I want to actually mine Bitcoin (SHA256) on modern GPUs

[NotFuzzyWarm]

The probability of a CPU ever finding a block is so small that it would be a waste of electricity since it wouldn't help secure the network.
Chances are pointless at that level.

I was told the same thing with USB mining and how much money I wasted on little USB sticks but here I am. 

Very true and against truly astronomically high odds you hit a block. 
That said, while using cpu/gpu *is* technically possible the much lower hash rate is several magnitudes lower that of a Compac-F and that much higher against ever finding a block in ones lifetime using them. But yes- it IS technically possible.

To me the main point is just get a Compac-F or the new R909 pod und forget about doing it with a cpu/gpu just because you can. Either way is very simple plug & play. Aside from an academic exercise to prove the code works and can actually mine to a node I don't see the point.

[Sledge0001]

Very true and against truly astronomically high odds you hit a block.  
That said, while using cpu/gpu *is* technically possible the much lower hash rate is several magnitudes lower that of a Compac-F and that much higher against ever finding a block in ones lifetime using them. But yes- it IS technically possible.

To me the main point is just get a Compac-F or the new R909 pod und forget about doing it with a cpu/gpu just because you can. Either way is very simple plug & play. Aside from an academic exercise to prove the code works and can actually mine to a node I don't see the point.

I can confirm that you can in fact mine BTC with CCminer for Nvidia GPU's and good old CPUMiner still works even with an old Core2Duo.

You just need to define the -a sha256d in each.

I know because I have done my own testing to see if and how it would work.

*Note* the max difficulty I saw with a 3060ti was around 7 at its peak and on the Core2Duo CPU is about .4 That allowed each to submit a share about every 2-8 seconds.



Here is the version I used with instructions for the GPU's

https://crazy-mining.org/en/mining/ccminer-how-to-install-and-use-download-and-configure-for-windows/

Code:

ccminer-x64 -a sha256d -o stratum+tcp://pooladd:port -u mybtcaddress.wokername  -p x

pause

GPU Results:

[2022-12-30 08:22:10] Stratum difficulty set to 4.37602
[2022-12-30 08:22:10] GPU #0: NVIDIA GeForce RTX 3060 Ti, 1782.21 MH/s
[2022-12-30 08:22:12] accepted: 12808/12814 (diff 8.205), 1790.72 MH/s yes!
[2022-12-30 08:22:15] GPU #0: NVIDIA GeForce RTX 3060 Ti, 1791.83 MH/s
[2022-12-30 08:22:19] accepted: 12809/12815 (diff 7.987), 1789.98 MH/s yes!
[2022-12-30 08:22:20] GPU #0: NVIDIA GeForce RTX 3060 Ti, 1764.46 MH/s
[2022-12-30 08:22:24] GPU #0: NVIDIA GeForce RTX 3060 Ti, 1794.85 MH/s
[2022-12-30 08:22:29] GPU #0: NVIDIA GeForce RTX 3060 Ti, 1793.41 MH/s
[2022-12-30 08:22:30] accepted: 12810/12816 (diff 5.804), 1789.73 MH/s yes!
[2022-12-30 08:22:31] accepted: 12811/12817 (diff 8.837), 1789.76 MH/s yes!
[2022-12-30 08:22:34] GPU #0: NVIDIA GeForce RTX 3060 Ti, 1792.61 MH/s
[2022-12-30 08:22:38] GPU #0: NVIDIA GeForce RTX 3060 Ti, 1792.04 MH/s
[2022-12-30 08:22:38] accepted: 12812/12818 (diff 6.183), 1789.74 MH/s yes!
[2022-12-30 08:22:41] accepted: 12813/12819 (diff 38.964), 1780.95 MH/s yes!



Here is where I acquired the CPU Miner:

https://github.com/tpruvot/cpuminer-multi/releases

Code:

cpuminer-gw64-core2 -a sha256d -o stratum+tcp://pooladd:port -u -u mybtcaddress.workername  -p x 

pause

CPU Results:
[2022-12-30 08:19:09] Stratum difficulty set to 0.0712169
[2022-12-30 08:19:22] CPU #1: 3303 kH/s
[2022-12-30 08:19:23] accepted: 8991/8996 (diff 0.064), 13174 kH/s yes!
[2022-12-30 08:19:28] CPU #3: 3541 kH/s
[2022-12-30 08:19:28] accepted: 8992/8997 (diff 0.132), 13216 kH/s yes!
[2022-12-30 08:19:48] CPU #0: 3147 kH/s
[2022-12-30 08:19:49] accepted: 8993/8998 (diff 0.152), 13250 kH/s yes!
[2022-12-30 08:19:50] CPU #2: 3284 kH/s
[2022-12-30 08:19:55] CPU #1: 3282 kH/s
[2022-12-30 08:19:55] accepted: 8994/8999 (diff 0.541), 13254 kH/s yes!


Never let anyone tell you that you can't. The results show you can if you want.

Is this efficient? Not really. But there is the possibility although remote to solve a block if the stars were perfectly aligned in your favor.

Success comes in cans not in cant's!!!!

[hZti]

The only real reason this would still be used is for fun or if you have somehow access to a botnet or malware and can run this for free on the victims computer. But even then it is not really profitable and also very illegal for just a small profit. Still nice that it can be done, since CPU mining is the form of mining that satoshi Nakamoto intended for bitcoin. As far as I know he never really though about asics.

[Sledge0001]

The only real reason this would still be used is for fun or if you have somehow access to a botnet or malware and can run this for free on the victims computer. But even then it is not really profitable and also very illegal for just a small profit. Still nice that it can be done, since CPU mining is the form of mining that satoshi Nakamoto intended for bitcoin. As far as I know he never really though about asics.

I wouldn't go so far to say that it has to be for illegal use, think of this case scenario.

If someone has a GPU farm that is sitting idle (post ETH2.0) and there was a pool that would allow them to participate, mine and earn BTC by sending whatever hashpower they could assuming the pool found blocks then I would assume that there could be a benefit there.

I am in full agreement with the assumption that ASICS were not in the original vision.

[kano]

The probability of a CPU ever finding a block is so small that it would be a waste of electricity since it wouldn't help secure the network.
Chances are pointless at that level.

I was told the same thing with USB mining and how much money I wasted on little USB sticks but here I am.  

Well why not try to understand the numbers and work it out for yourself then

I have explained the numbers in detail in many posts.

No doubt the other person you spoke to was useless and didn't give anything even close to the detail I have.
I do find it offensive comparing me to that since I do provide reasons and numerical details.

Success comes in cans not in cant's!!!!

Success comes in understanding - "can" is not the only requirement as you imply.

[Sledge0001]

Well why not try to understand the numbers and work it out for yourself then

I have explained the numbers in detail in many posts.

No doubt the other person you spoke to was useless and didn't give anything even close to the detail I have.
I do find it offensive comparing me to that since I do provide reasons and numerical details.

Success comes in understanding - "can" is not the only requirement as you imply.

Kano,
Make no mistake I know the numbers are against anyone that tries to mine for a block period unless they have a ton of hash-rate.

Many people did tell me and thought I was nuts for mining with USB sticks. You were not one of them.

Even with all that being said if there was enough hash and someone got lucky then it is possible.

True success comes from continually trying and eventually succeeding after many others have failed before you.

[NotFuzzyWarm]

Kano,
Make no mistake I know the numbers are against anyone that tries to mine for a block period unless they have a ton of hash-rate.
Many people did tell me and thought I was nuts for mining with USB sticks. You were not one of them.
Even with all that being said if there was enough hash and someone got lucky then it is possible.
True success comes from continually trying and eventually succeeding after many others have failed before you.

AFAIK Kano's primary objection is not the low hash rate & minuscule chance of finding a block when using CPU/GPU's - it's the power used to produce that *tiny* has rate. For running on one thread of a multi-core CPU while you are doing other things on your pc, fine. Doing it on PCs you do not own and are not authorized to use is obviously out of the question and highly illegal. Mining with multiple GPU's - well, the cost of power used adds up very fast and it's easy to put over 1kw & more into them. So unless your power is free, once you start going over 100w, just get a few sticks or a pod miner and be ahead of the game on all points...

[Sledge0001]

AFAIK Kano's primary objection is not the low hash rate & minuscule chance of finding a block when using CPU/GPU's - it's the power used to produce that *tiny* has rate. For running on one thread of a multi-core CPU while you are doing other things on your pc, fine. Doing it on PCs you do not own and are not authorized to use is obviously out of the question and highly illegal. Mining with multiple GPU's - well, the cost of power used adds up very fast and it's easy to put over 1kw & more into them. So unless your power is free, once you start going over 100w, just get a few sticks or a pod miner and be ahead of the game on all points...

@NotFuzzyWarm
To be clear I am (as we all should be) 100% against any illegal activities.

Hacking or using someone elses resources / hash rate or energy is theft. Make no mistake it should not be tolerated.

All I am saying is there are a ton of big GPU farms out there making pennies on shit coins...

If they were to unite and push them to a pool that allowed them to pool together and use their hash / resources one might be surprised at the outcome.

It can be done and again going back to the original concept 1CPU (or GPU in this case) = 1Vote.

The vision is alive!!!

[kano]

...
All I am saying is there are a ton of big GPU farms out there making pennies on shit coins...

If they were to unite and push them to a pool that allowed them to pool together and use their hash / resources one might be surprised at the outcome.
...

They wont even make pennies on bitcoin.

The power requirements of a GPU are thousands of times the return.

Be very clear, you cannot not get megawatts of power for nothing.
It's not possible without stealing it.

[Sledge0001]

...
All I am saying is there are a ton of big GPU farms out there making pennies on shit coins...

If they were to unite and push them to a pool that allowed them to pool together and use their hash / resources one might be surprised at the outcome.
...

They wont even make pennies on bitcoin.

The power requirements of a GPU are thousands of times the return.

Be very clear, you cannot not get megawatts of power for nothing.
It's not possible without stealing it.

@Kano,

Come on.... You can't say that with certainty.

Your statement seems to classify or refer to anyone who wants to take a chance with a CPU or GPU a criminal.  

I for one am not buying it.  

Regardless of ones energy use, who am I to stop someone from trying....

Let people mine with what they have.

The people want to be free!!!! 

[kano]

...
All I am saying is there are a ton of big GPU farms out there making pennies on shit coins...

If they were to unite and push them to a pool that allowed them to pool together and use their hash / resources one might be surprised at the outcome.
...

They wont even make pennies on bitcoin.

The power requirements of a GPU are thousands of times the return.

Be very clear, you cannot not get megawatts of power for nothing.
It's not possible without stealing it.

@Kano,

Come on.... You can't say that with certainty.

Your statement seems to classify or refer to anyone who wants to take a chance with a CPU or GPU a criminal.  

I for one am not buying it.  

Regardless of ones energy use, who am I to stop someone from trying....

Let people mine with what they have.

The people want to be free!!!! 

I am not selling anything. So there is nothing to buy.

Everything in my reply is valid and also an under estimate.

This discussion has nothing to do with freedom.
It's to do with doing something abhorrently bad.

Either an abhorrent waste of power, or stolen power.

e.g. a 300W ~1GH/s GPU will mine about 100000 times slower than a 3.3kW ~100TH miner.

So some tiny guy with ONLY 11 GPUs is using 3.3kW, ~11GH/s to PPS mine less than 0.00000004 BTC a day (4 satoshis)

That's really simple to see how bad an idea that is.
Alas I've said all this before so I guess you're just trying to harass.
I'll stop replying.

[Sledge0001]

I am not selling anything. So there is nothing to buy.

Everything in my reply is valid and also an under estimate.

This discussion has nothing to do with freedom.
It's to do with doing something abhorrently bad.

Either an abhorrent waste of power, or stolen power.

e.g. a 300W ~1GH/s GPU will mine about 100000 times slower than a 3.3kW ~100TH miner.

So some tiny guy with ONLY 11 GPUs is using 3.3kW, ~11GH/s to PPS mine less than 0.00000004 BTC a day (4 satoshis)

That's really simple to see how bad an idea that is.
Alas I've said all this before so I guess you're just trying to harass.
I'll stop replying.

Perfect your numbers are based upon the 4-5 bucks a day you can make with a 100TH miner assuming that your mining PPS at under 8 cents a KW/h

It has no bearing on a lucky solo miner who every well could disprove your logic with the most minimal of hashrate.

Not harassing just speaking based on the numbers. After all isn't that your basis? 

It proves that assuming someone gets really lucky (even astronomically... ahem) there are possibilities regardless of hashrate.

Is the cost per KW/h worth it? That shouldn't be a decision that one should make for others.

[NotFuzzyWarm]

Is the cost per KW/h worth it? That shouldn't be a decision that one should make for others.

No one is making that decision for others.
What we are doing is emphatically pointing out definite costs vs *possible* income and the fact that there are lower power cost options (sticks & pods) that also happen to be far faster as well as using far less power.

[Sledge0001]

No one is making that decision for others.
What we are doing is emphatically pointing out definite costs vs *possible* income and the fact that there are lower power cost options (sticks & pods) that also happen to be far faster as well as using far less power.

One can easily read through the posts above and plainly see how some are blatantly discouraging people to try or take steps to be inventive.

Sorry I don't and absolutely refuse to subscribe to that logic.

Imagine if someone re-wrote older code or created new code for a GPU / CPU miner program that was able to improve the efficiency drastically of GPU's/ CPU's to compete with smaller Asics. Then what would you say?

Impossible? I don't believe so. Does such code exist currently? No...  But one doesn't know for certain what the future will bring.

Currently the energy use of a GPU / CPU when mining BTC can not compete with an ASIC. This is fact.

So for now, we can all agree indisputably that even having the most powerful GPU compete head to head with a 110TH/134THUnit is bringing a knife to a gunfight.

Regardless now matter how improbable it maybe now, I personally would never discourage folks to try what they want to try.

Why would anyone go to such lengths to stunt creativity and possible invention? That is the bigger question.

So, I'll stand by my beliefs (based on this thread I am not alone) that everyone should choose their own route and try what they see fit to learn, mine and be creative.

 

[kano]

...
One can easily read through the posts above and plainly see how some are blatantly discouraging people to try or take steps to be inventive.

Sorry I don't and absolutely refuse to subscribe to that logic.

Imagine if someone re-wrote older code or created new code for a GPU / CPU miner program that was able to improve the efficiency drastically of GPU's/ CPU's to compete with smaller Asics. Then what would you say?

Impossible? I don't believe so. Does such code exist currently? No...  But one doesn't know for certain what the future will bring.
...

Changing the subject to alchemy and magic doesn't help.

Clearly you don't understand why a GPU wont do that.

---

Well then look at it a different way
If someone succeeds in doing that with a GPU then BTC will become worthless.
That's a certainty.

It's not possible to do that by magically writing better code, it would only be possibly by finding flaws in sha256.
Thus you can say goodbye to your BTC.

Also, if so then you can make a 100PH 3kW ASIC to blow away the GPU miners yet again.
But no one would do that coz BTC would already be worthless, so not worth spending 100's of millions on chip fabrication.

Yes you can pretend that everything is possible and that you can jump off a 10 story building and float to the ground with nothing.
Alas reality does have some very specific rules

[Sledge0001]

Changing the subject to alchemy and magic doesn't help.

Clearly you don't understand why a GPU wont do that.

---

Well then look at it a different way
If someone succeeds in doing that with a GPU then BTC will become worthless.
That's a certainty.

It's not possible to do that by magically writing better code, it would only be possibly by finding flaws in sha256.
Thus you can say goodbye to your BTC.

Also, if so then you can make a 100PH 3kW ASIC to blow away the GPU miners yet again.
But no one would do that coz BTC would already be worthless, so not worth spending 100's of millions on chip fabrication.

Yes you can pretend that everything is possible and that you can jump off a 10 story building and float to the ground with nothing.
Alas reality does have some very specific rules

Ahh Kano,

BTC isn't worthless and most likely won't ever be.... I mean I hit with a USB stick come on...

What would you have said if it would have hit off a GPU/CPU?

In any event I'm a buyer at a 100PH 3kW ASIC if you fab and code it up.

Regardless,  I do love your pessimism. Although I still don't share the same fundamental logic.

[BitcoinSoloMiner]

90Watts with 1.4GH/s

on software which is 9 years old  

i wonder if this could be improved

for example on new software which utilises AMD resizable bar

[kano]

...
Regardless,  I do love your pessimism. Although I still don't share the same fundamental logic.

It's not pessimism. It's fact.

A CPU, GPU, FPGA and ASIC miner all do the same VERY simple hash to attempt to brute force a number of leading zeros.

A GPU isn't going to magically be able to do that 10 times faster unless CPUs, FPGAs and ASICs all can do it 10 times faster.
As I already said, if they find a flaw in sha256 and make the GPU faster, that would work the same with all the others.
There's nothing special about a GPU that means you can improve it's performance and not be able to do the same to the others.
But if it's a flaw in sha256, it would suddenly mean that bitcoin would have a major problem, as will all the other world wide systems that use sha256.

There's already 3 speed ups.
About 3% by stopping when you have H=0 since you don't need the final A-G.
There also 2 forms of asicboost that gain about 10% using one or the other.
These aren't 50% improvements or even 25% improvements.
They're also very specific how they work and very simple to understand why.
... heck I even discovered the 3% H=0 zero myself less than a month after I came to the forum, in aug-2011, you can find my posts about it, alas someone had already worked it out well before that, and they were using it in mining, just I didn't know until I asked questions about what I worked out.

This whole GPU issue has come up coz there's all these people who spent money mining a scam coin and the scam coin has effectively scammed them.
They can't mine it any more - OMG there's gotta be something else they can do.
Oh well sux to be them. Yep go screw around with another scam coin - bye bye.
But, but, but, they gotta be able to mine Bitcoin on their GPUs coz they're desperate to believe it's possible even if it's been clearly know for years it's a waste of time.
Heck the rules for this section of the forum even state that.

[BitcoinSoloMiner]

Discussion of software for education purposes to understand mining related functionality can go here.

I think with hardware advancements to GPU, since asics took over, means that GPU efficiency is more than 3/25/50% if the software was advanced.

PCIe gen 4/5, more clock speeds and more memory on GPU these days, there must be software that can be created to improve on outdated CGMiner software and utilise these advancements

hashcat is a good example, its up-to-date software produces sha256 hashing rates of at least 3 times the CGMiner hashrates

i get 1.4GH/s on RX6600 on cgminer

---------------------------
* Hash-Mode 1400 (SHA2-256)
---------------------------

Speed.#1.........:  4219.1 MH/s (63.05ms) @ Accel:64 Loops:1024 Thr:256 Vec:1

thats 1.4GH/s vs 4.2GH/s for RX6600

Looking at the new 4090 hashrates for hashcat:

---------------------------
* Hash-Mode 1400 (SHA2-256)
---------------------------

Speed.#1.........: 21975.5 MH/s (48.66ms) @ Accel:32 Loops:512 Thr:512 Vec:1

thats 22GH/s!

For everyday people who are not willing to sacrifice their ears to have an asic in their home or room, they must be able to lottery mine BITCOIN (not some altcoin) with a GPU while their PCs are idle, this is a massive untapped market/area

look at how old S9s are being used as space heaters with 1/3 hashboards, fan modding and over/under-clocks

i think the bitcoin network deserves the hashrates provided by people with GPU who want to lottery mine such as others, OP and myself, they deserve an updated miner!

pandering to the hype/sensation from news and policymakers about how bitcoin is using alot of energy and destroying the world i think is not in the spirit of this community and its founder

this rhetoric of wasted electricity using CPU/GPU on bitcoin mining is tiring and as sledge said it is stunting creativity and possible invention

insulting and even hateful comments have no merit here

[kano]

I have already explained directly why those numbers are wrong.
It takes 3 hashes to mine, and 2 hashes for each nonce if you generate midstates.
I guess it slipped your mind ...

... and if you have a newer GPU that runs 5 times faster due to faster hardware and more cores,
than an older GPU, of course it will mine 5 times faster.

However, a 300W GPU mining full speed is not silent and still using 300W
without matching even a slow old cheap USB gekko miner in hash rate and using 20 times the power of that slow old USB gekko miner.

These GPUs also cost a lot of money - more than a 100TH/s miner.

This discussion about wanting to mine with a GPU has nothing to do with education,
other than for those who might learn how bad an idea it is.

The title says it's got nothing to do with education:
"I want to actually mine Bitcoin (SHA256) on modern GPUs"