Author

Topic: [ANN] cudaMiner & ccMiner CUDA based mining applications [Windows/Linux/MacOSX] - page 622. (Read 3426922 times)

full member
Activity: 199
Merit: 100
The new cgminer-heavycoin made some nice boost...
dga
hero member
Activity: 737
Merit: 511
You think I can adjust lookup gap with a knc titan for instance?

no, but with a specially designed (futureproof) Scrypt-N ASIC. It's absolutely possible to design one.

and this is why I don't like people claiming that Scrypt-N is ASIC-proof. Nothing is ASIC-proof.

Christian


+This.

How many miner devs does it take all saying the same thing before people will believe it? :-)

Implementing nearly any algorithm on an ASIC is merely a matter of time and money.  The hardware surrounding that ASIC may look more (lots of RAM) or less (entirely in SRAM) like a general purpose computer, but for anything you can imagine, there will always be some efficiency gap from a CPU to an ASIC that, if the money is right, will incent someone to do it.
legendary
Activity: 1400
Merit: 1050
if you wait one year you'll get them for 50 dollars...
At least it is one advantage of the gpu over the asic, it is always possible to resell them...
legendary
Activity: 1582
Merit: 1019
011110000110110101110010
I ordered 10 gridseeds for playing...
I have a contact Person in Hongkong who will send them much cheaper as
this stupid eBay or groupbuy prices.

Are you willing to share the contact info or a link to purchase?

member
Activity: 70
Merit: 10
Nothing is ASIC-proof.

Exactly, so many people lose sight of the fact that ASIC is an acronym.

Application Specific Integrated Circuit

Very specifically means it's an Integrated Circuit for Specific Application...

The only reason a coin/algo/mix will ever be "ASIC Proof" is if it flops so hard no one ever even began to care for it and thus never was worth thinking about how to build the IC for it. Smiley
legendary
Activity: 1400
Merit: 1050
I am looking into some new coin which should launch tonight (Madcowcoin: POS+3%premine+IPO (unsuccessul so far... so the dev might just run away with 0.1btc without launching the coin  Grin)
It is a scrypt-jane Nmin=5 and Nmax=30.
We don't have a start time (for now)... What could be used as starttime so that at launch the miners uses N=5 ?
(I assume the starttime is just a way to schedule the change of N factor, I may be wrong though...)
hero member
Activity: 812
Merit: 1000
You think I can adjust lookup gap with a knc titan for instance?

no, but with a specially designed (futureproof) Scrypt-N ASIC. It's absolutely possible to design one.

and this is why I don't like people claiming that Scrypt-N is ASIC-proof. Nothing is ASIC-proof.

Christian


You'd still get a performance hit...and high n factors would take a gazillion gb's, ram wise.
It's gonna be expensive to build it, even if it can do LG.
hero member
Activity: 756
Merit: 502
You think I can adjust lookup gap with a knc titan for instance?

no, but with a specially designed (futureproof) Scrypt-N ASIC. It's absolutely possible to design one.

and this is why I don't like people claiming that Scrypt-N is ASIC-proof. Nothing is ASIC-proof.

Christian
hero member
Activity: 812
Merit: 1000

Only way I can imagine it being feasible is make an asic with more than enough ram for a few n factor changes and then add more ram for the next gen.

LOOKUP_GAP also works in ASICs Wink  ... in principle


You think I can adjust lookup gap with a knc titan for instance?
member
Activity: 112
Merit: 10
I ordered 10 gridseeds for playing...
I have a contact Person in Hongkong who will send them much cheaper as
this stupid eBay or groupbuy prices.
hero member
Activity: 756
Merit: 502

Only way I can imagine it being feasible is make an asic with more than enough ram for a few n factor changes and then add more ram for the next gen.

LOOKUP_GAP also works in ASICs Wink  ... in principle
legendary
Activity: 1400
Merit: 1050
https://bitcointalksearch.org/topic/annscryptadaptive-kmc-kimocoin-kimotocoin-5-exchanges-524216

Good CoiN!

LAUNCH TIME: 1st APRIL 2014

Algorithim : Scrypt-n-Adaptive, Scrypt-ASIC RESISTANT!

My 750ti is ready to mine this .
1st of April seems to be such a bad time to launch a coin (... when already many coins look like bad jokes on normal day...)
sr. member
Activity: 300
Merit: 250
Last chance to vote vor FC... It's currently in 2nd place only...

you think cryptoaltex is the right exchange for fc?
look at the daily trade amount... is that exchange at all?

same can be said about nxt-e Wink

A small step for a coin, but a giant leap for coinkind.



that is why i have never used nxt-e or ptopex.com neither will i use cryptoaltex... one day the coin makes it to some real exchange, till then just mine and hold.
just hope those wouldbe exhanges wont ruin the coin.
hero member
Activity: 812
Merit: 1000
Scrypt-jane doesn't mean it has a variable N factor...could be fixed and it would still be some form of scrypt with chacha as mixing function.

in practice most jane coins are just forks of Yacoin, copying the formula for progressive N and only varying the start time and possible the nmin and nmax parameters.

In my opinion, all these coins progress too slowly to really fend off ASICs in the long tern...

Christian


You think so?

Only way I can imagine it being feasible is make an asic with more than enough ram for a few n factor changes and then add more ram for the next gen.
sr. member
Activity: 434
Merit: 250
https://bitcointalksearch.org/topic/annscryptadaptive-kmc-kimocoin-kimotocoin-5-exchanges-524216

Good CoiN!

LAUNCH TIME: 1st APRIL 2014

Algorithim : Scrypt-n-Adaptive, Scrypt-ASIC RESISTANT!

My 750ti is ready to mine this .
hero member
Activity: 756
Merit: 502
Scrypt-jane doesn't mean it has a variable N factor...could be fixed and it would still be some form of scrypt with chacha as mixing function.

in practice most jane coins are just forks of Yacoin, copying the formula for progressive N and only varying the start time and possible the nmin and nmax parameters.

In my opinion, all these coins progress too slowly to really fend off ASICs in the long tern...

Christian
hero member
Activity: 812
Merit: 1000
Bitleu, the Romanian coin is Asics proof . It is scrypt-Jane with variable N factor.
Scrypt-Jane with variable N factor, so...
Scrypt with variable N factor with variable N factor???

Scrypt-jane doesn't mean it has a variable N factor...could be fixed and it would still be some form of scrypt with chacha as mixing function.
hero member
Activity: 676
Merit: 500
Last chance to vote vor FC... It's currently in 2nd place only...

you think cryptoaltex is the right exchange for fc?
look at the daily trade amount... is that exchange at all?

same can be said about nxt-e Wink

A small step for a coin, but a giant leap for coinkind.



+1
hero member
Activity: 756
Merit: 502

FC/BTC market added to https://exarena.net

and soon also on newaltex.com (I voted with my feet, ermm wallet)

Christian
sr. member
Activity: 280
Merit: 250


And, you know what, I don't like those asic manufacturers/mining cartels/mafia dudes at all xD


This^

I really dont have a hatred of ASICs per se..  it's this bit that gets me.  If you could literally hop on newegg or amazon or wtf ever, order whatever the new kickass asic is
and have it show up a couple days later i really wouldnt care.  Its the "pre-order now so it'll make you half the money (if you're lucky) that you would have made when you ordered it!"

All the bastards do is announce the "new tech" and let you pre-order while theyre using it and developing the next best thing. When the next best thing is ready they start using that and pass off the leftovers as the latest and greatest at full price and you get to be behind the big boys yet again.
Jump to: