Author

Topic: [ANN][DASH] Dash (dash.org) | First Self-Funding Self-Governing Crypto Currency - page 2149. (Read 9723827 times)

sr. member
Activity: 426
Merit: 250
Mr Reappearing buywall on Cryptsy, I think things are about to get intersting people.

Yeah I've had to keep changing my purchase price because of him.  Smiley  He costs me $$. 
hero member
Activity: 518
Merit: 500
Mr Reappearing buywall on Cryptsy, I think things are about to get intersting people.
sr. member
Activity: 447
Merit: 250
i now mining dash with gpu, and i want to know if old x11 miner can be used to mine dash ?

edit : is x11 asic miner exist ?
Of course...I'm also mining.

I think x11 asic miner is still no present. But maybe I'm wrong.

A fair point to bring up occasionally but..
the question is rather, is it economically viable to design and build an asic machine for X11?

I designed some very tiny micro controllers and FPGA's in University for projects and we built a fast chip that could multiple two 4 bit numbers together.
You might think that its easy and can be done in 5 minutes, but it took days of involved boolean algebra, testing, errors, checking the faulty logic for latency overlap problems etc. Basically you can make any (speedy) function only from Nand gates and flip flops  but you will need very many of them and they are time consuming to setup.

So how long do you think it would take to design a logic circuit  to do 11 hashing algorithms from scratch ?!
The Sha256 maths on paper takes 16 minutes, and this is way more complicated.
https://youtu.be/y3dqhixzGVo

The months of development is not the killer issue though,
This is going to require a lot of logic in series and some of the hash functions will be quicker than others so you will have most of the circuit not doing anything for most of the time hence it will be costly to have relatively unused circuitry(more than common GPU's most likely)
You could optimize it of course in order to use some of the redundant circuit but this will bring in a new round of testing and errors (using the same part of an algorithm twice is risky as they may over lap-must be thoroughly tested)

So months down the line and 50k-100k later (investors getting nervous) drawing closer to an official release, you find that AMD have just released a new graphics card that is twice as quick as the last and Wolf0 has optimized the .bin again (TIA), thus quadrupling the hash rate so your promised pay back time has now gone up by 4, not such a hot seller on ebay any more.

And even if you get past all that, it has already been said that if there was an X11 asic machine developed, a fork could be taken immediately and now we have x12 instead and all that hard expensive work at the R&D labs for months was a complete waste of time (for this X11 coin anyway). Would any company take such a risk, with that in mind ?  (feathercoin done this already)

So do you take the first option, or do you go out and simply buy a few cheap Nvidia 750 ti's and wait for the next upgrades?

https://dashtalk.org/threads/x11-mining-optimisation-project.2584/page-6


I have personally designed an X11 ASIC - well the RTL only, which can be used for either FPGA or ASIC implementation.  It includes a mechanism to account for variations in performance between each hash type.  Unfortunately the economics are just not there yet to get this turned into a layout.

If you were to submit a proposal for a Masternode vote (when it becomes available) to release an open-source FPGA or ASIC miner it might be possible to gain the support necessary to achieve funding. I don't have the figures in front of me at the moment but I imagine it would require quite a rally in order to be economically viable.
legendary
Activity: 1456
Merit: 1000


Blocksize debate:

I've always liked the idea of the blockchain being a settlement layer with instant-transaction merchant layers on top. A blockchain can never complete with the capacity, functionality, flexibility etc of merchant payment layers no matter how fast. A supermarket, for example, could never use a blockchain directly, it has to be able to have so much meta-functionality and support like insured trades, instant reversibility when someone puts their conrflakes on the belt too close to the customer in front.

The nice thing about Dash is that it can do both.



tok - Can you elaborate on this please?  I have an ulterior motive for asking.  I'm meeting with a group of heavy hitting devs in the DCO world later this week.  I need all the ammo I can get.

.......
In the Bitcoin XT (Gavin Andreesen) vs Bitcoin Core (Peter Todd) debate, Gavin's point is that he wants the blockchain to be as diverse as possible in its potential use and therefore wants it to be potentially able to support a very high transaction capacity, including micro transactions. Peter Todd's view on the other hand is that he thinks the blockchain should now be left alone so that the payment processing meta layer can start to grow and mature.


I'm a bit confused.

Retail transactions happen at an aggregate level. I don't pay for every item with my credit card, debit card or cash. I wait for the kerching at the end.

Where do you shop? lol
legendary
Activity: 3066
Merit: 1188

I agree with the background.  Now for the punchline, "Why is DASH a better mousetrap?"  Feel free to use the following words; instantx, scalability, trustless.

LoL !

You said it    Wink

As far as blockchain times go, I never thought InstantX was a make or break thing at an industrial level since massive transactional clearing processors always get used to do the heavy lifting at the front end which are instantaneous anyway.

InstantX is a bit like an Airbus A320 doing a barrel roll at an airshow - i.e. it's potentially not essential in service but it indicates the amount of spare capacity there is in the design.

Dash's "industrial heavy lifting feature" is basically the decentralised specialisation that masternodes offer which comprehensively solves bitcoin's fungibility problem. Another "industrial level" solution they bring to bear is scaleability.

InstantX is a great spin-off of all that but at the same time though, we are talking about crypto and not fiat. That means that the raw blockchain will in fact see a lot more of the commercial "coal face" than the fiat bank clearing system does. For example, Dash is probably fast and flexible enough at its lowest level to penetrate new markets of the much heralded "unbanked millions". So I'm not saying it's just a fancy backflip to show of with - it could actually break the glass ceiling on many new markets.
hero member
Activity: 532
Merit: 500
By the way I haven't posted here so much lately but I'm very excited to see where the DEV team takes us next. DGBB is gonna be very nice but hard to explain to some people, No doubt that Dash is the most well rounded coin out there, no surprise from me if Eduffield comes out with somethign that blows bitcoin scalability out of the water.
hero member
Activity: 532
Merit: 500
Accidentally sent a test transaction to my mobile Dash wallet from old Darkcoin client I still had on local windows. There were no blocks for a while so I thought I had fucked up. Although instant X didn't work so I'm gonna resend and try again.

My instantX won't work for a few days now. Is it because there is 37% of the network are 0.11.2.22 ?
(I sometimes get a MasternodeB ranking is too low/spamming masternode in my debug log)

sent it again from the latest Dash wallet and instant X worked perfectly.
hero member
Activity: 671
Merit: 500
wisdom

I agree with the background.  Now for the punchline, "Why is DASH a better mousetrap?"  Feel free to use the following words; instantx, scalability, trustless.
sr. member
Activity: 434
Merit: 250
Quantum entangled and jump drive assisted messages
legendary
Activity: 3066
Merit: 1188


Blocksize debate:

I've always liked the idea of the blockchain being a settlement layer with instant-transaction merchant layers on top. A blockchain can never complete with the capacity, functionality, flexibility etc of merchant payment layers no matter how fast. A supermarket, for example, could never use a blockchain directly, it has to be able to have so much meta-functionality and support like insured trades, instant reversibility when someone puts their conrflakes on the belt too close to the customer in front.

The nice thing about Dash is that it can do both.



tok - Can you elaborate on this please?  I have an ulterior motive for asking.  I'm meeting with a group of heavy hitting devs in the DCO world later this week.  I need all the ammo I can get.

Naeborra !

The blockchain was never designed as a payments processor - it was designed as actual base money. The difference is that a payments processor plays around with numbers until things are "agreed", and then when the dust settles, sends the net balance to a more permanent settlement system for permanent "settlement".

There can be many layers in the payments processing process. In the example I described before, when you park your cornflakes too close to the person in front in the supermarket checkout queue and they get lumped in with the other person's transaction, the till staffer simply reverses it out of one and into the other. The software in the cash till buffers all that messing around and only later moves the settled balances onto the next "stage" (which isn't even the clearing bank stage yet).

A few hours later, all the till balances are dumped and reconciled and the transactions can sit in a payment processor ledger (like, say stripe) for a week where people get a chance to make claims, do chargebacks, resolve disputes and whatnot.

Maybe a week later the supermarket actually receives the money.

There are therefore several layers of "buffering" in the system - one at the cash till, another at the payment processor (Visa, Maestro etc). Only much later do the transactions actually hit the clearing banks and permanently "embed" themselves in the system.

In crypto, the blockchain is the "bottom layer" in the financial system - the permanent settlement layer. In a mature commercial economy, there's a very large gap between the payment processor and the settlement layer. i.e. you would never have a heavy duty commercial merchant doing blockchain transactions because:

a) the blockchain is not a payment processor and doesn't remotely support all the functionality needed at the point of sale
b) it's far too permanent and slow for heavy duty point of sale

On the bitcoin network, we're starting to see this "meta layer" of payment processor buffering appear. One example is the Lightning network but there are others such as coinbase that do payment buffering so that - although the bitcoin blockchain may have slow clearing times - transactions are instantaneous at the point of sale.

As the crypto economy matures, that layer will become progressively deeper and more of the front line transactional weight will be absorbed by the meta payment processing layer and possibly only a minority by the blockchain settlement layer.

In the Bitcoin XT (Gavin Andreesen) vs Bitcoin Core (Peter Todd) debate, Gavin's point is that he wants the blockchain to be as diverse as possible in its potential use and therefore wants it to be potentially able to support a very high transaction capacity, including micro transactions. Peter Todd's view on the other hand is that he thinks the blockchain should now be left alone so that the payment processing meta layer can start to grow and mature.

sr. member
Activity: 434
Merit: 250
Quantum entangled and jump drive assisted messages
Accidentally sent a test transaction to my mobile Dash wallet from old Darkcoin client I still had on local windows. There were no blocks for a while so I thought I had fucked up. Although instant X didn't work so I'm gonna resend and try again.

My instantX won't work for a few days now. Is it because there is 37% of the network are 0.11.2.22 ?
(I sometimes get a MasternodeB ranking is too low/spamming masternode in my debug log)
hero member
Activity: 532
Merit: 500
Accidentally sent a test transaction to my mobile Dash wallet from old Darkcoin client I still had on local windows. There were no blocks for a while so I thought I had fucked up. Although instant X didn't work so I'm gonna resend and try again.

Edit: It was 11.1.25 - Really need to clean up all my Darkcoin/Dash folders, it's a huge mess.
sr. member
Activity: 434
Merit: 250
Quantum entangled and jump drive assisted messages

The tx flood stress test was kind of a failure by the looks of the results,
only 20 transactions per second cause this!

hero member
Activity: 671
Merit: 500

Blocksize debate:

I've always liked the idea of the blockchain being a settlement layer with instant-transaction merchant layers on top. A blockchain can never complete with the capacity, functionality, flexibility etc of merchant payment layers no matter how fast. A supermarket, for example, could never use a blockchain directly, it has to be able to have so much meta-functionality and support like insured trades, instant reversibility when someone puts their conrflakes on the belt too close to the customer in front.

The nice thing about Dash is that it can do both.



tok - Can you elaborate on this please?  I have an ulterior motive for asking.  I'm meeting with a group of heavy hitting devs in the DCO world later this week.  I need all the ammo I can get.
sr. member
Activity: 434
Merit: 250
Quantum entangled and jump drive assisted messages
I think it's more than likely to be developed by "parties" who have a large budget to spend on cryptographic equipment, like governments, and who don't necessarily factor the costs in the same way.

At that point you have the world mining with GPUs and governments/secret agencies being able to "outrun" the general public.

Repeated forks every couple of months would kill asic development, even a government with an unlimited budget for 'defence' would not be able to keep up?!
(after all, they can't even finish their own military projects that always go over budget, take years and don't work )
legendary
Activity: 3066
Merit: 1188

Blocksize debate:

I've always liked the idea of the blockchain being a settlement layer with instant-transaction merchant layers on top. A blockchain can never complete with the capacity, functionality, flexibility etc of merchant payment layers no matter how fast. A supermarket, for example, could never use a blockchain directly, it has to be able to have so much meta-functionality and support like insured trades, instant reversibility when someone puts their conrflakes on the belt too close to the customer in front.

The nice thing about Dash is that it can do both.

sr. member
Activity: 434
Merit: 250
Quantum entangled and jump drive assisted messages
I have personally designed an X11 ASIC - well the RTL only, which can be used for either FPGA or ASIC implementation.  It includes a mechanism to account for variations in performance between each hash type.  Unfortunately the economics are just not there yet to get this turned into a layout.

This is way beyond anything we did, I would love to see how far you have got, do you have the program or the simulator results, what  
sort of speeds were predicted, could you tell ?
legendary
Activity: 3066
Merit: 1188

Letstalkbitcoin blocksize discussion

Totally fascinating talk with Gavin and Peter Todd. Massive relief from bitciontalk sniping and trolling. Proper civilised engineering discussion between people of disparate points of view. Andreas Antonopolis there as well.

A lot of it's pretty technical but general gist still comes across.

https://letstalkbitcoin.com/blog/post/lets-talk-bitcoin-217-the-bitcoin-block-size-discussion
legendary
Activity: 1708
Merit: 1049
i now mining dash with gpu, and i want to know if old x11 miner can be used to mine dash ?

edit : is x11 asic miner exist ?
Of course...I'm also mining.

I think x11 asic miner is still no present. But maybe I'm wrong.

A fair point to bring up occasionally but..
the question is rather, is it economically viable to design and build an asic machine for X11?

I designed some very tiny micro controllers and FPGA's in University for projects and we built a fast chip that could multiple two 4 bit numbers together.
You might think that its easy and can be done in 5 minutes, but it took days of involved boolean algebra, testing, errors, checking the faulty logic for latency overlap problems etc. Basically you can make any (speedy) function only from Nand gates and flip flops  but you will need very many of them and they are time consuming to setup.

So how long do you think it would take to design a logic circuit  to do 11 hashing algorithms from scratch ?!
The Sha256 maths on paper takes 16 minutes, and this is way more complicated.
https://youtu.be/y3dqhixzGVo

The months of development is not the killer issue though,
This is going to require a lot of logic in series and some of the hash functions will be quicker than others so you will have most of the circuit not doing anything for most of the time hence it will be costly to have relatively unused circuitry(more than common GPU's most likely)
You could optimize it of course in order to use some of the redundant circuit but this will bring in a new round of testing and errors (using the same part of an algorithm twice is risky as they may over lap-must be thoroughly tested)

So months down the line and 50k-100k later (investors getting nervous) drawing closer to an official release, you find that AMD have just released a new graphics card that is twice as quick as the last and Wolf0 has optimized the .bin again (TIA), thus quadrupling the hash rate so your promised pay back time has now gone up by 4, not such a hot seller on ebay any more.

And even if you get past all that, it has already been said that if there was an X11 asic machine developed, a fork could be taken immediately and now we have x12 instead and all that hard expensive work at the R&D labs for months was a complete waste of time (for this X11 coin anyway). Would any company take such a risk, with that in mind ?  (feathercoin done this already)

So do you take the first option, or do you go out and simply buy a few cheap Nvidia 750 ti's and wait for the next upgrades?

https://dashtalk.org/threads/x11-mining-optimisation-project.2584/page-6




Very interesting Sub-Ether. I've always known that an X11 asic was unlikely given the development costs and likely risk of ROI expectations being thwarted by ever faster GPUs hitting the market, but I've never seen the specifics spelt out like that. Thanks.

I think it's more than likely to be developed by "parties" who have a large budget to spend on cryptographic equipment, like governments, and who don't necessarily factor the costs in the same way.

At that point you have the world mining with GPUs and governments/secret agencies being able to "outrun" the general public.
Jump to: