Gridcoin [GRC] Mining Profitability in USD - NVIDIA GPU Mining (Up to USD$103/m per Card at ATH on Gamer Card!)

in gridcoin •  7 years ago  (edited)

One of the most common questions I get asked in the Gridcoin forums and social channels is: 'How much can I make with hardware X'? The difficulty in answering this question is that Gridcoin does not function like a conventional cryptocurrency with respect to mining. Your hardware is not put to use to endlessly hash data, but contributes to actual science which is less predictable with respect to performance.

Just like mining other crypto, with Gridcoin:

  • The amount you earn depends on how much you contribute to the network compute relative to everyone else.
  • You can join a pool to get paid more steadily, rather than having to get lucky staking a block.
  • More powerful hardware outcompetes less powerful hardware generally speaking.
  • GPUs outcompete CPUs.

However, unlike mining other crypto, with Gridcoin:

  • Your compute power is used to solve research problems, primarily for the scientific community. These problems are complex, and vary wildly between projects that the Gridcoin Network rewards contributions on. As a result, hardware performance will vary between projects, as will your mining rate.
  • Hashing power is meaningless, so hardware profitability cannot be readily compared
  • The daily Gridcoin mint is spread across the 'whitelisted projects', which are projects eligible for rewards (you run these projects - you get paid GRC). If the number of projects in the whitelist changes, this will significantly impact your daily GRC mint.

As a result of the above, the best way to maximise your yields is to contribute compute to the project that has the least competition from other Team Gridcoin members. I have been running a significant number of different GPUs on several of the most efficient projects to mine for the past month and a half, and collected data on the income they have generated. By calculating the average yield per GFLOP of compute and linearly extrapolating, we can predict the earnings of most GeForce cards in the series. Cards highlighted in yellow are cards for which I have actual data, which were used to linearly interpolate the rest of each series. Note that all values are averages, meaning none of the raw data remains in these tables. Further, all hardware is running stock settings - no overclocking or overvolting, which could further increase yield.


NVIDIA GeForce 10 Series


NVIDIA GeForce 900 Series

We will be skipping the GeForce 800 series as this only includes mobile models. This was because NVIDIA intended for the 800 series to use Maxwell microarchitecture but had already released 800 series mobile GPUs with Kepler architecture. Therefore the intended desktop 800 series GPUs were released as the NVIDIA GeForce 900 series instead.


NVIDIA GeForce 700 Series

Please note that some liberties have been taken in making these estimates:

  • I have not taken into account power costs as power prices vary significantly depending on where you live.
  • I have assumed that the GPU will be running 24/7
  • I have not optimised CUDA vs OpenCL (could yield about a 10% increase, depending on the project)
  • I have not considered FP64 performance, only FP32 (FP64 is relevant when running MilkyWay@Home)
  • These estimates are for the current GRC price (USD$0.032), and ATH (USD$0.13 in June). You will have to adjust the income estimates based on markets rates at the time of reading.

If you are not sure what GPU you have in your machine, you can find out in less than a minute! Hit the start button, and type 'run'. Open the run dialogue box by pressing enter and type 'dxdiag' without the speech marks. If asked whether or not you would like to check if your drivers are digitally signed, select 'no'. In the window that opens up, navigate to the 'Display 1' tab to find your GPU. Note that on my computer shown below, I am running an ancient Quadro 600, which was not listed above and has pretty abysmal performance.

If you would like to get involved, you can get started by installing the research software BOINC and the Gridcoin Wallet.


Image credit, in order of appearance:
Banner, @joshoeah
Tables done by me based off my own mining data
Banner, @vortac

HighQuality.gif

Authors get paid when people like you upvote their post.
If you enjoyed what you read here, create your account today and start earning FREE STEEM!
Sort Order:  

It's about damn time someone put something like this together = ).

To add to it:

Many BOINC projects run through CPUs only. This means that if you are a short term GPU miner and see better short term profits for your GPUs elsewhere, pointing your CPUs at some CPU specific BOINC projects and earning some GRC on the side is a distinct possibility.

Of course I'd advocate for pointing some GPUs at GRC as well because long term-- well let's just say the potential and dedication of Gridcoin and its community is clear.

  ·  7 years ago (edited)

A really good point @jringo! In fact, CPU mining yields roughly the same profits to GPU mining when comparing hardware from similar tiers and release years. If there is interest, I can repeat the process with CPU data.

Was putting it off because there is a lot more raw data to process for the CPUs.

@jringo actually beat me to it @dutch, I think there is a lot of competition in terms of crypto that can be GPU mined but hardly any where it's feasible to CPU mine. It would be great to show GPU miners they have an additional revenue stream they are missing out on.

Alright. I have started collating CPU data. =)

I think @scalextrix is right: target the CPUs of GPU miners. If you want a hand doing some grunt work data fetching let me know = )

I have just finished analysing my own data sets, and am open to accepting data from other crunchers. However, it is hard to verify data from other people to see if, for example, the machine was actually run 24/7.

seems like you're going to have to rely on the law of averages

Hello @dutch - could BOINK (I hope that I used the correct term) be used to search for a molecule that could activate telomerase and extend telomeres in human cells / organisms?

following you sir.

if I understand correctly, if you have a data set to test, yes! in the same fashion as WCG when they found a more efficient chemical set for solar panels:

https://www.greenoptimistic.com/boinc-solar-cell-design-system-20081210/

https://phys.org/news/2011-09-millions-molecules-screened-ideal-solar.html

It most definitely could, seeing as telomerase shape and function is relatively well understood. BOINC already processes molecular interactions on smaller scales, so I see no reason it cannot test for interactions with more complex compounds.

There are several exhaustive molecular interaction studies already being done.

  ·  7 years ago (edited)

let us know when you find the immortal molecular interaction pls

After following #gridcoin for awhile now, I am super excited to have become a token holder now. I bought some a few days back. #value! You are in inverse of your competition which has a huge market cap and no functionality. Gridcoin is all Steak and with a little Sizzle... the investors should show up soon.

  ·  7 years ago (edited)

Exactly! Golem has only one colossal advantage over us right now, which is that their marketing is spot on. That is about to change though, and Gridcoin rules the field in every other respect. We have a working product, that already allows compute to be purchased, and is built on a distributed research platform with 2 decades of development.

Brace yourself as GRC takes off for the moon in the next few months! =)

You deserve a resteem

Thank you! I appreciate that a lot. =)

Great job, thanks for all the calculations.

Currently on nicehash my gf970 is making $32 a month (undervolted) so it's quite on pair what we're seeing here. Would be cool to mine and accumulate some gridcoins in case of future price increase.

Anyway, following :)

That's really interesting, thanks for sharing! =)

I should probably add to the article that everything is running at stock, and could be overclocked and overvolted to increase raw yield.

Thanks for this - I have 3rigs each with 6 GX1060's which are currently mining Ethereum and I was about to point them at Zcash but I'll try with one rig on Gridcoin and compare results.

  ·  7 years ago (edited)

Great! It's probably worth pointing out that with Gridcoin, your daily mint grows over the course of the first month. Most people see their mint is low on day one and give up because they do not realise this.

To offset this, your daily mint will also tend back to zero over the course of about a month after you cease 'mining'. In other words, you will continue to be paid for weeks after you stop contributing.

thanks for that info

Dont forget you can profitably mine on your CPU too, just choose a project that does not use GPUs

Great post, just finished setting up Gridcoin and Boinc, just synced and starting to work on that Moo! Wrapper Project

Sheesh @dutch, how did I miss this post....

This post received a 1.4% upvote from @randowhale thanks to @dutch! For more information, click here!

staking computer GRC.png

Hey guys! Just wanted to share this. You can Earn while you Trade or Hodl GRC at Btcpop.co!

If there are any lazy stakers out their Btcpop stakes deposits for you and gives you proportionate rewards

These kinds of posts are very helpful when considering builds and BOINC projects. Thanks for this.

Very good information. Please introduce websites about this calculator. Thanks a lot for useful information.

The difficulty there is that there are a lot more input variables than mining other coins. I will have a crack at writing a dynamic calculator though, some time down the line.

Yes. I understand it. I read a few about this coin. I think it will grow up in future.it's platform is interesting and useful. It is using cpu,gpu & memory and maybe other processors to computing projects. But I can't find any information about profit of variables. Calculators help us to understand range of profit, not absolute profit so we need some calculators then start it.🤔

  ·  7 years ago (edited)

Cool, I love seeing data tables. Thanks for putting in the effort.

Semantic criticism: This data is about revenue, not profitability. Profitability data would account for initial costs as well as operating costs and then could be used to generate the ROI data and payback period, which is what you're after if you're hunting for the most profitable outliers.

Even if you're using someone else's hardware and not even paying for electricity, those costs do exist (for someone) and would exist for anyone looking to set up their own mining rig.

This data is nice and maybe there are some nuggets of useful information, but it seems to boil down to this: "more gflops = more grc", which isn't exactly illuminating.

I think you can take this a lot further. Maybe start by adding the wattage per card. Then you could have a function like "GFlops Per Watt", which might be helpful in itself. Add in the market price of each card, add in the rest of the rig costs (case/psu/mboard/ram/etc...), and a few columns for low/mid/high energy cost scenarios. At that point we can compare payback periods and actually be able to recommend things based on a new user's particular situation (money available to spend, their own power cost, etc...).

This is a good start though as a guide for existing hardware and users asking "How much can I make with hardware X".

Hey, thank you for the feedback! Always appreciated.

I agree that the data is closer to revenue than profit, but really it is neither due to the fact I have not taken into account exchange fees and several other small factors. I suppose the final number in USD is a tiny bit higher than the maximum profit.

The reason for the tables, which 100% do boil down to "more gflops = more grc" is that I found many users have no idea how to find their GPU specs. It takes enough effort to guide them into finding the make and model, so the intention here was to offload answering the profitability question onto this article.

I am open to the suggestion of adding a lot more complexity, but worry that it will once again result in new members to the Gridcoin Network deciding it is too much effort to read.

If I gather enough data, I may write a calculator with your suggestions implemented.

  ·  7 years ago (edited)

A much needed article for a question many newcomers keep asking. Thanks for taking the time to write this. Resteemed, upvoted and bookmarked for future use :)

  ·  7 years ago (edited)

Aren't there a bit more variables to it , thus these numbers are not a definitive factor? Such as the number of Gridcoin team members in a project and the number of hosts machines let alone cpu/gpu's? Thus " Max MAG " is just an " estimated value " along with " Max RAC ". Also different projects as mentioned about opencl vs cuda when you do the CPU chart , there are different extensions such as sse-sse4 & fgrpsse , fvp , avx , neon , pie and the list can go on so it also depends on the what the projects application is compiled to take advantage of in much the same as opencl/cuda etc. With that said , this is a great guide for beginners and a wonderful tool for them to assist in finding what is the best project for their machine when getting into earning Gridcoin. There are other factors that make it great that there is no way to make a " THIS IS THE ONLY WAY " and it would be nice if others could and would post their own experiences especially if you have access to a variety of hardware both GPU and CPU's. Possible additional information to log is the number of team Gridcoin members in the project at that time , how many computer systems team Gridcoin has crunching on that project at that time and Gridcoin's RAC/MAG percentage in that project at the time of logging. Unfortunately the factors are fluid and the output is always changing so its only a snapshot per MAX or MIN anything at that time. Boinc & Gridcoin together is like the New York stock exchange , what projects do you want to invest into?!

I agree 100% - the aim with these tables was to give a very quick, rough answer to the age old profitability question. If I want to add all the extra complexities that you mentioned, as well as those I listed in the original post, it would have to be presented as a calculator. The tables would be far too complex, or multidimensional.

If I get a couple of days off, I would be open to writing a calculator in Python or similar.

Regarding more data, the difficulty is that I do not know the accuracy of data submitted by other users. It is, for example, impossible for me to verify if they have actually been running their hardware on the same project 24/7 for a full month.

Very good post, to add to it, i see that most pc's have one gpu meaning one project at a time, i get around 150,000 credits a day with my 6 core, so i believe it would be more profitable to buy a lets say 32 or more core pc instaed of a 1080 gtx, as you could work on more projects at once, if a workstation can run boinc, id like to get to that point someday XD

You'd be right - 32 logical CPU cores can easily outperform a single 1080. I am currently analyzing my data on CPU profitability and will share that in the near future too.

Great post man. I am also Preparing similar calculations for today after prize change.

I'm confused, what is ATH?

  ·  7 years ago (edited)

All Time High. It was the peak price GRC reached a month ago. We are already well on our way back to that, after dropping with the rest of the market for the last few weeks.

Ah, thanks, yeah I see GRC as being massively undervalued. The fact that its POW at its core demonstrates real world work done to help humanity far outshines all the useless puzzle solving/energy wasting crypto-currencies like bitcoin.

@dutchoh man..
We need more people like you on this channel. Damn cryptos.
I initially found you using my machine. Are you interested in presenting more technical stuff here, in the future?

Wrong link? Or am I missing something?

I notice he's posted this same message elsewhere. Could be a bot? Or just copypasta.

Probably bot or spammer, yeah. The reply was very fast after posting.

  ·  7 years ago Reveal Comment

..... no?

please vote me

Please be on topic instead

you mean flag you?

hahahahhahah

please