Why there could be a mini revolution occuring in Crypto currency.

Hash algorithm Complexity and how it plays a part with regards to efficiency; how this effects market distribution and a sociological network effect , some of the Key parameters of Cryptocurrency or decentralized digital money.

A significant and important prediction to the theory of Multi Hash algorithm and the Cryptocurrency market in whole.

*** note if you are new to cryptocurrency *** This document and writing is of a semi technical nature and covers mining distribution sociological aspects and market effects, feel free to discuss on our forum for more clarification of any details, having said that my intention is that you do not need to have an in depth knowledge of cryptography or cryptocurrency to understand the economic implications of this writing and the following projected theory.

An Explanation and compelling hard and circumstantial evidence using a “Theory of Complexity” as to why, Multi Hash + Random function crpyto (i.e 6 or more hash algorithms + a random function) maybe the permanent future of crypto currency and the possible implications that will have for Quark.

I have some pretty large news if i am correct in this analysis, we may be witnessing a mini revolution in crpytocurrency.

I’d love the get Vitalik Buterin or anyone else in the field of complexity including Jim Rickards feedback as I respect both.

The final proof of Scrypt ASICS is indeed very compelling as shown in the total network hash rate and many GPU users complaining of dwindling profits, the profit side of this is to a degree also multiplied by a drop in the value of the final destination of most if not all new Scrypt currency , Bitcoin.

That aspect aside for the moment as the writing does not focus on the reality of SCRYPT ASICS only their inevitability, and probably reality right now, but amongst this torrent of the SCRYPT currency there maybe a mini Crypto revolution of sorts occurring.

Cryptocurrency mining and distribution is derived of three primary groups in terms of network power:

  1. ASIC SHA256 BITCOIN miners.
  2. GPU Graphic card SCRYPT miners.
  3. To a lesser degree CPU computer processor Multi Hash+ miners.

But in terms of network % distribution the group looks more like this :

  1. GPU Graphic card SCRYPT miners
  2. CPU computer processor multi Hash+ miners.
  3. ASIC SHA256 BITCOIN miners.

This document primarily focuses on the GPU and CPU mining group and their relationship and a theory of the two groups merger through the advantages of algorithm complexity, which will lead to a permanent decentralization solution .

with the future arrival of SCRYPT ASICs, the GPU base of miners are under significant threat of centralization.

To explain that from a political power point of view the GPU base of miners are the staple of the SCRYPT algorithm, and represent a large % of the decentralization of cryptocurrency!

With ASICs on the very near horizon GPU farm owners need to face specific decisions for their profit and benefit:

  1. Move profit capital to an ASIC (if they can even gain access to them initially) which will in turn centralize the market into ASIC devices.
  2. Move to another mining algorithm.

What will undoubtedly occur is both of these options, but let me explain where complexity theory comes in and evens the field almost indefinitely, and more importantly if you get though this document, why that matters.

A multiple Hash algorithm such as Quark with 6 hashes or more + a random function can be mined by a GPU but the power to yield performance ratio is much reduced as compared to Scrypt IN relation to a CPU.

now that might sound like gibberish on first read, but its the light bulb point of this document so i intend to explain it clearly  –

So to do that:

lets line up two Hash algorithms SCRYPT “litcoin and clones” Versus 6 hash+ “Quark and clones” ) and Two pieces of hardware GPU “graphics card” v  CPU “computer processor”

data from reference guide here :  https://litecoin.info/Mining_hardware_comparison

SCRYPT comparison GPU V CPU:

Graphics Card GPU

Computer Processor CPU

Now this is simpler than it could appear, GPU mining is much more efficient at 770kh#  compared to the 65kh# yielded by the CPU clearly GPUs are better at this single hash algorithm because it can more efficiently mine this single hash process.*

Now lets run exactly the same hardware but change the algorithm to 6hash+  (6 Hash + a random function)

6Hash+ comparison GPU V CPU:

Graphics Card GPU

Computer Processor CPU

Things become clearer, Now the GPU yields 1370kh# v 860Kh# for the CPU notice the gap has significantly closed ?

with SCRYPT the GPU has an almost 1100% advantage as compared to 6 Hash+ not even 50% advantage.

So what does this all mean politically and for the market and related sociological effects ?

Because 6Hash+ uses 6 different algorithms and then a random function it has a higher Complexity rating , than a single hash algorithm (SCRYPT) this complexity mean that up to a certain cost point ASICS are eliminated, we would never say never but certainly in the short term they would never be viable in the market.

now a CPU is designed by comparison to process complex multi threaded processes, it can yield 865Kh# compared to just 65kh# processing SCRYPT, the GPU still pulls though at 1370K# on Multi hash due to the brute force of the GPU processor* but it is likely that further efficiency will be hard to gain or yield because of the complexity of the multi Hash+ algorithm.

so while code (miner software) may improve the yield for both CPU and GPU there will probably not be a large spread in this closer ratio 1370kh# v 865kh# due to complexity of the 6Hash+ algorithm.

I need to underline how significant this is, this means that multi hash because of its complexity can be the friend of all well distributed crypto, and it significantly keeps all the original players in the economic market happy plus expands the base of distribution back to where Bitcoin intended it to be !

furthermore unless i have miscalculated, this effect is an “economically hard effect” that is to say the effect is probably a permanent reality and I will explain why.

because as SCRYPT becomes the domain of ASIC and thus becomes centralized (because of its relative simplicity) the network effect of a decentralized market will seek to re-balance that equation, and here is where I might lose some people but in a decentralized environment that is everything, as the effect of decentralization moves forward in time.

As the market seeks to balance this equation it will do so with Multi hash algorithms because they close the gap between both GPU and CPU and they do this though complexity of the algorithm. not to understate the gravity of the statement as it relates to cryptocurrency;  the GPU and CPU mining and distribution comprise probably 80% to 90% of the Cryptocurrency market in terms of decentralization, ASICs are a centralizing force, so they represent a huge hash power for their respective algorithm  but a small % of participants.

Without decentralized mining Crypto currency is essentially dead,  its a non starter.

Why will Hash complexity be the probable end point for the solution?

because it is relatively simple and malleable to program and essentially provides infinite options for complete centralization resistance, GPUs can yield reasonable hash returns (as demonstrated) and so can CPU for for their respective power usage, furthermore the design is future proof (unless i am incorrect in some way), the CPU and GPU field has been leveled, with FPGA and ASIC pushed almost out of the market.

almost a perfect market solution.

Summary and theory prediction.

Now even if you didn’t fully understand most of the specifics of this document.I can make some bold predictions based on it:

I can derive from both the hard data and the circumstantial evidence that Multi Hash + will soon dominate the Crypto currency market, SHA256 and SCRYPT will still be relevant along with other single algorithm variants, but the new and very probably permanent member will the multi hash +, this is due to its relatively simple to program in complexity + the options at variation, that adds enough risk that centralizing forces are outside feasibility.

Due to the random nature of the free market , plus the random nature of the actual Multi hash+ algorithm, this system effectively protects against a potential overwhelming centralizing force(s) that exist outside a profit motive, that is to say Banks or Governments working at the behest of Banks.

Why this is so important?

Because we can not look at crypto currency in traditional centralized measures or metrics, because it essentially has a multi dimensional aspect.

Because decentralized systems always try to move towards neutral control, but humans tend to affect centralization, these are two forces that tend to need to be in balance, when out of balance the system is non efficiently functioning and seeks towards equilibrium.

My theory predicts that Multi hash is the best decentralized option for the largest amount of the participants of this system, i.e  the “computer user”, and this is essential for a functioning decentralized system that will tend towards neutral control.

General Statement:

If i’m correct (and time will tell), this means that Quark and its proto SIF have achieved a “first of a first”, another strange accident of the free market that no one could predict, where Bitcoin has got the protocol right and correct, Quark was the first with what I predict will be a time proven and permanent variant of a decentralized PoW solution.

The Quark Algorithm and or multi hash + variants will dominate Crypto currency in the future as the permanent Proof of Work solution for decentralization,  and for the general benefit of the most of the market, it will do this because it benefits the two largest permanent participants in the market as far as % of participants, the GPU and CPU market – essentially the “home computer” market.

Kolin Evans.

* I did not talk about the “memory hard” function of SCRYPT, because it is not specifically relevant to the theory nor does it change the results, but I do understand why the variation in # between the 1st and 2nd sudo experiment , happy to discuss on the forum or anywhere.
* I do not believe non-PoW currencies are viable alternatives, the random nature of PoW goes a long way to add an essential network effect for distribution and decentralization. ( i could be wrong, let time tell)

Where could the theory show a significant fault?

if a multi hash algorithm can be adapted to GPU and yield a similar Hash rate to a single hash like Bitcoin or other single Hash algos Keacck etc, adjusting for Moore’s law etc.

i.e if a GPU can yield 500 ~ 700Mh# + that’s M for Mega not K on a real multi hash + a random function , then I will state that I was incorrect.

where as the top card now (priced $500 to $600 USD) can just get 2.5M# on a multi hash called Darkcoin, and supposedly 2.1M# on Quark, of which i know for certain that Quark uses a random function in the hashing mechanism.

A multi hash that does not seem to use a random function and actually only mines a single hash at a time yields around 10M# Myriad coin  https://bitcointalk.org/index.php?topic=483515.740, but on examination its not actually a multi hash and it doesn’t pass the test, (so far so good.)

This would mean a significant ratio gap spread between GPU and CPU , which would mean that GPU will still dominate but ASICs are still not feasible, but it would open the door in the market for another algorithm to gain market share.

** side note this also means that GPU prices will drop but not as much as if Multi hash did not exist.

 

–update– 23 sept 2014

I am now more educated re Myriad – Myraid is a CPoW when looked at in its whole as the reward is determined by all competing algorithms, theory still holds, and prediction is on line as CPoW is taking over the PoW market.

Advertisements

6 comments

  1. The closer gap between CPU and GPU on “6 hash+” is likely more due to lack of GPU optimization than to “higher complexity”. Also, “6 hash+” is a very suitable ASIC target, much more so than scrypt.

    If you really want to improve the viability of CPUs you could adopt an extremely memory intensive PoW like my Cuckoo Cycle (see this reddit discussion http://www.reddit.com/r/dogecoin/comments/210y4a/how_to_stop_asics_and_multipools_xpost_github/cg8qnge)

    1. You are missing the point, but some of that is due to me not clarifying, cPoW (Complex Proof of work)

      where by you can see example coming out now x11 (Dark) and other Quark clones combined with a moderately quick distribution will provide the Resistance.

      the GPU optimization has yet to be realized so, unless we have some reality or evidence to point too, that is just theory.

      the point being that Developers can package hash functions and random function and this shifts the ASIC frame work, sure an ASIC can be made; maybe a replaceable chip design etc.

      but this would mainly be used by “government agencies” i.e Bank interests.

      so would not be realized by the wider market. so is of little value and can be avoided easily.

      but i will look at your link.

    2. Just read that, its just a talk about its implementation, is there any other reference ?

      I.e has it been implemented anywhere and has anyone got # rates from various devices yet ?

      so it would be a memory intensive CPU algo , basically designed for server farms but prevent robots due to high mem usage ?

      its possibly viable but cPoW as i stated right now keeps most of the market happy as it has both parties at a reasonable ratio –

      everyone talks about robots etc , yet no one cares about them as they don’t represent the type of % fix permanent monopoly that ASICs do.

      but a supposedly true “CPU” only algo would be interesting to see in trial.

      1. The whitepaper is there as

        https://github.com/tromp/cuckoo/blob/master/cuckoo.pdf?raw=true

        Rate follows more or less from
        “running time for the current implementation on high end x86 is under 24s/GB single-threaded, and under 3s/GB for 12 threads.”

        So if you pick a memory size of 4GB (cuckoo128), and you have 12 threads on your CPU,
        then you’ll need about 4×3 = 12 seconds for one instance.

        | so it would be a memory intensive CPU algo , basically designed for server farms but prevent robots due to high mem usage ?

        Correct. But also designed for people with a high-end PC, to use their idle cycles.

        | as it has both parties at a reasonable ratio –

        But both parties will be unhappy to get killed by FPGAs…

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s