Topic: [LTC] Changing the litecoin Proof of Work function to avoid ASIC mining? - page 6. (Read 33318 times)

Check this topic : https://bitcointalksearch.org/topic/vertcoin-first-scrypt-n-first-stealth-address-privacy-without-mixer-404364.

Vertcoin : Adaptive N-factor in Litecoin

 

The new Cuckoo Cycle proof-of-work system that I made available yesterday at

  https://github.com/tromp/cuckoo

should be more robust against time-memory tradeoffs.

You modified it?

Oh great just what we need, a home-made made-up crypto algorithm.

Didn't they have a don't do this at home label on the damn thing?

Please everyone avoid the damn thing until real cryptographers have had a few years to thoroughly look it over like they probably did with the real Blake that this guy claims to have hacked up somehow.

-MarkM-

Novacoin changes stuff around, they think nothing of screwing around with interest rates for example, setting them nice and high knowing their cronies won't object as long as they plan to change them back low again once they and their cronies have gained lots from the high interest.

Gosh knows what other screwing around they might decide at any moment to do.

-MarkM-

The majority have agreed to change the POW. Fork away!

More discussion:
http://www.reddit.com/r/litecoin/comments/1u2fst/a_majority_of_litecoin_holders_agreed_in_a_vote/

Do we need false resistance?
The problem with picking an algorithm like scrypt is that you are starting with a false assumption of "resistance"(think of a cheap watch that says water resistant)
resistance is not the same as proof! e.g Litecoin started by saying GPU,FPGA,ASIC resistant this has now changed over time to just be Asic resistance, can you see the problem?

Why is it a problem for Scrypt?
Because as the algorithm became more popular and other people starting working on ways to mine scrypt they found ways to mine it on GPU and FPGA thus the so called "resistance" is a flawed method!

What's so bad about Scrypt?
to create the resistance the developers have created a linear function which consumes resources on any platform CPU, GPU, FPGA, ASIC

this is bad for Asic development because it costs money for extra silicon space

the reason Asic's are so fast is the they are Application-specific integrated circuit and the more cores you can get on a small amount of silicon space/design area the better as it will result in cheaper and faster designs and thus Asic mining hardware

so you can understand that Scrypt Asic's will cost more and also not be able to scale the cores to get good performance in terms of hashes per second for the end user thus it will be a very long time to get your ROI for a Scrypt based Asic device  

The exponential growth of technology is often described by Moore's law and has shown for many years how we will always look for new more efficient and faster ways of doing the same thing  

Better Idea?
Pick a fast and well balanced algorithm that can work well on all platforms so when you use it on GPU,FPGA,ASIC you get good performance and lower costs to produce any type of hardware

I think it would be best to have good balance of efficiency, power usage, performance and cost effective implementations on all platforms so when someone wants to spend excessive amounts of money on R&D for a Asic miner they should pick an algorithm that would use very little silicon space = more chips a wafer = cheaper $ per chip

I have picked and modified Blake-256 to fit this exact role as it has a better parallelism than SHA-256 and is faster and uses less resources  

First coin to be made with Blake-256 is Blakecoin (named after the algorithm) and is also merge mine capable  

unlike Scrypt based coins with Blake-256 we can make new coins and then just merge mine them on pools, we can then just mine pools that merge the coins we like rather than this pick and choose battle that is going on in the Scrypt based alt coin world  

I am not a salesman or a marketing guy so I built Blakecoin from a technical point of view with efficiency and long term use in mind and forgot to add the spin and hype  

anyway hope I have explained why Scrypt based Asic's are such a risky investment  

Merry Christmas everyone

These Scrybt Asics are still vaporware and from what I read in the asic miner thread they more expensive per MH then GPU , only advantage they have is less power consumption

Bump for move votes. Not sure if it even matters anymore

This statement isn't necessarily true, the correct statement is that higher decentralized hashrate leads to a stronger network.
Here's a comment by coblee on that: https://bitcointalksearch.org/topic/m.1723744
You can see some more info at https://litecoin.info/Comparison_between_Litecoin_and_Bitcoin#SHA256_mining_vs_scrypt_mining

No proof yet but from this thread https://bitcointalksearch.org/topic/gridseed-gc3355-hybrid-scryptsha256-asic-355268 there should be chips in developer's hands in the next few days. The performance is nothing to write home about (60kHash/sec scrypt plus another 1.75 GH/s or so SHA256d in dual mode), but where they lead others will follow.

Jasinlee's Fibonacci project is now claiming to be ASIC too https://bitcointalksearch.org/topic/ann-fibonacci-info-and-questions-368468 though originally it was FPGA. This employs off-chip SDRAM so is likely to be more cost effective IMHO.

Check out cypherdocs gold down btc up thread where this economic effect was discussed as a possible cause of btc price rise.  New people come to btc seeing huge gains.  Mining appears on the surface to be easy money.  But once people realize mining is a zero sum losing game they move on to straight investment.  Wouldn't surprise me if eventually we will see a crypto coin released with a ready made ASIC that goes with it.

I voted NO not becuase i wan't this coin to fail but becuase i don't think it would be practically possible to make sush dramatic changes and get most people to agree on a new Nfactor. I don't wan't the coin to fail yet since i need to finance my new multi GPU setup  

A new coin only with a different Nfactor would probably die off pretty fast for obvous reasons.

Just having the threat of possibly changing Litecoins algorithm could be enough to keep ASIC's at bay (or keep them secret).

No one will invest in developing an ASIC for Litecoin if there is the chance that the ASIC will then become worthless.  

I used to be against ASIC's being made for Bitcoin while I was mining Bitcoin with my GPU's.  The supply could not meet the demand initially and it felt unfair.  With all the huge delivery delays and scams, I was not prepared to risk investing in ASIC's.  Also, for a about a month, Bitcoin became extremely vulnerable to a 51% attack.  Anyone with just a few ASIC's could have overpowered the network.

Now however, Bitcoin ASIC's are readily available, the network is becoming very secure once again, and the massive investment in technology has only added to the credibility if Bitcoin.  

If Litecoin can follow the same path, I think it will only be good for it.

I haven't actually read anywhere that someone has come close to having a working ASIC for Litecoin though, which is surprising.  Bitcoin was at $15 with 25 btc per block when jgarzik received his first 48gh Avalon.

YaCoin with its increasing N factor already covers that niche. Current memory requeriment is at 4MB (for Litecoin's 128kB), will be 32MB in 2015, 256MB in 2020, 1GB in 2030, etc. Whoever doesn't like the effect of scrypt ASICs on Litecoin can just trade their coins and will likely lose less value than in a hard fork -- which, if it just raises N once, will be just a temporal patch.

I answered no to the pool, and I have to say I am surprised that the majority of vote is Yes (LTC haters?)
I agree that ASIC technology decreases mining decentralization but ONLY during the first stage.
ASIC is the way to go to secure the network, gain credibility and attract more investors.

Up until now I haven't seen any proof of LTC ASIC, have you? ...

You mean like Bitcoin did?

NO - because it is dangerous to change fundamental properties of a coin after its release. That is crypto 101.

This.  My vote is no, let litecoin develop on its own even if ASICS will come out for it to mine them.

No, because I want LTC to strengthen (higher hashrate leads to stronger network).