Author

Topic: Nexus - Pure SHA3 + CPU/GPU + nPoS + 15 Active Innovations + More to Come - page 324. (Read 785514 times)

legendary
Activity: 868
Merit: 1058
Creator of Nexus http://nexus.io
CryptoHobo,

Add this in your util.cpp
uint64 mpz2uint64(mpz_t n)
{
    unsigned int lo, hi;
    mpz_t tmp;

    mpz_init( tmp );
    mpz_mod_2exp( tmp, n, 64 );

    lo = mpz_get_ui( tmp );
    mpz_div_2exp( tmp, tmp, 32 );
    hi = mpz_get_ui( tmp );

    mpz_clear( tmp );

    return (((uint64)hi) << 32) + lo;
}


Add this to your util.h
uint64 mpz2uint64(mpz_t n);


PrimeSieve in prime.cpp:
delete nonce = mpz_get_ui(zTempVar);
add nonce = mpz2uint64(zTempVar);


And at top of PrimeSieve in prime.cpp:
delete unsigned long nonce = 0;
add uint64 nonce = 0;

Worked for me Smiley
Viz.
legendary
Activity: 1050
Merit: 1000

Thanks for the info. Another method is to add the code below to the header file and then replace all instances of unsigned long with word_t in the source files.

#ifdef _WIN64
  typedef unsigned long long int word_t;
#else
  typedef unsigned long int word_t;
#endif


I added this to core.h (presuming that was the correct header), then did a search and replace through the other source files. Seems to be running better on my windows 8 system but crashes on the windows 7 system this could be because since my last build ive removed and reinstalled cygwin then built gmp from scratch or because i've bodjed the code lol
legendary
Activity: 868
Merit: 1058
Creator of Nexus http://nexus.io
Phzi,

You are correct - nothing can be proved until source code is released. We will all be the judges then [until this time both sides are only opinion working their ways to veracity]. And yes, it was created from Peercoin Core, which is partially [Satoshi's Vision for Another] why I will continue this trend.

Everything comes in time my friend  Smiley
Viz.

edit: added more detail

hero member
Activity: 700
Merit: 500
Viz, so as we know, GPU miner already exists for CPU channel. At least one dev (Christian) already has it. Hats off to them, knowledge is power, if I was good enough, I would make one but I'm still far from it.
But why don't you make one and make it public then? I mean, it's more fair than just let few people have it.
Way better than "we won't work on GPU miner for CPU channel".
Just my personal opinion.
The multi-channel approach and trying to offer a CPU channel is failing pretty badly so far.  The original promise was to "make the distribution as fair as possible".  If that was really going to be achieved, CPU mining is a very poor choice, GPU miner should have been released BEFORE launch (I talked about this before the coin was launched) to allow time for optimizations, and Viz would be taking a smaller dev fee.

Instead, the coin launched with a severely un-optimized CPU miner, and a PoW algorithm that is actually fairly efficient to run on GPUs.  But, only a handful of people know how, so the distribution is almost certainly skewed to those select few.
member
Activity: 111
Merit: 10
Viz, so as we know, GPU miner already exists for CPU channel. At least one dev (Christian) already has it. Hats off to them, knowledge is power, if I was good enough, I would make one but I'm still far from it.
But why don't you make one and make it public then? I mean, it's more fair than just let few people have it.
Way better than "we won't work on GPU miner for CPU channel".
Just my personal opinion.
sr. member
Activity: 518
Merit: 275
If you fail...just dont fail again

Shkembe,

As I've mentioned, the best source of information is coinshieldtalk.org. You will find fresh updates on the chatbox, I am rarely here on Bitcointalk anymore. Pool is almost done, in final testing phases. Keep in mind I cannot take any existing pool software to build a Coinshield pool, so I am coding it from scratch. This doesn't happen overnight.

Exchanges will be given source code upon their interest to add us, nothing more.
Viz.


At coinshieldtalk.org you have 92 posts vs 1270 here, I think you can take us more serious. As I see you have no intention to promote your coin and perhaps you expect help from the comunity to focus exchange's interest to it. If yes just tell us. Personally I am thinking to abandon this coin.


We are currently building all our infrastructure. Unlike other coins (which some people have become accustom to), quality is not sacrificed for quantity (If this was the case, we could have released 11 coins each with 1 new feature). We have reached out to a few exchanges which were recommended to us on our chatbox, however please reach out to anyone about us (exchange or not) help spread the word about stopping scam/clone coins, protecting good coins, and about the most secure coin with 11 new features that is ONLY 15 days old.

This is what I outlined previously as re-quoted by viz:

@Viz, whats your marketing plan? We need to build a strong community if you want this project succedeed

In the next few hours the forum (coinshieldtalk.org) will be going offline for some updates.

Unlike other coins, CoinShield's value directly stems from our community. The more people vote, comment, create petitions, and exchange shitcoins for CSD the more CSD's value will rise.

CoinShield is the first ALT. coin made for miners and investors who are tried of owning non-Verified coins, supporting non-Verified coins, being scammed by non-Verified coins, and losing money on non-Verified coins, Our internet marketing won't focus on trying to persuade the crypto community that we are the best non-centralized currency in the world (we all know that is bitcoin). Instead it will be focusing on introducing CoinShield to:
  • the entire mining community
  • all the pool operators
  • all the ALT. coin investors

Once CoinShield is fully operational (all channels are launched, voting and exchange are ready) we will be buying adspace on various sites (bitcointalk, coinmarketcap, etc.) to help inform the community. We may incentivise users to add CoinShield on their signature here on bitcointalk. Currently we are building our followers on twitter to let the community know what coins are being killed in real time (follow us now if you currently do not: https://twitter.com/CoinShieldTalk).

I feel our job is to inform the community and keep them up to date with what we are killing and what is being voted on. Internet Marketing that focuses on anything else is just to create hype and make a bubble. We want CoinShield to be a stable economy where wealth can be built while clones/scams can be killed.

With this said, If you have any ideas on what else you think we should do please PM me or post here.


With that being said, Viz does post on our chatbox on a regular basis for those who are following it.

My thanks go out to everyone who is keeping up with us as we grow and build Coinshield.

Cheers,
KryptoKash
hero member
Activity: 700
Merit: 500
Work on a GPU miner for the GPU channel would seem more fitting as logic goes [GPU Channel Launch in ~2 weeks]. I'm sure the little CPU miners with few cores such as myself would appreciate a reasonable difficulty level for their amount of computing power.

Not all of us have Supercomputers
Viz.
This is just silly.  If an algorithm is more efficient on a GPU, then of course someone will be mining on their GPUs.  Better that the miner is publicly released instead of only privately available.

CPU mining is kinda silly anyway, and mainly works to encourage botnets.  

---

As for people commenting on insufficient marketing and a lack of exchanges, I think this is to be fully expected.  Nobody in their right-mind should trust closed source cryptography.  Until the CoinShield wallet source code is released, this project is simply a bunch of promises with no proof, and no good exchange is going to adopt it.  Providing source specifically to exchanges is equally stupid - why would you trust them but not the masses, and how would they know the source code is what client's are running unless there is a reproducible binary build process (which I am sure Viz has not created).

Viz's continued claim that the source code is private to keep people from cloning his work is an extremely poor excuse.  One of the base ideas of this project was to absorb the economies of blatant coin clones.  If he had any faith in the idea, I don't see why he would be even slightly threatened by people using his work.  I'm sure he's using other people's GPLv3 licensed work - CoinShield certainly wasn't built from scratch.
legendary
Activity: 868
Merit: 1058
Creator of Nexus http://nexus.io
Supercomputing,

Yes sir I do have a code base [OPENCL] that Bitslapper has been working on. And yes GPU channel is Active on Testnet. You'll need to use the command SetChannel(2) in your mining LLP to tell Coinshield what type of block you will be working on.

Thank You,
Viz.

sr. member
Activity: 278
Merit: 250
Work on a GPU miner for the GPU channel would seem more fitting as logic goes [GPU Channel Launch in ~2 weeks]. I'm sure the little CPU miners with few cores such as myself would appreciate a reasonable difficulty level for their amount of computing power.

Not all of us have Supercomputers
Viz.

@Videlicet

Do you have an existing code base (OPENCL/CUDA) for the GPU channel that I can review? Is the GPU channel activated on testnet?
legendary
Activity: 868
Merit: 1058
Creator of Nexus http://nexus.io
go6ooo1212,

I begin work on an open source GPU miner for the CPU channel

GPU Channel Activates at Unix Timestamp 1413914400. GPU Blocks are Rejected until then.
Viz.
legendary
Activity: 1512
Merit: 1000
quarkchain.io
I've missed something - is there working win binanrie for GPU mining ?
sr. member
Activity: 466
Merit: 250

Shkembe,

As I've mentioned, the best source of information is coinshieldtalk.org. You will find fresh updates on the chatbox, I am rarely here on Bitcointalk anymore. Pool is almost done, in final testing phases. Keep in mind I cannot take any existing pool software to build a Coinshield pool, so I am coding it from scratch. This doesn't happen overnight.

Exchanges will be given source code upon their interest to add us, nothing more.
Viz.


At coinshieldtalk.org you have 92 posts vs 1270 here, I think you can take us more serious. As I see you have no intention to promote your coin and perhaps you expect help from the comunity to focus exchange's interest to it. If yes just tell us. Personally I am thinking to abandon this coin.
If you decide to abandon this coin please send to me > 2RQZ4VGqwqX2au4AMrKXbmA5fPdp37DbyHTeznhi6Q5eM9hoChT
sr. member
Activity: 433
Merit: 250
legendary
Activity: 868
Merit: 1058
Creator of Nexus http://nexus.io
Shkembe,

fudge,

Direct this one at Krypto Kash, he's the guy for marketing. I can though breakdown what I'm working on, and how this will work hand in hand with his marketing:

A] - Getting solid GPU developer working on the GPU miner edit: Bitslapper is on it, GPU launch is going to be 3 weeks from today, I'm releasing some new updates that will go live on 10/7/2014 that make some modifications to the difficulty which will modulate the nTargetTimespan instead of having nChainMod be another modular like the block time, this will keep block time at 1/2 the time when released rewards are too high.

B] - Developing Open-Source Pool, which will go live on 10/7 with the new updates, donation only [so free if you want it to be], using LLP server templates.

C] - Proof of stake will then go live 2 weeks after GPU launch, which will prompt me to develop the voting channels on the main site. Once these are functioning and live, 2 weeks later the first exchange channel will be opened up. From there a new channel will be opened every 1-2 weeks.

D] - Once all mining channels are active, and the voting/exchange channels are active on the main site, this is when the whole Coinshield system will be fully functional. Every new coin added to an exchange channel, petitioned, appealed will be known by the community through social media, forums, etc.

E] - After the core of my developmental work is done, I will then be free to work on my education/exchange company, to educate real people about crypto-currencies. We will be using standard marketing techniques to let people know about classes that will be held for free. These classes will be taught by me - face to face - to help average everyday people get use of crypto. This will work as an entry point through Coinshield, as they will have the protection and advice to know what investments to go to/stay away from. I have already begun this, found a couple decent people that are starting to spread the word, etc. This will grow in organization.

F] - During this process, a money transmitting license will be getting obtained by me, in order to start the Coinshield Exchange. I will be developing a payment database system that will ride on top of the Coinshield Network to then allow these average people to utilize something as simple as a debit card to use their crypto-currencies, while allowing the Exchange System to keep track of all their tax reports, etc for the IRS. Through the education process and peer to peer marketing, we will be working on getting businesses to accept this currency, face to face, person to person. The biggest problem I've seen thus far from talking to average people about Crypto is they don't understand it, so this will hopefully bridge this.

My biggest goal here, is to relate to the everyday people, and help them secure their money outside of FIAT. I spent a long time ensuring Coinshield could act as a stable currency with the Time-Released Supply, Upgraded Security [in light of quantum computing], Unified Time, etc. The internet is going to be a section of our overall marketing, Crypto needs some outside investors. Keep in mind circumstances do shift around the time-lines, so don't take these timelines exactly literally. This is an objective breakdown, with optimal timelines


Unix Timestamps for Coinshield Activations:

1412964000 - 1.0.1 Updates go live at this Timestamp. Version 1 blocks will be rejected after this update. Have your binaries updated by then. Will release binaries in next day or two.

1413914400 - GPU Channel Launch / Version 2 GPU blocks will begin to be accepted by the Coinshield Network

1415124000 - POS Channel Activation - Proof of Stake blocks will begin to be accepted by the Coinshield Network. Will require new binaries before the Timestamp. These will be released one week in advance.


@Viz, whats your marketing plan? We need to build a strong community if you want this project succedeed

In the next few hours the forum (coinshieldtalk.org) will be going offline for some updates.

Unlike other coins, CoinShield's value directly stems from our community. The more people vote, comment, create petitions, and exchange shitcoins for CSD the more CSD's value will rise.

CoinShield is the first ALT. coin made for miners and investors who are tried of owning non-Verified coins, supporting non-Verified coins, being scammed by non-Verified coins, and losing money on non-Verified coins, Our internet marketing won't focus on trying to persuade the crypto community that we are the best non-centralized currency in the world (we all know that is bitcoin). Instead it will be focusing on introducing CoinShield to:
  • the entire mining community
  • all the pool operators
  • all the ALT. coin investors

Once CoinShield is fully operational (all channels are launched, voting and exchange are ready) we will be buying adspace on various sites (bitcointalk, coinmarketcap, etc.) to help inform the community. We may incentivise users to add CoinShield on their signature here on bitcointalk. Currently we are building our followers on twitter to let the community know what coins are being killed in real time (follow us now if you currently do not: https://twitter.com/CoinShieldTalk).

I feel our job is to inform the community and keep them up to date with what we are killing and what is being voted on. Internet Marketing that focuses on anything else is just to create hype and make a bubble. We want CoinShield to be a stable economy where wealth can be built while clones/scams can be killed.

With this said, If you have any ideas on what else you think we should do please PM me or post here.

Do realize it takes time for things to grow properly and Coinshield is 15 days, 18 hours, 49 minutes old
Viz.

legendary
Activity: 1559
Merit: 1001
Epic Private Internet Cash

Shkembe,

As I've mentioned, the best source of information is coinshieldtalk.org. You will find fresh updates on the chatbox, I am rarely here on Bitcointalk anymore. Pool is almost done, in final testing phases. Keep in mind I cannot take any existing pool software to build a Coinshield pool, so I am coding it from scratch. This doesn't happen overnight.

Exchanges will be given source code upon their interest to add us, nothing more.
Viz.


At coinshieldtalk.org you have 92 posts vs 1270 here, I think you can take us more serious. As I see you have no intention to promote your coin and perhaps you expect help from the comunity to focus exchange's interest to it. If yes just tell us. Personally I am thinking to abandon this coin.
legendary
Activity: 868
Merit: 1058
Creator of Nexus http://nexus.io
Work on a GPU miner for the GPU channel would seem more fitting as logic goes [GPU Channel Launch in ~2 weeks]. I'm sure the little CPU miners with few cores such as myself would appreciate a reasonable difficulty level for their amount of computing power.

Not all of us have Supercomputers
Viz.
sr. member
Activity: 278
Merit: 250
@Videlicet

Thanks for the info. Another method is to add the code below to the header file and then replace all instances of unsigned long with word_t in the source files.

#ifdef _WIN64
  typedef unsigned long long int word_t;
#else
  typedef unsigned long int word_t;
#endif


FYI, I will release one more enhancement to the CPU code before I begin work on an open source GPU miner for the CPU channel before year end. I know as a fact that a high-end GPU will be at least 12x faster than the best desktop CPU.
legendary
Activity: 868
Merit: 1058
Creator of Nexus http://nexus.io
Supercomputing,

I found out why it is missing shares. You are updating the nNonce with mpz_get_ui [unsigned long] when the nNonce is [unsigned long long]. Linux interprets this as 8 bytes, but not Windows 7, this give false primes.

uint64 mpz2uint64(mpz_t n)
{
    unsigned int lo, hi;
    mpz_t tmp;

    mpz_init( tmp );
    mpz_mod_2exp( tmp, n, 64 );

    lo = mpz_get_ui( tmp );
    mpz_div_2exp( tmp, tmp, 32 );
    hi = mpz_get_ui( tmp );

    mpz_clear( tmp );

    return (((uint64)hi) << 32) + lo;
}


This did the trick for me.

-------------------------------

Shkembe,

As I've mentioned, the best source of information is coinshieldtalk.org. You will find fresh updates on the chatbox, I am rarely here on Bitcointalk anymore. Pool is almost done, in final testing phases. Keep in mind I cannot take any existing pool software to build a Coinshield pool, so I am coding it from scratch. This doesn't happen overnight.

Exchanges will be given source code upon their interest to add us, nothing more.
Viz.
sr. member
Activity: 278
Merit: 250
sometimes miner loses wallet connection
v1.2 miner, 1.0.1 core
[
[METERS] 5 Block(s) ACC=3 REJ=1| Height = 13917 | Diff = 6.051853 | 20:21:51
[METERS] 5 Block(s) ACC=3 REJ=1| Height = 13917 | Diff = 6.051853 | 20:21:51

[METERS] 5 Block(s) >  what that mean : did you find 5 bloks ?
Yes that's what that means, he found 5 blocks.  Though I'm curious as to why there are 3 accepted and 1 rejected... that makes 4 Smiley

...I'm curious as to why there are 3 accepted and 1 rejected... that makes 4

5 blocks were found, 3 of which were accepted and 1 explicitly rejected. The missing block was implicitly rejected to prevent orphan blocks. What happened is that at least two blocks were found on the network within one second of each other and before one was submitted, the miner received a new block height from the wallet and started to work on that new block height.

I will add a counter to the next version to make the stats easier to understand.
sr. member
Activity: 278
Merit: 250
Version 1.2

Tested on Linux.
Can be ported to Windows with minimal effort.

https://www.dropbox.com/s/j8m9a9cgipraaa8/CoinShieldV1.2.zip?dl=0

Greater than 4x increase in performance.
Drastically reduces the number of calls to wallet server for new block hashes - a single wallet can now handle hundreds of mining threads.
Added support for prime septuplets and octuplets with prime cluster stats.

Awesome, 4x performance so this is who is taking all the blocks.  Grin
how do you know it 4 time faster ?
how i can see speed ? or hash rate ?

how do you know it 4 time faster ?

Run a benchmark on the same hardware at the same base difficulty, then compare the results.

how i can see speed ? or hash rate ?

View the Prime Clusters Found statistics: 19 blocks found, 19 accepted, 0 explicitly rejected in approximately 14 hours and 21 minutes.

[METERS] 19 Block(s) ACC=19 REJ=0 | Height = 14077 | Diff = 6.094174 | 14:21:41

  Prime Clusters Found: Four=29323 | Five=1055 | Six=17 | Seven=0 | Eight=0

[METERS] 19 Block(s) ACC=19 REJ=0 | Height = 14077 | Diff = 6.094174 | 14:21:52
Jump to: