[XPM] [ANN] Primecoin Release - First Scientific Computing Cryptocurrency - page 161.

oroqen

sr. member

Activity: 280

Merit: 250

Quote from: mustyoshi on July 09, 2013, 08:52:14 AM

Quote from: eule on July 09, 2013, 08:44:23 AM

nMaxSieveSize = 1000000u ~130 - 230 pps
nMaxSieveSize = 100000u ~230 - 400 pps
nMaxSieveSize = 10000u ~ 350 - 800 pps

Running the second one now, no idea if this will make me mine more blocks or none. Also not sure about the other thing you mentioned, goes above my head. Grin

I'm not sure that reducing the sieve size is the best optimization, since by reducing it you also reduce the amount of potential numbers you are testing for the PoW...

Who knows what that actually does to the potential of finding a correct PoW though.

Running different Sieve values per thread rather than the current all threads same Sieve, could spread out that risk, even if it's just 1 out of 4 running at 10k and the other 3 at 1M

mustyoshi

sr. member

Activity: 287

Merit: 250

Quote from: eule on July 09, 2013, 08:44:23 AM

nMaxSieveSize = 1000000u ~130 - 230 pps
nMaxSieveSize = 100000u ~230 - 400 pps
nMaxSieveSize = 10000u ~ 350 - 800 pps

Running the second one now, no idea if this will make me mine more blocks or none. Also not sure about the other thing you mentioned, goes above my head. Grin

I'm not sure that reducing the sieve size is the best optimization, since by reducing it you also reduce the amount of potential numbers you are testing for the PoW...

Who knows what that actually does to the potential of finding a correct PoW though.

nmersulypnem

full member

Activity: 238

Merit: 100

Quote from: maka on July 09, 2013, 08:39:44 AM

Quote from: nmersulypnem on July 09, 2013, 08:30:41 AM

Quote from: maka on July 09, 2013, 08:08:02 AM

Quote from: eule on July 09, 2013, 07:43:33 AM

I thought difficulty will rise a lot when GPU miners appear?

The prime number searching range for Primecoin is roughly 2^100 ~ 2^2000 ( 100 digits ~ 2000 digits ). For the sake of simplicity, let's assume the primes are about 2^100. According to the Prime number theorem, the prime number distribution of this range is about 1 prime number in 230 integers. When difficulty grows from 7 to 8, it means the miner needs to find a prime chain of length 8 instead of 7, making the likelihood to 1/230 of before. In other word, it takes 230 times longer to find a 8-chain than a 7-chain if we assume the mining power remain the same. It would take 230*230*230 = 12167000 times longer to find a 10-chain, that's 10 million times more difficult.

Now let's take the GPU and other possible device into account. The GPU is very efficient of calculating hash, but in the specific case of primecoin, it doesn't seem have the same kind of advantage. Unless there are some specially designed hardware for prime searching, the mining speed at most will be less than 1000x faster.

So it is quite clear that even with 10000x more miners equipped with monster machines, it is still not even close to reach difficulty 10.

Where do you see that range in the code? I don't see it...

static const CBigNum bnPrimeMax = (bnOne << 2000) - 1;
static const CBigNum bnPrimeMin = (bnOne << 255);

I think an FPGA would be perfect for this...

eule

hero member

Activity: 756

Merit: 501

nMaxSieveSize = 1000000u ~130 - 230 pps
nMaxSieveSize = 100000u ~230 - 400 pps
nMaxSieveSize = 10000u ~350 - 800 pps

Running the second one now, no idea if this will make me mine more blocks or none. Also not sure about the other thing you mentioned, goes above my head. Grin

maka

newbie

Activity: 54

Merit: 0

Quote from: nmersulypnem on July 09, 2013, 08:30:41 AM

Quote from: maka on July 09, 2013, 08:08:02 AM

Quote from: eule on July 09, 2013, 07:43:33 AM

I thought difficulty will rise a lot when GPU miners appear?

The prime number searching range for Primecoin is roughly 2^100 ~ 2^2000 ( 100 digits ~ 2000 digits ). For the sake of simplicity, let's assume the primes are about 2^100. According to the Prime number theorem, the prime number distribution of this range is about 1 prime number in 230 integers. When difficulty grows from 7 to 8, it means the miner needs to find a prime chain of length 8 instead of 7, making the likelihood to 1/230 of before. In other word, it takes 230 times longer to find a 8-chain than a 7-chain if we assume the mining power remain the same. It would take 230*230*230 = 12167000 times longer to find a 10-chain, that's 10 million times more difficult.

Now let's take the GPU and other possible device into account. The GPU is very efficient of calculating hash, but in the specific case of primecoin, it doesn't seem have the same kind of advantage. Unless there are some specially designed hardware for prime searching, the mining speed at most will be less than 1000x faster.

So it is quite clear that even with 10000x more miners equipped with monster machines, it is still not even close to reach difficulty 10.

Where do you see that range in the code? I don't see it...

static const CBigNum bnPrimeMax = (bnOne << 2000) - 1;
static const CBigNum bnPrimeMin = (bnOne << 255);

mustyoshi

sr. member

Activity: 287

Merit: 250

Quote from: ReCat on July 09, 2013, 08:24:58 AM

Quote from: mustyoshi on July 09, 2013, 08:18:17 AM

Left my somewhat optimized version running over night and came back to 3 additional blocks mined.
Getting roughly ~60 pps with 500K sieve size and basic optimization flags. Running on a single thread of a stock i3 2120.

Dude. Post it. Keeping optimized miners secret is a bad move for your rep and the coin. Don't make me get my own optimized one to post.

Other than the sieve size change. I also opted to pass the blockheader hash through the functions, instead of making calls to block->GetHeaderHash() every time. I'm not sure what kind of improvements that actually made since I also changed the sieve size when I first did it.

But it stops me from running double SHA256 like 3 times more than I used to.

nmersulypnem

full member

Activity: 238

Merit: 100

Quote from: maka on July 09, 2013, 08:08:02 AM

Quote from: eule on July 09, 2013, 07:43:33 AM

I thought difficulty will rise a lot when GPU miners appear?

The prime number searching range for Primecoin is roughly 2^100 ~ 2^2000 ( 100 digits ~ 2000 digits ). For the sake of simplicity, let's assume the primes are about 2^100. According to the Prime number theorem, the prime number distribution of this range is about 1 prime number in 230 integers. When difficulty grows from 7 to 8, it means the miner needs to find a prime chain of length 8 instead of 7, making the likelihood to 1/230 of before. In other word, it takes 230 times longer to find a 8-chain than a 7-chain if we assume the mining power remain the same. It would take 230*230*230 = 12167000 times longer to find a 10-chain, that's 10 million times more difficult.

Now let's take the GPU and other possible device into account. The GPU is very efficient of calculating hash, but in the specific case of primecoin, it doesn't seem have the same kind of advantage. Unless there are some specially designed hardware for prime searching, the mining speed at most will be less than 1000x faster.

So it is quite clear that even with 10000x more miners equipped with monster machines, it is still not even close to reach difficulty 10.

Where do you see that range in the code? I don't see it...

eule

hero member

Activity: 756

Merit: 501

He did:

Quote from: mustyoshi on July 08, 2013, 09:38:57 PM

Has anybody else tried messing with the nMaxSieveSize variable?
By chopping a zero off of it, I seemed to be able to get almost a 5x increase in pps. I know that reducing the sieve size would effectively reduce the amount of numbers checked for each run of the sieve and then reduce the likelyhood of finding a block, but is pps more important than search size?

ReCat

sr. member

Activity: 406

Merit: 250

Quote from: mustyoshi on July 09, 2013, 08:18:17 AM

Left my somewhat optimized version running over night and came back to 3 additional blocks mined.
Getting roughly ~60 pps with 500K sieve size and basic optimization flags. Running on a single thread of a stock i3 2120.

Dude. Post it. Keeping optimized miners secret is a bad move for your rep and the coin. Don't make me get my own optimized one to post.

eule

hero member

Activity: 756

Merit: 501

Quote from: mustyoshi on July 09, 2013, 08:18:17 AM

Left my somewhat optimized version running over night and came back to 3 additional blocks mined.
Getting roughly ~60 pps with 500K sieve size and basic optimization flags. Running on a single thread of a stock i3 2120.

I tried ommiting two zeroes from nMaxSieveSize and got ~400-800 pps using 6 cores of an old i7, quickly changed back to normal version though due to not finding blocks in a seconds interval. Roll Eyes

Now i fell i must try it again, either with one or two zeroes less than normal. Grin

mustyoshi

sr. member

Activity: 287

Merit: 250

Left my somewhat optimized version running over night and came back to 3 additional blocks mined.
Getting roughly ~60 pps with 500K sieve size and basic optimization flags. Running on a single thread of a stock i3 2120.

maka

newbie

Activity: 54

Merit: 0

Quote from: eule on July 09, 2013, 08:11:22 AM

@ maka: Thanks for that explanation, makes sense! The quite quickly sinking block reward made me think that.

You are very welcome. Yeah, I didn't mined a single block in the last 24 hours Angry

eule

hero member

Activity: 756

Merit: 501

Quote from: ReCat on July 09, 2013, 08:13:07 AM

What will the value of this coin be with how few of them are in existence?

Have a look at the two forum exchanges:
https://bitcointalksearch.org/topic/xpm-primecoin-ltc-google-docs-exchange-252611
https://bitcointalksearch.org/topic/xpm-to-ltc-exchange-launched-with-escrow-252875

I think we're still in a price finding phase, although it's clear no one wants to get rid of their precious primecoins for pocket money.

ReCat

sr. member

Activity: 406

Merit: 250

What will the value of this coin be with how few of them are in existence?

eule

hero member

Activity: 756

Merit: 501

@ maka: Thanks for that explanation, makes sense! The quite quickly sinking block reward made me think that.

Boing7898

sr. member

Activity: 686

Merit: 259

I was getting 100-150 primespersec yesterday with my Phenom X4 965 but today I'm getting barely 50. Did something change? o.o

Vorksholk

legendary

Activity: 1713

Merit: 1029

Upping personal bounty for a standalone cpu-miner that is as fast OR faster than built-in primecoin one, and is capable of working with pools and getting shares to 2.5BTC. Must provide windows and linux binaries to the public, as well as source code.

maka

newbie

Activity: 54

Merit: 0

Quote from: eule on July 09, 2013, 07:43:33 AM

I thought difficulty will rise a lot when GPU miners appear?

The prime number searching range for Primecoin is roughly 2^255 ~ 2^2000 ( 76 digits ~ 602 digits ). For the sake of simplicity, let's assume the primes are about 10^100. According to the Prime number theorem, the prime number distribution of this range is about 1 prime number in 230 integers. When difficulty grows from 7 to 8, it means the miner needs to find a prime chain of length 8 instead of 7, making the likelihood to 1/230 of before. In other words, it takes 230 times longer to find a 8-chain than a 7-chain if we assume the mining power remain the same. It would take 230*230*230 = 12167000 times longer to find a 10-chain, that's 10 million times more difficult.

Now let's take the GPU and other possible device into account. The GPU is very efficient of calculating hash, but in the specific case of primecoin, it doesn't seem have the same kind of advantage. Unless there are some specially designed hardware for prime searching, the mining speed at most will be less than 1000x faster.

So it is quite clear that even with 10000x more miners equipped with monster machines, it is still not even close to reach difficulty 10.

simondlr

sr. member

Activity: 424

Merit: 250

Sunny, I have a question related to the scientific benefit of this altcoin. I did some further reading on what is being generated and on cunningham chains. It's a very rusty initial dive into this, so forgive me if I am wrong.

Primecoin won't really find new primes (for now at least), since they are so large to reasonably process in a wallet, and they would be at the end of longer chains, rather than the current length of chains (might be wrong assumptions here)?

The idea is to find plenty variations of cunningham chains in order to hopefully better understand the distribution of primes, and whether they could eventually give clues to the distribution of primes in general?

techbytes

legendary

Activity: 1694

Merit: 1054

Point. Click. Blockchain

Quote from: Boing7898 on July 09, 2013, 06:39:47 AM

Quote from: techbytes on July 09, 2013, 06:38:46 AM

Quote from: maco on July 09, 2013, 04:32:17 AM

Does anyone recommend overclocking a CPU? This is my processor: http://ark.intel.com/products/52210/Intel-Core-i5-2500K-Processor-6M-Cache-up-to-3_70-GHz

It's currently at 3.3 GHz and the Max Turbo Frequency 3.7 GHz

Anyone recommend boosting it for the Max Turbo? and do you believe that will help?

I wouldn't bother overclocking just yet. My older laptop is finding more blocks than my i7 desktop. I'm going to fireup my intel atom and see what it can do. Grin

-tb-

I tried with my atom but it kept showing 0 primespersec, even after 30 minutes.
If you get it working, tell me!

I'm getting 0 PPS with my intel atom as well but cpu is pegged at 100% so it is doing something. I'm going to let it run and will let you know if and when it gets a block.

-tb-

Topic: [XPM] [ANN] Primecoin Release - First Scientific Computing Cryptocurrency - page 161. (Read 688812 times)