost matches a 1070's performance with only 75 watts. Currently, most mining clients don’t support the 1660 or 1660ti. Linux drivers currently only have limited support for the 1660 and 1660ti. The 1670 or 1670ti will be the next GPU that will be optimal to buy for mining. They will likely have a 150-160 watt TDP, with an optimal mining powerlimit around 90-95 watts. plan accordingly Is the 1670 confirmed? |
|
|
|
|
fullzero (OP)
Legendary
Offline
Activity: 1260
Merit: 1009
|
|
March 18, 2019, 03:21:19 PM |
|
agreed that both the 1660 and 1660ti are excellent mining cards, at least for everything but Ethereum/XMR. They do OK in ethereum AMDs still hold XMR throne. I doubt we'll see anything better than 1660ti. I assumed something was in the works as nvidias "Mining GPU" however the crash left a sour note in there mouths. Doubt we'll ever see mining GPU from nvidia again.
1660ti does 507 H/s with XMR and 30+ MH/s using 60-70 watts for Ethash. It was never a good idea to use the nerfed GPUs that were targeted for mining. It is always better to retain the possibility of reselling GPUs to gamers in the future. This is the best aspect of the 1660s, or 1660tis as they are mainstream Gaming GPUs. Nvidia's stock price only took a dive with the release of the RTX 2000 series GPUs. As Ray tracing is generally held to be irrelevant to gamers and is useless for miners; the 1600 series should resolve this problem. |
|
|
|
|
fullzero (OP)
Legendary
Offline
Activity: 1260
Merit: 1009
|
|
March 18, 2019, 03:22:06 PM |
|
1600 series Nvidia GPUs are significantly better than everything else for mining. The 1660ti almost matches a 1070's performance with only 75 watts. Currently, most mining clients don’t support the 1660 or 1660ti. Linux drivers currently only have limited support for the 1660 and 1660ti. The 1670 or 1670ti will be the next GPU that will be optimal to buy for mining. They will likely have a 150-160 watt TDP, with an optimal mining powerlimit around 90-95 watts. plan accordingly Is the 1670 confirmed? I don't believe it is. |
|
|
|
|
|
dragonmike
|
|
March 18, 2019, 03:37:43 PM |
|
Don't completely agree. Yes, the 16xx series is good for mining, but it's not the best by a huge margin. As a matter of fact, core-bound algos scale very well across the range.
1660 -> 1660Ti -> 2060 -> 2070 etc all have very similar hash/watt and hash/dollar. It's now very much down to how much you want to spend, and build your rigs accordingly.
Basically, only the 2080Ti and Titans are disconnected. They are WAY more expensive and ratios are off.
|
|
|
|
|
|
P00P135
|
|
March 18, 2019, 03:49:29 PM |
|
Has anyone done some algo testing with the plain 1660's with DDR5 yet?
|
|
|
|
|
|
lunobird
|
|
March 18, 2019, 04:22:57 PM |
|
2080 rtx is the new king. I don't touch low class stuff like the 1660 ti. Compare mining specs on grin.
|
|
|
|
|
|
fanatic26_
|
|
March 18, 2019, 06:47:38 PM |
|
2080 rtx is the new king. I don't touch low class stuff like the 1660 ti. Compare mining specs on grin.
If you are interested in a multi year payback time sure. You would have to be an absolute fool to spend money on RTX cards for mining at this point.... |
|
|
|
|
badbart
Member
Offline
Activity: 449
Merit: 24
|
|
March 18, 2019, 07:16:27 PM |
|
2080 rtx is the new king. I don't touch low class stuff like the 1660 ti. Compare mining specs on grin.
With low electrical cost a 2080 will take you 1,000 days to payback your initial investment. |
|
|
|
|
|
lunobird
|
|
March 19, 2019, 06:57:49 AM
Last edit: March 19, 2019, 04:20:48 PM by lunobird |
|
2080 rtx is the new king. I don't touch low class stuff like the 1660 ti. Compare mining specs on grin.
If you are interested in a multi year payback time sure. You would have to be an absolute fool to spend money on RTX cards for mining at this point.... oh really lets see here. https://www.reddit.com/r/gpumining/comments/avyaq1/gtx_1660_ti_mining_performance/MTP-Zcoin (CryptoDredge 0.17) 1.84 Mhs Core +80 / Mem +600 Powerlimit 75 Intensity 6 90 Watts My 2080 rtx gets double that at 3.6 mhz core +125/ Mem +175 Power limit 65 . 150 Watts. That means you would have to buy two of these gtx 1660 ti to match a 2080 rtx performance and your using up more energy in the process at 180 watts total vs 150watts. In this multi year bear/accumulation game its all about efficiency because your just 1 watt away from getting liquidated, and this proves the 2080 rtx is more efficient and worth paying a little bit more for. I've seen 2080 rtx on sale for $660 last week on B&H. You get the extra rtx cores for gaming, better resale value, more future proof, more dense rig setup and more energy efficiency. Most miners care about daily profits on whattomine and have fast they can break even and be stuck with junk hardware. These are your typical home hobbist miners that will never be able to expand and always be small timers I think two steps ahead and focus on whats going to get me the maximum amount of money in a few years of mining and not selling a single coin till we hit next parabolic bull cycle. I always pick out my gpu's based on a rig setup that will allow me to mine the most popular new coins like grin and beam early before asics hit them and I pick hardware that has most efficiency for the long game so i don't get gpu liquidated. |
|
|
|
|
|
fanatic26_
|
Maths are hard for some people i guess 1660Ti = $299 MSRP RTX 2080 = $799 MSRP If you can find the 2080 on a 1 off low price, you will be able to do the same thing for the 1660 once its a month or two old. You are cherry picking your data to prove your point even though you are standing on very shaky ground. You even had to single out one particular algo and a random useless shitcoin to prove your point. Lets look at something more popular...say ETH hash rates. 1660Ti = 29.5MH @ 90w (Source: https://www.reddit.com/r/gpumining/comments/avyaq1/gtx_1660_ti_mining_performance/ ) 2080 = 42.3MH @ 168w (Source: https://www.reddit.com/r/gpumining/comments/9ixng5/rtx_2080_mining_hashrateperformance/ ) So on the most popular algo you can get a pair of 1660s doing 59mh @ 180w or a single 2080 doing 42.3mh @ 168w.... PHi2 1660Ti = 9.2mh @ 90w 2080 = 9.97mh @ 168w The 1660 dominates here.... x16r 1660TI = 17.8mh @ 90w 2080 = 30mh @ 168w Close enough to be nearly identical.... See when you cherry pick results its easy to make claims that your way is the best. The unfortunate side effect is someone might listen to your bs and make a poor decision after reading this because they dont know how to do proper research on their own. |
|
|
|
|
|
lunobird
|
|
March 19, 2019, 04:34:38 PM |
|
Maths are hard for some people i guess 1660Ti = $299 MSRP RTX 2080 = $799 MSRP If you can find the 2080 on a 1 off low price, you will be able to do the same thing for the 1660 once its a month or two old. You are cherry picking your data to prove your point even though you are standing on very shaky ground. You even had to single out one particular algo and a random useless shitcoin to prove your point. Lets look at something more popular...say ETH hash rates. 1660Ti = 29.5MH @ 90w (Source: https://www.reddit.com/r/gpumining/comments/avyaq1/gtx_1660_ti_mining_performance/ ) 2080 = 42.3MH @ 168w (Source: https://www.reddit.com/r/gpumining/comments/9ixng5/rtx_2080_mining_hashrateperformance/ ) So on the most popular algo you can get a pair of 1660s doing 59mh @ 180w or a single 2080 doing 42.3mh @ 168w.... PHi2 1660Ti = 9.2mh @ 90w 2080 = 9.97mh @ 168w The 1660 dominates here.... x16r 1660TI = 17.8mh @ 90w 2080 = 30mh @ 168w Close enough to be nearly identical.... See when you cherry pick results its easy to make claims that your way is the best. The unfortunate side effect is someone might listen to your bs and make a poor decision after reading this because they dont know how to do proper research on their own. Zcoin is not a shitcoin it's the number #3 coin by market cap for best masternodes and it's in the first page of the top 100 coins. If I wanted the best eth miner I would get a Vega vii. Not 1660 ti. And it's true 2080 rtx sucks at eth. It has an achiles heel. But it rocks all other algos like a boss. |
|
|
|
|
|
fanatic26_
|
|
March 19, 2019, 04:46:40 PM |
|
Zcoin is not a shitcoin it's the number #3 coin by market cap for best masternodes and it's in the first page of the top 100 coins.
If I wanted the best eth miner I would get a Vega vii. Not 1660 ti.
And it's true 2080 rtx sucks at eth. It has an achiles heel. But it rocks all other algos like a boss.
Lol its number 96 out of 100 on coinmarketcap with a horrible market volume that makes it the very definition of a shitcoin. I mean it was popular a lil over a year ago before settling down into the basement with all the other dying coins. Again none of that addresses your "two steps ahead" philosophy or your claim its so much more efficient even at 2.5 times+ the cost. You seem to have a hard on for new overpriced cards saying you would buy the Vega VII as well. Not sure if you followed mining after the release of the original Vega but the drivers are shit and it will take months and months and months before people bother adding support for it in their miners. Its yet another poor purchase choice in the current market. |
|
|
|
|
|
|
|
fanatic26_
|
|
March 19, 2019, 05:07:39 PM |
|
Cuckaroo29 is a GPU focused algo specifically designed to be phased out by next year in favor of the ASIC focused Cukaroo31 algo so if you are looking long term this is not the algo to be focusing on. That’s why Grin introduced two proof-of-work algorithms. The primary one is designed to be ASIC friendly and the secondary one ASIC resistant. At launch, the secondary one PoW will represent approximately 90% of the blocks found while the primary one will be roughly 10%. This percentage will evolve such that in two years 100% of the blocks are mined with the primary PoW, thus encouraging ASICs manufacturers to develop a machine for the primary algorithm. Source: https://blog.blockcypher.com/an-introduction-to-grin-proof-of-work-103aaa9f66ce |
|
|
|
|
nars28
Newbie
Offline
Activity: 25
Merit: 0
|
|
March 19, 2019, 05:10:58 PM |
|
Funny title. Radeons RX470 bought in 2016 for 165€ are hashing ethash 29.2 MH/s and Cryptonight V4 @ 1000 H/s. Consuming around 90 watts. So yeah 1600 series are "significantly" better than everything else   |
|
|
|
|
|
P00P135
|
|
March 19, 2019, 08:10:38 PM |
|
I think two steps ahead and focus on whats going to get me the maximum amount of money in a few years of mining and not selling a single coin till we hit next parabolic bull cycle. I always pick out my gpu's based on a rig setup that will allow me to mine the most popular new coins like grin and beam early before asics hit them and I pick hardware that has most efficiency for the long game so i don't get gpu liquidated.
Man you sound so delusional its hilarious |
|
|
|
|
badbart
Member
Offline
Activity: 449
Merit: 24
|
|
March 19, 2019, 08:23:44 PM |
|
I think two steps ahead and focus on whats going to get me the maximum amount of money in a few years of mining and not selling a single coin till we hit next parabolic bull cycle. I always pick out my gpu's based on a rig setup that will allow me to mine the most popular new coins like grin and beam early before asics hit them and I pick hardware that has most efficiency for the long game so i don't get gpu liquidated.
Man you sound so delusional its hilarious Same here good stuff but is eth going to swithc to PROGPOW? What kind of performance will a 470 have in progpow? Rx 400 and 500 eries are cheap right now if they preform well on prog they are a steal. |
|
|
|
|
|
lunobird
|
|
March 19, 2019, 09:00:59 PM |
|
Cuckaroo29 is a GPU focused algo specifically designed to be phased out by next year in favor of the ASIC focused Cukaroo31 algo so if you are looking long term this is not the algo to be focusing on. That’s why Grin introduced two proof-of-work algorithms. The primary one is designed to be ASIC friendly and the secondary one ASIC resistant. At launch, the secondary one PoW will represent approximately 90% of the blocks found while the primary one will be roughly 10%. This percentage will evolve such that in two years 100% of the blocks are mined with the primary PoW, thus encouraging ASICs manufacturers to develop a machine for the primary algorithm. Source: https://blog.blockcypher.com/an-introduction-to-grin-proof-of-work-103aaa9f66ceHence the reason why you should get a 2080 rtx now and mine early and get all the meat. The asic boys can have my scraps after they release their miner later at end of year. Getting in early now is way more profitable than getting in late with more powerful hardware. Plus I'm not focusing on grin only. Theres tons of algos the 2080 rtx does well at and much more efficient than a 1660 ti. |
|
|
|
|
|
