Archaea
[H]F Junkie
- Joined
- Oct 19, 2004
- Messages
- 11,828
I've had a couple Fury X and recently started mining (last couple weeks). I picked up three 1080TI and two arrived last night. I swapped out the cards and here are my initial observations.
Fury X cards:
http://xfxforce.com/en-us/products/...md-radeon-r9-furyx-liquid-cooled-r9-fury-4qfa
1080TI cards:
http://www.pny.com/geforce-gtx-1080ti-blower-edition-2
It's more profitable with nicehash to mine on the 1080TI pair of cards, and they are less power hungry too.
With the Fury X cards I was cutting power by about 20% for mining using AMD Wattman, much more of a cut than that and the nicehash miner would crash. With the 1080TI I'm cutting power by about 35% for mining using MSI Afterburner, much more of a cut than that and the nicehash miner crashes.
So power use is much lower on the 1080TI (stock Fury X at full 100% load is 275 watts, stock 1080TI at full 100% load is 250 watts). Since the 1080TI mines faster, uses less watts stock, and in addition can be have it's power cut back an additional 35% without much affecting performance - it's a winner overall for this mining challenge.
HOWEVER: Disadvantages to 1080TI mining I've seen so far are noteworthy!
1) My desktop user experience with mining on the 1080TI pair is worse. YouTube videos skip frames and stutter for instance. Whereas, with Fury X you couldn't even tell you were mining when the pair of GPUs were mining. I don't know if this is a difference in the mining algorithm or what. (Fury X pair was using Dagger Hashimoto, and the 1080TI are using Lyra2Rev2 and Equihash. Any suggestions here? I'd prefer to be able to use my PC normally when mining with nice hash even at the expense of a small amount of mining power. It's not terrible with the 1080TI, but you can tell it isn't a flawless desktop experience.
2) Fans on 1080TI (even at 65% power usage) are louder than Fury X and temps higher. (though certainly more heat is generated in the room with the Fury X)
3) you can't mine at full TDP or overclocked with the 1080TI (at least without an external fan blowing into the case. Even with the case side off the cards will thermal throttle at 85* if run at stock speeds/voltages. (Not a problem with the watercooled Fury X --- though I chose to undervolt on the Fury X too simply to save power and heat generation.)
At full stock speeds the Fury X will mine at 65* and 52% fan usage (but there is a LOT of heat generation in the room)
when TDP is lowered 20% they mine at say 64* and 32% fan usage (less heat generation in the room)
At full stock speeds the 1080TI will mine at 85* (and then throttle core speed slightly, then boost, repeat)
At 65% TDP the 1080TI will mine at 66* with fan speed manually set to 66%. The result of tuning is 325 watts power use instead of 500 watts power use as stated in nicehash Equihash miner after being tuned. The the difference in mining speed is very minimal as you can see in the screenshots. So far changing RAM speed doesn't seem to affect operation. For some unknown reason lowering core speed on the slider actually increases heat on my cards? Seems like a bug. Raising core speed increases hashing speed though -- so that's why you see the tuned 1080TI with a 200Mhz manual bump to core.
Stock Fury X pair
Mining tuned Fury X pair - (-20% power) (don't pay attention to the payout rate (that is variable)) - the MH/s is the stat to look at. I often saw the tuned at 57 or 58MH/s -- as compared to about 59MH/s stock. This is the screenshot I have though.
Stock 1080TI Pair
Mining tuned 1080TI Pair - (66% fan manual, -35% Power Limit, + 200Mhz core clock) (don't pay attention to the payout rate (that is variable)) - the GH/s is the stat to look at.
Fury X cards:
http://xfxforce.com/en-us/products/...md-radeon-r9-furyx-liquid-cooled-r9-fury-4qfa
1080TI cards:
http://www.pny.com/geforce-gtx-1080ti-blower-edition-2
It's more profitable with nicehash to mine on the 1080TI pair of cards, and they are less power hungry too.
With the Fury X cards I was cutting power by about 20% for mining using AMD Wattman, much more of a cut than that and the nicehash miner would crash. With the 1080TI I'm cutting power by about 35% for mining using MSI Afterburner, much more of a cut than that and the nicehash miner crashes.
So power use is much lower on the 1080TI (stock Fury X at full 100% load is 275 watts, stock 1080TI at full 100% load is 250 watts). Since the 1080TI mines faster, uses less watts stock, and in addition can be have it's power cut back an additional 35% without much affecting performance - it's a winner overall for this mining challenge.
HOWEVER: Disadvantages to 1080TI mining I've seen so far are noteworthy!
1) My desktop user experience with mining on the 1080TI pair is worse. YouTube videos skip frames and stutter for instance. Whereas, with Fury X you couldn't even tell you were mining when the pair of GPUs were mining. I don't know if this is a difference in the mining algorithm or what. (Fury X pair was using Dagger Hashimoto, and the 1080TI are using Lyra2Rev2 and Equihash. Any suggestions here? I'd prefer to be able to use my PC normally when mining with nice hash even at the expense of a small amount of mining power. It's not terrible with the 1080TI, but you can tell it isn't a flawless desktop experience.
2) Fans on 1080TI (even at 65% power usage) are louder than Fury X and temps higher. (though certainly more heat is generated in the room with the Fury X)
3) you can't mine at full TDP or overclocked with the 1080TI (at least without an external fan blowing into the case. Even with the case side off the cards will thermal throttle at 85* if run at stock speeds/voltages. (Not a problem with the watercooled Fury X --- though I chose to undervolt on the Fury X too simply to save power and heat generation.)
At full stock speeds the Fury X will mine at 65* and 52% fan usage (but there is a LOT of heat generation in the room)
when TDP is lowered 20% they mine at say 64* and 32% fan usage (less heat generation in the room)
At full stock speeds the 1080TI will mine at 85* (and then throttle core speed slightly, then boost, repeat)
At 65% TDP the 1080TI will mine at 66* with fan speed manually set to 66%. The result of tuning is 325 watts power use instead of 500 watts power use as stated in nicehash Equihash miner after being tuned. The the difference in mining speed is very minimal as you can see in the screenshots. So far changing RAM speed doesn't seem to affect operation. For some unknown reason lowering core speed on the slider actually increases heat on my cards? Seems like a bug. Raising core speed increases hashing speed though -- so that's why you see the tuned 1080TI with a 200Mhz manual bump to core.
Stock Fury X pair
Mining tuned Fury X pair - (-20% power) (don't pay attention to the payout rate (that is variable)) - the MH/s is the stat to look at. I often saw the tuned at 57 or 58MH/s -- as compared to about 59MH/s stock. This is the screenshot I have though.
Stock 1080TI Pair
Mining tuned 1080TI Pair - (66% fan manual, -35% Power Limit, + 200Mhz core clock) (don't pay attention to the payout rate (that is variable)) - the GH/s is the stat to look at.
Last edited: