That did it! Ordering some Thai while it benches for nicehash. Then, onto power testing. Goal will probably be to maximize power use by cards, conditional on total power use <= M M tbd, will mostly be a function of observed refrigerator and monitor draw.
-
Show this thread
-
Early findings: (i) i7-5820k @ 3.8GHz draws only an extra 30W when full-tilt running CryptoNight (350H/s), that's awesome (ii) running a 1080TI FTW3 @ 100% pulls an extra 269W, about spec. Every 10% reduction in power target reduces draw by about 30W. Will test performace drop.
1 reply 0 retweets 5 likesShow this thread -
Fascinating. 1080TIs seem to see 1:1 gains in hash rates as power increases until about 80% TDP, when the return drops off dramatically 1080s drop off dramatically at . . . 65% TDP? Less? Interesting implications for power use optimization anyway
1 reply 0 retweets 1 likeShow this thread -
Ok. Final tally, for 4x1080TI @ 80%, 3 x 1080 @ 70%, 35 CPU cores >= 3GHz is . . . 1779 Watts Comfortably under my 1920W continuous limit although a bit dull numerologically main remaining concern is the fridge. rated 6.5A full-load, which would be (eep) 780W, insta-game over
2 replies 1 retweet 4 likesShow this thread -
-
Replying to @twitsshits @eigenrobot
Or assuming you are not a robot( lol yeah right) live off of rice and beans and other shit you don't need to refrigerate
1 reply 0 retweets 1 like -
Replying to @twitsshits
considered just eating at work, but wiferobot is moving in in a month and she'll want a functional fridge probably the things I do for love
1 reply 0 retweets 0 likes -
Replying to @eigenrobot
You could go the other direction and keep the apt cold enough that you don't need a fridge
1 reply 0 retweets 1 like -
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.