Version 2 of GMiner versus T-rex miner, actually mining real Eth on 3070s

[WanderingPhilospher]

Ok, I am running a GMiner versus T-rex miner. Version 2.0.
I am running the latest releases of each miner.

@Didz ran a test and you can see the results here:

https://bitcointalk.org/index.php?topic=5345767.0

I think he did a good job and his post was insightful. There were some different suggestions in some of the comments so I decided to implement a few of those.

So here are the test parameters:

2 x RTX 3070s running on GMiner
2 X RTX 3070s running on T-rex

All 4 cards are the same exact brand/make/model
All cards have same settings
All cards are running on the same rig, so every card has the same CPU, RAM, PSU, etc.
Only difference is they are on different PCIE slots (obviously)
All cards are running on a real-time, mining eth pool

Why only 4 cards? I wanted the test to be fair and run simultaneously, in the same pool, on the exact same cards. I didn't want different times or network hash
or network difficulty to skew the data.
These 4 cards are the only ones I have that are the exact same brand/make/model.


I did read the test that 2 miners did; found here:

https://2miners.com/blog/best-ethereum-mining-software-for-nvidia-and-amd/

From what I gather, they set up a test pool with low difficulty to run their tests. Which is fine, it does show data/results.

But I am not interested in what the miners show as hash rate, nor am I interested in what the pool shows as the hash rate.
And I am not interested in how many shares each miner submitted because, as we know, each share is weighted differently. A share
of 100 G outweighs a share of 10 G.
But what I am interested in, is the bottom line, how much eth/money did the miners mine!

So the test is running live now. I don't know how long I will let the miners run, but for awhile.
Right now GMiner is running GPUs 0 and 1 and T-rex is running GPUs 3 and 4.
I do plan on switching the miners GPUs up ever so many hours so each miner mined with same cards and PCIE slots.
EDIT: swapping every 6 hours

Comments and suggestions on anything to change with the tests are welcomed.

You can watch/check in on the test in real time (both with the undercover aliases as 007s):

GMiner stats:

https://eth.2miners.com/account/0x00718dd53e4d73fc0fa3ed7935c887a62d7f84ed

T-rex stats:

https://eth.2miners.com/account/0x007231da7f49a3fed2100e6e743f06be45fcec89

Just for your knowledge/awareness; each miner is showing all cards running at 58 MH/s; so each miner is showing total hash rate of 116 MH/s.
You can use that when you look at what the pool is showing the hash rate of each miner at. But again, show me the money!

I'll keep track of how many total shares each miner submitted to the pool, just for those who are curious. But (last time), show me the money,
that's what I am more interested in!

[666mrga999]

Good experiment.
I believe it is sad that old ethminer outperforms "tuned and optimized" miners according to 2miners tests.

[WanderingPhilospher]

Good experiment.
I believe it is sad that old ethminer outperforms "tuned and optimized" miners according to 2miners tests.

True!  I think I will do another test once this one is over.  Whichever miner makes more money during this test, I will run that miner against ethminer to see if the no dev/dev fee really does play into earnings. ethminer has no dev fee while GMiner has .65 and T-rex has 1.0. That will be an interesting test as well.

[mak013]

Really nice. And for what time do you plan testing? The main problem is 6 hour switch i think. 1 time a day you will sleep this time May be switch every 12 hours will be better?

[WanderingPhilospher]

Really nice. And for what time do you plan testing? The main problem is 6 hour switch i think. 1 time a day you will sleep this time May be switch every 12 hours will be better?

Yeah, the 6 hour switch was to try and get closest network hash/difficulty and in case I ended it at 24 hours then each miner had 4 swaps, but I think 12 hours would be ok as well, even 24.

They have been running for almost 24 hours now; 24 hours may be good enough for the test. That would include ample time for dev fees and GPU swapping to all be factored in.

[WanderingPhilospher]

Ok, so the first round of results are in.

I ran both miners for about 24 hours and 15 minutes.

Gminer earned = Eth 0.003349   BTC 0.00020561 USD 7.48

T-rex earned    = Eth 0.003545   BTC 0.00021765 USD 7.92

I am running one more test and I will post those results tomorrow and then I will give my thoughts on miners and the tests.

[P00P135]

Ok, so the first round of results are in.

I ran both miners for about 24 hours and 15 minutes.

Gminer earned = Eth 0.003349   BTC 0.00020561 USD 7.48

T-rex earned    = Eth 0.003545   BTC 0.00021765 USD 7.92

I am running one more test and I will post those results tomorrow and then I will give my thoughts on miners and the tests.

Trex always performs better pool side for me.  Slightly higher hashrate and less stale shares compared to gminer and ethminer so its worth the 1% fee for me at least.

[WanderingPhilospher]

Ok, I ran another test, this time on ethermine, for 12 hours.

Two main reasons:

1.  I wanted to see the difference with smaller difficulty shares. Ethermine is 4.92 and 2Miners is 8.72.
2.  Something seemed off at the end of the first test. Almost like someone put a GPU or two on T-rex's wallet.
The graph looked extremely strange the last 4 hours. But I have no way to prove anything, it just looked odd.

You can see the results here:

Gminer:
https://ethermine.org/miners/771158cd4ea143aed3e2f5a733616e1dc881da7a/dashboard


T-rex:
https://ethermine.org/miners/7722b72718eb275e2e31017288bed5e27ea04401/dashboard

The eth amount will continue to change for a few minutes, but as of right now:

Gminer = $3.65 USD = 0.00172 ETH

T-Rex =  $3.64 USD = 0.00172 ETH

So if the first test was on the up and up, then it appears T-Rex does better with higher share diff pool
but they basically do the same on lower share diff pool.

Looks like they earned a little more with lower hashrate pool (2Miners) compared to Ethermine.

So those are the numbers...the data.

Remember, facts don't care about your feelings.

Hope you learned something!

[ZeusAlpha]

Hey there's a new version of GMiner out and would love to see some updated benchmark results!  Thanks!

[dedizonesv2]

Already before making attempts between mining software, you will need to know what the pool really gives you in money? and how it gives it to you (protocol) D-PPLNS, PPLNS, PPS + POP?

I have tested many pools, I have been mining for a few years now and it is very hard these days to find a really profitable pool ...

I take the example of

2miners: 1% Fee & 1% MEV
Ethermine: 1% fee pool & 20% MEV
Hiveon: 3% Fee pool & 10% mev
And much more...

So if we take your two test pools, you already don't have the same level of fees ... so even if your software gives you more ETH, you also need the pool fees to be correct and the same!

I am currently on sparkpool and crazypool, and nothing to do with other pools: someone had the good idea to make a more realistic site of what I earn on the pools: https://www.miningpoolsprofits.com/ and as we say the figures do not lie, we can see that this pool is the most profitable!

So if you go with good mining software, good rig settings, good connection to you and the pool, and a pool that doesn't suck your profits, it's no secret you'll have the sum you want at the end of the day. and more

Whatomine announces a theory of:
0.0451 / 24H

I watch how much the pool (crazypool) gives me today
0.0668 ETH

The miningpool profit site
0.0615 ETH / 24H

I have 2 hours left to finish the day, we can still find a block ...

So a word of advice, take a good look at your pools, the "REAL" pool fees,

For my part I use Gminer and Trex

I win the same with both because the share rate is the same for a few shares. except one has 1% fees and the other 0.65%.

You have to think like a farmer, even you have the best agricultural machinery, seed, if the field where you are going to plant, the land is poor, you will not have a good harvest.

If you didn't believe me do the test on the top 5 pools on the site over a 1 week period, you'll see!

Good pick

[WanderingPhilospher]

Already before making attempts between mining software, you will need to know what the pool really gives you in money? and how it gives it to you (protocol) D-PPLNS, PPLNS, PPS + POP?

I have tested many pools, I have been mining for a few years now and it is very hard these days to find a really profitable pool ...

I take the example of

2miners: 1% Fee & 1% MEV
Ethermine: 1% fee pool & 20% MEV
Hiveon: 3% Fee pool & 10% mev
And much more...

So if we take your two test pools, you already don't have the same level of fees ... so even if your software gives you more ETH, you also need the pool fees to be correct and the same!

I am currently on sparkpool and crazypool, and nothing to do with other pools: someone had the good idea to make a more realistic site of what I earn on the pools: https://www.miningpoolsprofits.com/ and as we say the figures do not lie, we can see that this pool is the most profitable!

So if you go with good mining software, good rig settings, good connection to you and the pool, and a pool that doesn't suck your profits, it's no secret you'll have the sum you want at the end of the day. and more

Whatomine announces a theory of:
0.0451 / 24H

I watch how much the pool (crazypool) gives me today
0.0668 ETH

The miningpool profit site
0.0615 ETH / 24H

I have 2 hours left to finish the day, we can still find a block ...

So a word of advice, take a good look at your pools, the "REAL" pool fees,

For my part I use Gminer and Trex

I win the same with both because the share rate is the same for a few shares. except one has 1% fees and the other 0.65%.

You have to think like a farmer, even you have the best agricultural machinery, seed, if the field where you are going to plant, the land is poor, you will not have a good harvest.

If you didn't believe me do the test on the top 5 pools on the site over a 1 week period, you'll see!

Good pick

I am not sure if you were confused by my tests...but I ran both miners on the same pool at the same time. I wasn't comparing different pools, to see what one would earn based on the pool but based on the miner used. I ran a second test privately so without a doubt there wasn't any outside influence on the speed of either miner.

I absolutely understand different pools will pay different based on what they pay and their fees, but that was not the purpose of my tests. It was to see two mining software mining on same exact pool at exact same time and then compare those results.