Performance with 2x EVGA GTX1080Ti’s
It took me a bit longer to write this review than I originally planned (life gets in the way), but I did receive my second EVGA GTX 1080 TI today, so I thought I would add some figures quickly to show the performance with two cards. As you know, when it comes to mining rigs, each additional card you can run will help offset the “system overhead”, not only in ROI calculations (as each card will contribute to the overall ROI) but also in greater overall power efficiency since additional each card will also help offset the systems share of power usage. So with two cards now in the system we can get a little bit better idea of those benefits.
My test rig is pictured above with the two GTX 1080Ti’s in it. My next article idea is to write a guide on how to build a mining rig from start to finish, and I think I am going to use this test platform to base it upon. In case you are wondering the test rig consists of the following components:
- MSI Z170A SLI Motherboard (Socket 1151)
- Intel Pentium G4400 Skylake (1151) Processor
- Intel SSD 600p Series 128GB (M.2 2280)
- GSkill Ripsaw DDR4 RAM (4 GB)
- CORSAIR RM1000i 1000W GOLD PSU
- 2x EVGA GTX1080Ti
- Windows 10 Pro 64-bit edition
Getting back to this review, the performance readings with optimized 78% PL, +150 Core, and +300 Memory settings with two GTX 1080Ti’s installed in the rig is shown below:
As you would expect, the original card’s (GPU1) performance is the same as before and the new card’s performance is also identical. I guess I would be surprised if the results were any different, so no big revelations here as the performance scaling is exactly as anticipated. The real benefit though which EWBF program will not monitor directly, is the overall system wattage only went up to 466 Watts. Referring the the earlier single card wattage readings we seen before, we were drawing 266 Watts for the single GPU and system combined.
So the difference between these two readings is 466W-266W or 200 watts. So the addition of an extra GTX1080Ti only raised the system wattage by 200 watts (which by the way is real close to EWBF’s reading). I hope you can see where this is leading, as now the overall system wattage will be divided by two cards when comparing Sols per Watt. To see what I mean look at two calculations below:
|Total Sols (Equihash)||Total System Power (Watts)||Overall Efficiency|
|One GPU Readings||700 Sol/s||266||2.63 Sol/W|
|Two GPU Readings||1400 Sol/s||466||3.00 Sol/W|
So while EWBF looks only at the GPU efficiency, we still need to manually calculate out the overall system efficiency counting the wattage overhead consumed by the rest of the system (CPU, Motherboard, RAM, FANS, PSU, etc.). As you can see by the above, adding additional GPUs will raise the overall efficiency as each GPU will help offset a portion of the rest of the system. So if the above ratio holds, adding 2 more GPUs (for total of four, which is my plan) would bring the total wattage to somewhere close to 866 watts and hash rate would increase to 2800 Sol/s, bringing the overall system efficiency to around 3.23 Sol/W.
Inputting the figures we have obtained into What-to-Mine.com we can get rough estimates on the profit potential at this point in time. I used a rather high $0.125 figure for power, so your rate may be lower, but I like to keep myself grounded in reality and try to add a few cents to the cost account for other things. Also the What-to-mine front page calculator does not account for pool or miner fees, thus another reason I go high on electricity. You can input these fees on the specific coins page if you want.
Granted, network difficulty and exchange prices will change over time, possibly dramatically, but as of right now the net profit is close to $16.00 per day, or roughly $8.00 for each card. So just figuring for the cards alone we would be looking at around a 100 day ROI, and factoring in the entire system it would be closer to 125 days, or about 4 months for break-even. From past experience it will probably be closer to 6 months, but we have had a good run for the past 18-months so while I do expect things to slow down again, I do not think we will go back to no profitability as in the 2014/2015 period.