1697 replies to this topic

[Foogles]

Just a thought, what if you were to instead of running one process on, say, 4 threads, ran 4 different mining processes on 1 thread each using different nonces? Would you expect similar output, better, or worse?

[Anavrins]

Foogles, on 22 January 2016 - 02:36 AM, said:

Just a thought, what if you were to instead of running one process on, say, 4 threads, ran 4 different mining processes on 1 thread each using different nonces? Would you expect similar output, better, or worse?

This is what YTCIKristMiner does, it is of course better, it's actually how it should be done, otherwise you'll be using 4 threads with the performance of only 1.

Edited by Anavrins, 22 January 2016 - 02:42 AM.

[Foogles]

Hmm, interesting. The reason I ask is because I am using YTCIKristMiner with 2.2ghz clock speed on 8 cores, and getting 1.4 MH/S, which is what I was getting on my other computer with 1.9ghz clock speed on 2 cores...

[apemanzilla]

Anavrins, on 22 January 2016 - 02:41 AM, said:

Foogles, on 22 January 2016 - 02:36 AM, said:

Just a thought, what if you were to instead of running one process on, say, 4 threads, ran 4 different mining processes on 1 thread each using different nonces? Would you expect similar output, better, or worse?

This is what YTCIKristMiner does, it is of course better, it's actually how it should be done, otherwise you'll be using 4 threads with the performance of only 1.

Err, what? I don't know where you're coming up with this information. Using one Thread per core in Java will be able to fully utilize all cores available to the CPU. Separating them out into their own processes isn't much different, and just adds overhead.

[Foogles]

Well I can tell someone just started mining again

[Anavrins]

apemanzilla, on 22 January 2016 - 03:12 AM, said:

Err, what? I don't know where you're coming up with this information. Using one Thread per core in Java will be able to fully utilize all cores available to the CPU. Separating them out into their own processes isn't much different, and just adds overhead.

What he's saying is if using different nonce for each of the threads matter, it does.
And YTCI miner does that https://github.com/Y...Miner.java#L241

[apemanzilla]

Anavrins, on 22 January 2016 - 03:18 AM, said:

apemanzilla, on 22 January 2016 - 03:12 AM, said:

Err, what? I don't know where you're coming up with this information. Using one Thread per core in Java will be able to fully utilize all cores available to the CPU. Separating them out into their own processes isn't much different, and just adds overhead.

What he's saying is if using different nonce for each of the threads matter, it does.
And YTCI miner does that https://github.com/Y...Miner.java#L241

Yes, you have to make sure that each thread ends up hashing different nonces. In the case of turbokrist (My GPU miner) it's as simple as just incrementing the nonce by the number of threads running instead of by 1.

[bauen1]

I mined my first block using a cpu miner after 9 hours (i only mined at 0.9MH/s to keep my fan off)
Edit: second block, i'm literally the only one mining

Edited by bauen1, 23 January 2016 - 11:38 AM.

[ry00000]

I've got a Core 2 Quad with a GeForce GTX 920. How many Krist am I going to get? Also, with 8 gigabytes of ram.

[bauen1]

ry00000, on 23 January 2016 - 12:42 PM, said:

I've got a Core 2 Quad with a GeForce GTX 920. How many Krist am I going to get? Also, with 8 gigabytes of ram.

Run YTCIKristMiner (search this topic for the link) and the higher the hash rate the better your chances are
Post it and i can tell you what you might get

Edited by bauen1, 23 January 2016 - 12:48 PM.

[ry00000]

Anybody have a Lua api for Krist ready? I MIGHT be able to translate Apemanzilla's Java API in lua but I'm not really sure.

[ry00000]

I've just made a public Krist address, if anyone wants to use it, the address is kpx6e6td32 while the password is "public" without quotes.

Hey apeman I challenge you to bruteforce the address: kdrwv2ralf I just mashed keys on the keyboard to create a password.

There's no Krist in it, I forgot to mention

I can't even get 1 MH/s on this computer! (Not mine, my grandma's)

I get like 0.809 MH/s with YTCI!

[bauen1]

ry00000, on 23 January 2016 - 01:08 PM, said:

Anybody have a Lua api for Krist ready? I MIGHT be able to translate Apemanzilla's Java API in lua but I'm not really sure.

https://github.com/a...er/kristapi.lua dunno if it works

[ry00000]

My PC is much more powerful, I should be able to solve a few blocks on mine

[bauen1]

ry00000, on 23 January 2016 - 01:18 PM, said:

My PC is much more powerful, I should be able to solve a few blocks on mine

Whats your CPUs clock speed ? i have the almost same specs as you with a clock speed of 1,7 GHz

[ry00000]

Please add in the password functions the Java API had. This is my only API request for now. I need them for... er... a.. An automatic Krist account generator!

I think it's also 1.7 GHz.

Also, there's an address called literally koalazarsq, this is the actual address and it has a TON of krist.

[bauen1]

ry00000, on 23 January 2016 - 01:22 PM, said:

Please add in the password functions the Java API had. This is my only API request for now. I need them for... er... a.. An automatic Krist account generator!

I think it's also 1.7 GHz.

Also, there's an address called literally koalazarsq, this is the actual address and it has a TON of krist.

If your gonna try to brute force passwords, do it outside of cc

[ry00000]

well why not? It's a sort of programming challenge for me, do a KST bruteforcer inside CC!
It's a little hard to actually implement it.

[bauen1]

ry00000, on 23 January 2016 - 01:26 PM, said:

well why not? It's a sort of programming challenge for me, do a KST bruteforcer inside CC!
It's a little hard to actually implement it.

I used that api and let me tell you the brute force program is less than 10 lines of code but it is slow, very slow.

[ry00000]

Just give me the API.
I'll handle everything myself.