[ANN] [SUP] Supcoin, PLUCK algorithm, Just Launched, Mine Now!

[Nthused]

From the OP:

"I've been working on a big coin for the past month, and it's been on my mind for over a year now with all the ideas for it finally having a solid solution and plan to accomplish it. Sadly, Supcoin is not that coin. Basically, I need a coin to test some things, and that is this coin!"

So guys, this is a test coin? When will the main coin release?

It's still a coin by earlz. One of the best programmers with the most knowledge of the bitcoin code out there working on altcoins.

This is what the coin is all about:

Supcoin



Premise

Hello,

I've been working on a big coin for the past month, and it's been on my mind for over a year now with all the ideas for it finally having a solid solution and plan to accomplish it. Sadly, Supcoin is not that coin. Basically, I need a coin to test some things, and that is this coin! I'm not ready to discuss my big coin project just yet (there will be a pre-ann in a couple months), but this coin is my way of making sure that the codebase I'm using is good.

Features:

• C'mon, you've always wanted to greet someone by giving them some SUP
• Based on bitcoin 0.10 (includes headers-first, which significnalty speeds up syncing times)
• 5 minute blocks
• Block reward halves every 100 days, starting at 10 coins per block
• PLUCK, memory hard hashing algorithm based on SHA256, using 128Kb of memory
• Digishield difficulty adjustment
• addrindex support (use -addrindex to the daemon and you can use RPC calls to query the balance of addresses on the blockchain)
• OP_CHECKLOCKTIMEVERIFY support, for advanced timelock contracts (from Viacoin)
• All BIPS enabled, including BIP66
• Traditional pure PoW
• Reasonably small premine of 100 blocks worth, with bounties
• First of it's kind Proof Of Permission: Whitepaper

Premine breakdown: (1,000 SUP)

• 50 SUP for me, to do whatever with(ie, to sell)
• 350 SUP for a PLUCK Nvidia GPU miner
• 300 SUP for a PLUCK AMD GPU miner
• 50 SUP for a NOMP module compatible with PLUCK(including pull requests and acceptance into the project) --CLAIMED (python module by ocminer: https://github.com/ocminer/pluck-hash )
• 200 SUP for someone who can takeover this coin with a real marketing plan, etc
• 50 SUP for whoever can make a decent logo (see below) -- CLAIMED

Also, I accidentally premined 1 additional block, worth 10 SUP. This will be given to Wolf` as a tip for his work in testing the coin, algorithm, and spotting some huge bugs.

PLUCK:

PLUCK is a configurable hashing algorithm using SHA256. I developed PLUCK because most algorithms are not actually memory hard. They are constructed in such a way that many things can be chosen to not be stored into memory and can instead be recalculated. PLUCK does not suffer from this exploit. See also this article for more explanation.

This is similar to a merkelhash in a way, but is specifically designed for PoW to not allow breaking the memory hardness requirement. ASICs will always be possible, but would require expensive and large memory caches that can not be optimized away in anyway.

The core weakness in scrypt is that most variables stored in memory can easily be calculated more than once to save on memory costs. Although this is in theory less effecient, it ends up making ASICs and such faster by requiring less (expensive) cache memory.

This hashing algorithm is designed so that this can not take place. Recalculating anything stored in memory is as expensive as recalculating all the progress made thusfar on the hash. In the worst case, it may be possible reduce memory costs by 50% by recalculating the first half of the SHA256 hash values, but this would of course have extremely diminishing returns since this requires many SHA256 hashes to be computed.

Normally, I would scoff at the idea of writing my own hashing algorithm. However, I've taken precautions to ensure that this is at least as secure as sha256. Sha256 has a block input size of 512 bits, or 64 bytes. In ALL hashes generated, at least 256 bits of this input consists of an unmodified SHA256 hash generated previously. This ensures that the algorithm carries over SHA256's collision resistance and preimaging protection. Because of this property, I feel it is a safe cryptographic hash. The worst case is that the memory-hardness aspect is broken. The hash itself should never be broken (ie, collisions or preimages) unless SHA256 is broken in some way.

Supcoin

So, I'm a pretty busy guy. I'm still working a lot on the coin that comes after this. However, I intend to do maintenance on Supcoin at least until that coin is released. Supcoin should ideally be trivial to maintain as I strive to write coins that don't need forks and such to fix things. However, any issues that come up I will fix. I will not be adding gimmicky features however. I might add some non-forking features eventually just to test them out on a live network, but I'm not making any promises on that.

I will not do any kind of coin swap or anything like that when I release the next coin I'm working on, however, I might offer some kind of rewards for people that held some amount at some point in time, as an additional distribution mechanism.

Anyway, I'm a developer, not a marketer or community manager or anything like that. If you are someone who could do a "takeover" of the project, then have at it! I'll be holding onto 200 SUP to reward for a successful community takeover with a leader that proves they aren't just trying to do a pump and dump (ie, after some amount of time)

GPU MINER BOUNTY
In order to raise tangible funds to motivate GPU miner devs, I am now hosting a donation address for Bitcoin.

1K6cUbtsn5xinD8XfjnTn7UnsMriLhZaGD

The first GPU miner to be created (AMD or Nvidia) and published under an open source license will receive all the money from this fund.

I'm putting in 0.25 BTC of my own money. I encourage everyone to donate what you can to help make this a lucrative bounty.

[scamkiller]

AMD gpu miner plzzzzzzzzzzzzzzzzzzzzz!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

[kitaco]

AMD gpu miner plzzzzzzzzzzzzzzzzzzzzz!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Then the nethash will be 50000KH/s as every farm around the world spools its AMD's up, completly putting little miners out of business.

[Enzyme]

Hey,

Settings: ccminer.exe -a pluck -o stratum+tcp://sup.suprnova.cc:7777 -u Tempest.1 -p x
My miner has a very high hash and nothing gets accepted or declined.

Help!

[kitaco]

Hey,

Settings: ccminer.exe -a pluck -o stratum+tcp://sup.suprnova.cc:7777 -u Tempest.1 -p x
My miner has a very high hash and nothing gets accepted or declined.

Help!

You're not using ccminer30.exe

Look for djm34's new 3.0 binaries in this thread.

[Enzyme]

Hey,

Settings: ccminer.exe -a pluck -o stratum+tcp://sup.suprnova.cc:7777 -u Tempest.1 -p x
My miner has a very high hash and nothing gets accepted or declined.

Help!

You're not using ccminer30.exe

Look for djm34's new 3.0 binaries in this thread.

I tried ccminer30.exe and my kernel driver crashes.
Do I need to close as many programs as possible?

GTX 760

[kitaco]

Hey,

Settings: ccminer.exe -a pluck -o stratum+tcp://sup.suprnova.cc:7777 -u Tempest.1 -p x
My miner has a very high hash and nothing gets accepted or declined.

Help!

You're not using ccminer30.exe

Look for djm34's new 3.0 binaries in this thread.

I tried ccminer30.exe and my kernel driver crashes.
Do I need to close as many programs as possible?

GTX 760

keep changing --throughput to a lower value until it no longer crashes. Try 4 and go down. My 670 only likes --throughput 1.15

[Enzyme]

Hey,

Settings: ccminer.exe -a pluck -o stratum+tcp://sup.suprnova.cc:7777 -u Tempest.1 -p x
My miner has a very high hash and nothing gets accepted or declined.

Help!

You're not using ccminer30.exe

Look for djm34's new 3.0 binaries in this thread.

I tried ccminer30.exe and my kernel driver crashes.
Do I need to close as many programs as possible?

GTX 760

keep changing --throughput to a lower value until it no longer crashes. Try 4 and go down. My 670 only likes --throughput 1.15

I've mined other GPU coins and many aren't as laggy.

Is this algo very intensive?

[kitaco]

Hey,

Settings: ccminer.exe -a pluck -o stratum+tcp://sup.suprnova.cc:7777 -u Tempest.1 -p x
My miner has a very high hash and nothing gets accepted or declined.

Help!

You're not using ccminer30.exe

Look for djm34's new 3.0 binaries in this thread.

I tried ccminer30.exe and my kernel driver crashes.
Do I need to close as many programs as possible?

GTX 760

keep changing --throughput to a lower value until it no longer crashes. Try 4 and go down. My 670 only likes --throughput 1.15

I've mined other GPU coins and many aren't as laggy.

Is this algo very intensive?

Lags my PC a lot, but I'm not using the mining rig except for...mining. Not really creating much heat either.

[logocreator]

i think i am mining something, on 2 pc's,

[kitaco]

Stop selling 1-2 coins for dirt cheap. This coin should be worth at least .01 PER. Make it happen guys!

[tamara163]

Hey Wolf, how much faster is your optimized CPU miner?

If I wanted to learn how to make optimized miners where would I start? Assuming I know barely anything about coding but am eager to start.

I ran some quick tests on it to get current numbers - today I managed to improve SHA256 a ton, raising the stock miner hash by 32% by itself while using no special CPU features. Anyway, the following results are the stock cpuminer that Earlz wrote from his github, compiled with -march=native and -Ofast, versus my current working build compiled with the same, run on my i7-5930K (OC'd to 4.65Ghz) with 12 threads:

Stock (NSFW): https://ottrbutt.com/miner/pluckhashstock-02202015-2.png
Modified (NSFW): https://ottrbutt.com/miner/pluckhashwolf-02202015.png

The only processor features /explicitly/ used by my code is SSE2, which all 64-bit x86 CPUs are guaranteed to have. That is not to say the compiler didn't use others - but it had the same opportunity for the stock miner.

If you want to learn more about miners, optimizations, and lower level coding, I recommend you learn C, which is what I started with.

And.. I'm sorry, but I have to recommend you also become pretty good in x86 assembly. It's not easy, but it gives you an in-depth understanding of your target platform, and even if you don't use it often, it will help you write better code.

where to download optimized CPU miner?

[kitaco]

Hey Wolf, how much faster is your optimized CPU miner?

If I wanted to learn how to make optimized miners where would I start? Assuming I know barely anything about coding but am eager to start.

I ran some quick tests on it to get current numbers - today I managed to improve SHA256 a ton, raising the stock miner hash by 32% by itself while using no special CPU features. Anyway, the following results are the stock cpuminer that Earlz wrote from his github, compiled with -march=native and -Ofast, versus my current working build compiled with the same, run on my i7-5930K (OC'd to 4.65Ghz) with 12 threads:

Stock (NSFW): https://ottrbutt.com/miner/pluckhashstock-02202015-2.png
Modified (NSFW): https://ottrbutt.com/miner/pluckhashwolf-02202015.png

The only processor features /explicitly/ used by my code is SSE2, which all 64-bit x86 CPUs are guaranteed to have. That is not to say the compiler didn't use others - but it had the same opportunity for the stock miner.

If you want to learn more about miners, optimizations, and lower level coding, I recommend you learn C, which is what I started with.

And.. I'm sorry, but I have to recommend you also become pretty good in x86 assembly. It's not easy, but it gives you an in-depth understanding of your target platform, and even if you don't use it often, it will help you write better code.

where to download optimized CPU miner?

I'm obviously not going to dump the result of a lot of work on the forum so that the diff goes even higher...

Yep, releasing it makes the optimizations completely useless.

[earlz]

Hey Wolf, how much faster is your optimized CPU miner?

If I wanted to learn how to make optimized miners where would I start? Assuming I know barely anything about coding but am eager to start.

I ran some quick tests on it to get current numbers - today I managed to improve SHA256 a ton, raising the stock miner hash by 32% by itself while using no special CPU features. Anyway, the following results are the stock cpuminer that Earlz wrote from his github, compiled with -march=native and -Ofast, versus my current working build compiled with the same, run on my i7-5930K (OC'd to 4.65Ghz) with 12 threads:

Stock (NSFW): https://ottrbutt.com/miner/pluckhashstock-02202015-2.png
Modified (NSFW): https://ottrbutt.com/miner/pluckhashwolf-02202015.png

The only processor features /explicitly/ used by my code is SSE2, which all 64-bit x86 CPUs are guaranteed to have. That is not to say the compiler didn't use others - but it had the same opportunity for the stock miner.

If you want to learn more about miners, optimizations, and lower level coding, I recommend you learn C, which is what I started with.

And.. I'm sorry, but I have to recommend you also become pretty good in x86 assembly. It's not easy, but it gives you an in-depth understanding of your target platform, and even if you don't use it often, it will help you write better code.

where to download optimized CPU miner?

Wolf0 is keeping his changes private for now at least. (maybe he'd sell them to you? idk). However, I'm working on improving the CPU miner as well. I've pushed a fairly trivial optimization to github that increases hashrates by about 15% in my testing.

[fmz89]

AMD gpu miner plzzzzzzzzzzzzzzzzzzzzz!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Then the nethash will be 50000KH/s as every farm around the world spools its AMD's up, completly putting little miners out of business.

yeah absolutely, but isnt fair, nvidia miner very happy now ,nethash 1,2mhs now
why not cpu miner only,thats fair

[kitaco]

AMD gpu miner plzzzzzzzzzzzzzzzzzzzzz!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Then the nethash will be 50000KH/s as every farm around the world spools its AMD's up, completly putting little miners out of business.

yeah absolutely, but isnt fair, nvidia miner very happy now ,nethash 1,2mhs now
why not cpu miner only,thats fair

Think about how I feel only having NVIDIA cards when new coins launch where AMD's outperform by miles . Quiet  

I'm getting 2.5x as much on my CPU thn NVIDIA card anyway.

[fmz89]

AMD gpu miner plzzzzzzzzzzzzzzzzzzzzz!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Then the nethash will be 50000KH/s as every farm around the world spools its AMD's up, completly putting little miners out of business.

yeah absolutely, but isnt fair, nvidia miner very happy now ,nethash 1,2mhs now
why not cpu miner only,thats fair

Think about how I feel only having NVIDIA cards when new coins launch where AMD's outperform by miles . Quiet  

I'm getting 2.5x as much on my CPU thn NVIDIA card anyway.

i'm getting 0.8khs with fx6300@4ghz
just have 1 pc at home run24/7 0.8khs
please give me wolf0 optimization
i hope no amd miner even i have amd gpu, nethash will blow away,maybe the price will down to

[scorpio18]

Spread the hash guys

[NightShade00013]

Spread the hash guys

I wish. . . Rnova is running even more workers now than last night.

I am getting about 1.1Kh on a fx 8350 throttled up to 4.3Ghz temps around 49 with a evo 212 hyper while x11 mining with a pair of radeon 6850's that are pulling 1.3Mh, wish I could spin them up on this coin as well but alas the miner is not available yet.

Also have an old dual core Athlon 64 4200+ that is pulling about .20Kh

Kinda sucks mining on a pool that is getting slammed since the other one is running 1300 Kh

[ocminer]

Spread the hash guys

I second this and please do so guys, we need to spread the hashes a little !