I have wondered this for a while and a recent RT article brought it back to mind again. The RT article compares the BTC network to the Visa network not in transactions per second but energy per transaction and the amount of energy it takes to power the BTC hashrate. https://www.rt.com/usa/270550-bitcoin-energy-hog-critics/
The article doesnt take into account more efficient ASIC's nor increased hashrate.
So I tried to assess the LTC cost per transaction but would like to know accurately with a computer doing the math based on fact instead of me based on guestimation so looking for input. I assessed the value as follows, leading up to the per transaction costs; see below
Average power usage per GH by ASIC type
4500Watts : 1GH Titan
9000watts : 1GH A2
14000watts : 1GH Generic & low MH units
doesnt include A/C which is another 20-50% of their ASIC electric costs but incurred by all 3
Network 1.2 TH / 1200 GH
%Total shot in the dark here%
50% Network hashrate A2
25% network hashrate KNC Titan
25% network hashrate generic & low MH units (&posibly some crazy guy with a GPU)
Which means that the total power consumption would be something like
A2's = 50% of 1.2TH, 600GH @ 9000watts per GH, A2's use roughly 5400KW / 5.4MW
KNC Titan = 25%, 300GH @ 4500watts per GH, Titans use roughly 1300KW / 1.3MW
Other asic = 25%, 300GH @ 14000watts per GH, other ASIC use roughly / 4.2MW
That would mean the LTC network @ 1.2TH would be using just under 11MW of electricity
The current LTC blocksize is 1MB @ 576 blocks per day & using BTC average transaction size 250b thats 4000 transaction per MB or 2,304,000 transactions per day and I may be off but that seems to come to about 4.7watts per transaction.
At todays utilization rate, hashrate and average transaction frequency, the numbers above put the LTC network at around 1KW per transaction which is really expensive.
Is there some where that anyone knows that actually uses the block chain to calculate the above?
So if the bitcoinXT wallet is pushing 8MB blocks, wouldnt that make BTC capable of being 8 times more energy efficient per transaction then it was before? And even with that I am thinking it doesnt compare to scrypt. Could one not argue that the carbon footprint for Scrypt mining is much more sustainable and comparable to conventional fiat systems then the SHA256 algo. Does anyone have a link that puts together a viable assessment of the cost per transaction to support SHA 256 algo vs the cost of transaction to use Scrypt? I know that scrypt leveraging memory is more energy efficient but I dont know by how much.
100% of everything above is extrapolate and conjecture, but ultimately the information I am trying to assess I think is of value.
Seems if there are some more big green anti-BTC movements, maybe some well placed LTC marketing could give LTC the push it needs.