Lack of fallover server support makes this miner UNUSABLE in a production environment.
Fixing that should have been a priority a WHILE ago.


 Did some comparison testing on my 1070 ti rig between 0.5.5 and EWBF 0.3.4b yesterday - what I found was that DSTM gave a little higher hashrate when the cards were being run at a low power setting (under 70%, or 124 watts) but at that 124 watt point the hashrate was almost identical and EWBF started giving better hashrate above that point.
 The differences were small though, and DSTM seems to vary hashrate a lot more which makes it hard to do a good comparison.

 I also dislike the fixed 2% "developer contribution" level, MORE than makes up for the hashrate differences I've seen vs EWBF and EWBF lets YOU set the contribution rate (no, I don't use 0 - but 2% IMO is a RIPOFF when other software is so close on performance).

EVGA G2 850 would be a good deal at that price, but the GQ line uses one of those junk "fancy name" sleeve bearing fans instead of a good Ball Bearing fan like the G2 series uses.

Without CUDA support, there is ZERO ability of ANY miner to use the card for anything.
That's what CUDA is all about.

 There might be older CPU-based mining software for SOMETHING (most likely Monero) that supports a non-AES CPU.

Where the Tesla shines vs NVidia "consumer" cards is in Double Precision Floating Point operations - this specifically includes the current-generation Titans, not just the GTX 10xx series.
Cryptocoin usage is a waste of the card, as cryptocoin is all Integer operations.

They're intended for usage in specific server cases with a lot of airflow through the case - thus the "no fan" passive cooling setup.
A lot of AMD workstation-specific cards in the FirePro line are set up the same way.

Try running it on the BOINC Milkyway project for Gridcoin earnings sometime - it SHOULD blow away the R9 280x / HD 7970 cards that most of THOSE folks use in their "high performance machines" for similar reasons (the Tahiti GPU was AMD's highest performing "consumer" DP FP chip).

 220 and 240 are generally used interchangeably when talking about US power - the "nominal" value for US split-phase is 117/234 volts, but ACTUAL voltage at the outlet varies widely even excluding "brownout" conditions.

 Vega is a VERY good and efficient Monero (XMR) miner, it's only so-so at best directly mining ETH (more than twice the cost and ballpark twice the power draw of the RX 470/480/570/580 cards for well UNDER twice the hashrate).
 Since AMD RX 5xx cards have dropped semi-sorta-close to MSRP level pricing in the last couple weeks, there is no reason to even LOOK at anything else on ETH mining (unless you can get a REAL GOOD deal on a used R9 290/290x or the like).


 Gold-rated power supplies in general offer ballpark 2% higher efficiency when run from 220 instead of 110 - the exact amount varies a bit depending on the load ON the PS and the specific PS, but at 50% plus load values 2% is a good "ballpark" figure for all of them.

For the SAME price, I'd go with the 1080.

 Almost identical hashrate at the same power level for almost identical efficiency, but better residual value and can be pushed to a hair higher hashrate if you choose to do so (but NOT a good idea on the Zotac Mini, those coolers are a bit marginal for a 1070 ti OR a 1080).

 I normally buy the 1070 ti though because they NORMALLY cost about 10% less than a "same model" 1080 while being almost identical capability (except if you push them HARD then they "only" manage 95% or so at the HIGH end of the TDP range).

 The 1070 ti boosts the GPU a hair higher than the 1080 when they have the same power available, since the 1080 has to split the power between more cores.
 This seems to make up for, or slightly MORE than make up for, the higher core count the 1080 has.
 It's a SMALL difference either way though.

 VoskCoin was using DSTM to get to 4.7 efficiency, not EBWF - the testing I did that he refers to in that video and his thread used EBWF and was only gettting to about 4.5ish at best- and my more recent "fine tuning" testing didn't give substantially more, that 106 watt TDP point was very very close to the optimal (104 watts in my more recent testing is good for a HAIR more but close enough to be within the limits of test measurement).

 I DID do some testing with DSTM yesterday on my 1070 ti rig - slight improvement over EBWF at the "efficient" point, but a lot more variable - and it showed no improvement or a LOSS compared to EBWF past about 124 watts (70% TDP).

 Also keep in mind that the 1070 ti uses GDDR 5 vs the 1080 using GDDR 5x - which doesn't seem to affect equihash a LOT but might have some small effect.
 Contrary to often stated opinion, memory clock DOES affect equihash noticeably on most NVidia cards it's commonly mined on - and the latency of 5x vs it's higher bandwidth can sometimes be an issue as ETH miners know well.

I would go with a Sapphire over ANY of those 3, but I concede I don't have a lot of recent experience with ASUS cards (I do have one of their "Turbo" 1080 ti blower cards).

 XFX - I stopped having anything to do with them a while ago, WAY too many "die young" issues.

 Gigabyte - I might THINK about an Aorus model, the rest of their line went "cheap junk" on the fans a while back and I refuse to put up with them any more.
 The specific Gigabyte card listed uses the SAME fans as several of my 1070 models - fans that have been DYING IN ABOUT A YEAR in my experience with them. BAD IDEA.
 They cool well - when they work - just don't plan on them WORKING for very long.

 Dual BIOS isn't a big deal - but high quality BALL BEARING fans on a card you are going to use in a high-temp high-workload 24/7 environment is CRITICAL, and everything Sapphire I have ever owned has had a long-lasting BALL BEARING fan on it.


 The only cards that specify "memory type" that I have seen in recent years is the Sapphire "mining specific" RX 470 line - but those have not dropped in price to match recent price drops on various RX 570/580 models so I can't recommend them any more.

He better be a BIG investor if he wants to play in the ASIC mining world - he's looking at needing to hire a team of VERY VERY GOOD chip designers, put out MILLIONS in working with a Fab to create test chips, lots more in testing the chips to make sure they meet the target specs, then millions more to go into production of the chips - THEN he gets to start paying to put together the miner design and manufacture side of the business.

 There are other threads that have ALREADY gone into this.

For XMR, the Vega blows both cards away badly while costing quite a bit less (if you can FIND one right now) than the 1080 ti - 1900 XMR hash tends to be a hair on the LOW side for the Vega RX 56 when configured correctly, compared to under 1000 for the 2 cards you list.
 Down side - you HAVE to run Windows 10 to get that kind of hashrate due to driver issues, any other driver/os setup you are lucky to get past 1300.

 For ETH, the R9 295x2 is a dual R9 290x on one board - cooling is marginal compared to the actual 290x, but it should be capable of mining ETH at somewhere between 50 and 60 MH/s vs the 1080 ti (WHICH IS NOT A GOOD ETH MINING CARD barely managing 35ish at most for WAY too much money), but does eat a lot more power to do so.

 IMO you should be looking at other choices if you really want to mine ETH or XMR - and on the 1080 ti you should be looking at algorithms it is GOOD at, like equihash (ZEC/ZEN and such) or lyraR2v2 and the other "traditional" NVidia algorithm strongholds.

 ETH might be the best choice for the R9 295x2 but a pair of RX 470/570/480/580 cards can nearly match the hashrate for WAY LESS THAN HALF THE POWER DRAW and run a lot cooler while doing so.
 If you can find that R9 295x2 cheap enough and your power cost is cheap enough, though, it MIGHT be worth getting.

 Also keep in mind that the R9 295x2 has been out of production for YEARS and is currently a full generation behind state-of-the-art for AMD, while the 1080 ti and Vega ARE current generation.

Memory DOES make a difference on equihash - though not a big one on the 1080 ti.

 On my Aorus (not the extreme), I found that +100 core +100 memory was the best setting for speed/efficiency that was stable - +150 memory gave a few more sol/s but wasn't long-term stable as those cards seem to push the memory pretty hard already, +200 memory gave me LOWER sol/s probably because it pushed the card up into the next "strap" range with higher latency settings.

 50% TDP (125 watts) gave 505 hash - don't bother this is a waste of the card.
 60% TDP (150 watts) gave 620 hash - this seems to be the best efficiency point that still has decent hashrate.
 70% TDP (175 watts) gave 680 hash - efficiency dropping but not a LOT yet
 80% TDP (200 watts) gave 725 hash - efficiency dropoff getting noticeable
 90% TDP (225 watts) gave 760 hash - ditto
100% TDP (250 watts) gave 780 hash - waste of power unless you have VERY VERY cheap power, and even then it's a bit iffy - though the cards still runs pretty cool at this setting as long as it's NOT in a hot room.


 Other cards may vary, though all of the cards I have to date (1 Gigabyte Windforce, 1 EVGA SC, pair of Aorus) all seem to match nearly identically up to about the 80% point, then the EVGA and Windforce start dropping a bit as temp doesn't let them boost the core as much.

Zen and ZCash use the same algorithm, so they should use the same settings.

 For efficiency, 106 watts give or take a couple seems to be the most efficient point for that algorithm - aim for +200 core +700 memory, per the testing I've done on my EVGA SC 1070 ti cards.

 I should have a Zotac Mini on hand Tuesday (long shot Monday) to see if it works well with the same settings.

 (edit) Zotac mini arrived this morning, has gone for 8ish hours so far at the same settings no issues and same performance.

 What do you mean by "run them as one unit"?

 Technically, burst doesn't have direct competition.

 Storj, Sia, and maid (can't remember exact name on that one) are targeting "rented storage" not pure "mining from the HD" and are more competition for each other, as I understand those coins.

 Doesn't seem to be one for a machine that is used for other things that grab the video card.

 Folding At Home.

 At this point, given how many issues I've had with Gigabyte cards the last 3 weeks, I can't recommend Gigabyte PERIOD (except perhaps the Aorus, jury is still out on those).

 3 cards with dead fans, 1 with *2* dead fans, and I've had *2* of the ITX cards just flat out DIE this past week.



 BTW - the "common" 1070 ti has a 29 watt higher TDP than a "common" 1070, why is it a supprise that it would run hotter?

I'd be inclined to run those Tesla units on the BOINC "Milkyway" project to earn Gridcoin with.
Same thing if those are the ORIGINAL Kepler-based Titans.

Those cards are not real good at mining cryptocoins (which use all integer operations) but they have VERY HIGH double precision floating point performance which is what the Milkeyway project needs for good performance.

When I was mining ETH with my R9 290 cards (all Sapphire Reference design, with a bios from TheStilt), they were generally pushing just a hair under 30 Mh/s running 1100 core / 1250 memory as I recall - under LINUX so the only undervolt was out of the modded BIOS.
No idea what the "aux" setting you reference is, and TDP at stock.

 It's also dry, so inexpensive evaporative cooling would work during the hot parts of the days.

 I don't have anything that does wifi, never have and probably never will.
 Not willing to put up with the much lower reliability vs Ethernet, and I don't waste money on laptops when I don't travel enough for them to make sense.