Bitcoin Forum

Hi guys, we've just reviewed the latest GEFORCE RTX 2080 and compared its mining performance against the GTX 1080 on 4 different algos. Just wanted to share what we've found as the results are pretty interesting.  8)

The setup

Our desktop mining rig is based around an Intel Core i7-5930K CPU and 32GB of memory. We’re running Windows 10 build 1083 and made use of NVIDIA’s release-day GeForce drivers (version 411.63) for the RTX 2080. These drivers seemed stable on our setup and we saw no unusual errors under the hood on our mining software.

All tests were conducted with beta version 0.5.1 of the CudoMiner multi-coin crypto mining software. We focused on the proof-of-work algorithms Ethash, CryptoNight variant 1, Equihash, and Lyra2RE v2. Cudo Miner supports concurrent CPU+GPU mining, but we turned off CPU mining for these tests to ensure it didn’t interfere with our results for each GPU.

https://imgur.com/a/MHUxtn0 (https://imgur.com/a/MHUxtn0)

Analysing the results

Although both cards consume essentially the same amount of power when idle, the RTX 2080 draws a significantly larger amount of power under full load, with the worst case scenario of a 27% increase over the GTX 1080 when running Equihash.

The RTX 2080 boasts an average of 60% hashrate increase over the GTX 1080 across the four tested algorithms.

Hashrate

Ethash   
NVidia GTX 1080 Founders OC: 20.2 Mh/s (28.7 with Ethpill)    
Gigabyte RTX 2080 OC: 35.3 Mh/s (+75%)

Cryptonight Variant 1
NVidia GTX 1080 Founders OC: 507 h/s   
Gigabyte RTX 2080 OC: 769 h/s (+50%)

Lyra2REv2
NVidia GTX 1080 Founders OC: 42.1 Mh/s   
Gigabyte RTX 2080 OC: 72.1 Mh/s (+70%)

Equihash
NVidia GTX 1080 Founders OC: 391 Sol/s   
Gigabyte RTX 2080 OC: 566 Sol/s (+45%)

See the following images for tables and graphs:
https://imgur.com/a/ZvPdYco
https://imgur.com/a/iQmZkdT

Based on launch-day market conditions, we found the Ethash-based Ethereum to be the best coin to mine on both cards. Over a projected month of mining, the RTX 2080 delivers $0.12 per watt, with the GTX 1080 coming in at $0.09 per watt. This translates to a monthly revenue of $17.91 for the RTX 2080 versus $10.23 for the GTX 1080, before electricity costs.

https://imgur.com/a/iQSHiJ8
https://imgur.com/2AxxaVi
https://imgur.com/bXOOTNI

Other considerations

We also trialled the free ETHlargement Pill tool, which optimises memory structures on GDDR5X memory-based GPUs such as the GTX 1080 and claims to boost mining performance for Ethash-based coins like Ethereum. This gave our GTX 1080 setup a roughly 40% increase in output. We don’t know right now if similar tools may be released in the future that will similarly optimise newer generation cards such as the RTX 2080, but if so, this could result in further revenue boosts.

Both cards were tested in our rig with their factory settings unchanged for core clock, memory clock and supply voltages. Overclocking or voltage modding could change the performance for one or both cards, in which case the performance and efficiency gaps between the two may be different.

Wrapping up

The RTX 2080 shows an increase in power and efficiency over the previous generation of cards, but is it worth putting down the extra cash? Right now, if you’re looking to get one just to boost your mining revenues, the answer would seem to be no.

Our tests have shown that although there is a slight increase in revenue, the incremental improvement over last-generation’s tech is far outweighed by the high retail cost of the RTX 2080. In the future, it’s possible that new algorithms may take advantage of the new generation of AI-driven computational abilities, but right now there’s nothing in the crypto mining world that takes advantage of the full potential of NVIDIA’s new offering.

If you’re looking for a quick return on your investment, at the moment your best bet is still to try and get a good deal on the second-hand market for previous-generation cards such as the GTX 1080.

What are your thoughts about the results? Has anyone tested the 2080 on their rig?

You can view the review video here https://www.cudominer.com/rtx2080benchmark/ (https://www.cudominer.com/rtx2080benchmark/)

why not show hashrate?

+1

useless review...

And why on base clocks?

Stable max efficiency anyway you achieve it is what matter these days.

garbage review.  go away

Jesus that was just a horrible review man.  Just drop hash numbers next time and call it a day you big dumbo.

But why comapre it against 1080 and not 1080ti? Your review sucks.

Easy guys! I forgot to add a link to an image with a table showing all hashrates. Post updated. No need to get so excited 8)

Hashrates

Ethash   
NVidia GTX 1080 Founders OC: 20.2 Mh/s (28.7 with Ethpill)    
Gigabyte RTX 2080 OC: 35.3 Mh/s (+75%)

Cryptonight Variant 1
NVidia GTX 1080 Founders OC: 507 h/s   
Gigabyte RTX 2080 OC: 769 h/s (+50%)

Lyra2REv2
NVidia GTX 1080 Founders OC: 42.1 Mh/s   
Gigabyte RTX 2080 OC: 72.1 Mh/s (+70%)

Equihash
NVidia GTX 1080 Founders OC: 391 Sol/s   
Gigabyte RTX 2080 OC: 566 Sol/s (+45%)

Thanks for your efforts and work.

Could you precise the versions of the miners and the parameters please ?
A GTX 1080 ti have better hashrate for most algos at a lesser price.
Could you test lyra2z with    
CryptoDredge 0.9.1  please ? https://bitcointalk.org/index.php?topic=4807821.0

We wanted to compare it to the same model of an earlier generation. 2080Ti has been delayed by a week. As soon as I get hold of one, a full 1080Ti vs 2080Ti test will be performed. ;)

pricewise, RTX 2080 is against GTX 1080 ti. It makes sense to compare it to both GTX 1080 and GTX 1080 ti.

need to compare with lyra2z algo

why are u testing it against a 1080 ?

you should be testing it against a 1080TI , the 1080TI NIB is like $200 cheaper and anyone with half a brain would buy that over 1080

Yeah this reviewer must not be a miner , this is not  a GAMING site dude, we don't give a fuck about anything thats not related to how much two cards cost in relation to each other and in relation to Hash/$

You job as a reviewer here is to test the shiny new 2080 against the next best lower gen card which is again * drum roll* $200 dollars cheaper, what we would me more interested in is how much power

does a 2080 use vs the card closest to it in price performance : Vega 64 and 1080TI  maybe even a 1070TI is you wanted to go crazy.

Nice effort but I will wait for a real review.

What miners do you use?

Your numbers for the 1080 are not so good. 1080s do 40Mhs on eth, 62Mhs on Lyra2v2 and 650 Sol/s on Equihash.

For lyra2v2 I use cryptodredge, claymore+ethpill for eth and equihash is ASIC algo.

Algos which will interest people around here would be: x16r/s(try Z-enemy miner), Equihash144/192 (EWBF 0.5), ProgPoW.

Can you people chill and tell the guy what to do instead of blurting random insults?

Where did you get these values? Did you mine on the default settings?
These are the real values for overclock and  optimization.
Ethash   
NVidia GTX 1080 : 25 Mh/s (35 with Ethpill)    
Cryptonight Variant 1
NVidia GTX 1080 : 650h/s   
Lyra2REv2
NVidia GTX 1080 : 51,8 Mh/s   
Equihash
NVidia GTX 1080 : 588 Sol/s   


https://www.youtube.com/watch?v=Ufo0u6kBCu8
50 sec

There is no benefit from the tests of the RTX 2080 on non-optimized software.

Fair point Marvell2, we will be doing more reviews in the next couple of weeks so will definitely post stats for 1080Ti vs both RTX 2080 and 2080Ti when it's finally out.

I don't believe on the hashrate give in ETHhash Algo. because my gtx 1080 hashrate can reach to 25 to 30 mh/s with etherpill it can reach maximum of 39mh/s to 40mh/s with small OC.

What do you think for RTX 2080 maybe it also have a software to increase the hashrate like ethpill?
Only few algo's has differences on hashrate there is a little improvement but the price still worth it if you can buy it for cheaper price.


I think it's fine because there RTX 2080 is just like GTX 1080 so you can compare GTX 1080ti and RTX 2080ti because it has the same TI in the end of their model name.