Topic: ASIC = The end of decentralized mining - page 8. (Read 22806 times)

I took this into account:
1.51 Gbps / 8 (bits/bytes) / 64 (bytes/block) / 2 (SHA-256 blocks per Bitcoin hash) / 0.0037 (Watt) = 398.5 Mhash/Joule

Edit: actually this is wrong. 1.51Gbps was achieved at maximal clock (200 MHz), for which power consumption was not reported. the 0.0037 Watt was estimated with a clock running at 50 Mhz, not 200 Mhz. The actual performance number that can be used is the quoted 19.75 mJ/Gbits which correspond to 49.4 Mhash/J (see my edited post above).

When they come, I'll be ready
I hear their voices inside
The stars in the heavens are moving
Soon they will align

Gavin, god of Bitcoin
Let me die with a FPGA in my hand
Raise your hands, GPUs in the wind
Brothers of Bitcoin together again
With blood in our voices we ride
We'll fight till we win or we'll fight until we die  

270days on BFL vs 440days on Cablepair guy's FPGA.

Difficulty increase and block reward halfing make it even worse, not to mention you get full trade in value with the BFL single

If you cool the Single right you get 900Mhash on the top firmware, with more firmware coming soon according to BFL, so maybe even 1Ghash. Wait time is now down to 50 days.

Have you run the numbers for breaking even on any investment, GPU or FPGA? Including electricity?

Saying: FPGA manufacturers are enjoying sales in the many tens of thousand
...is talking about unit quantity sales and not money, if you meant money then you need to learn to phrase your thoughts right.

so true ... he sounds like a horse and buggy salesman after automobiles were developed ... he won't answer this, but ask him how many months it would take for an 'investment' in his product to break even, including his assumptions for difficulty increases and reward reductions ... makes your comments even more entertaining

No, it has never previously been changed.

Question: has this ever happened before? I know with GPU mining, there have been code cleanups, updates for compatibility/optimization for driver/SDK changes, and forks for newer hardware, but has the bitcoin code ever changed the way the SHA256 hashing works?

Not that I have to prove anything to you or that tens of thousands of dollars is really that much money (what are you 13 or something?)

To date we have sold almost 50 ModMiner Quads, I encourage you to please do the math.

if you dont believe me just ask around there are many people on this forum who have purchased 5 or 10 units in one shot.
im ignoring you now.

Yes son,  many tens of thousands, we believe you 

You've got it all wrong cablepair.  The fact that we DON'T currently have ASIC miners available to anyone who wants to buy them makes that risk of a rich guy spending his trust fund on a custom ASIC chip that much greater.  When we DO have ASIC miners available to the public, that risk will be mitigated.  By your own anecdote, you've aptly demonstrated exactly the reason why publicly sold ASICs are exactly what Bitcoin mining needs to continue keeping the network secure.

If you don't believe the Bitcoin protocol is correct, then just outright state it.  Create your own fork where ASICs are impossible, either by changing the algo every 3 months, or by some other method of your choosing.  But Bitcoin is SHA256 - it is defined by that very algorithm.  To change that would be to admit fault in the code Bitcoin code and idea itself.  Lots of other people have found "faults" in Bitcoin and "fixed" it through their own fork.  So far, none of them have been successful to any sort of similar degree as Bitcoin.  But if you believe Bitcoin is faulty, then please do take your FPGA's and start hashing on your own blockchain with your own idea of what the algorithm should look like.  See how many people follow you there.  

Your point?  Who cares if ASIC's can't adapt?  No one is changing the Bitcoin code...

No one is disputing that FPGA's are #1 right now.  Pretty sure everyone agrees on that.  Your point?

To everyone else I respect your opinion a great deal, and to be honest I hope ASIC comes to the Bitcoin world in the spirit of helping the network and not hurting it, its when people like Vladimir come on giving warnings, thats when I get fired up.


Believe it or not I care first and foremost about the Bitcoin network and if you look at the things I have done for the Bitcoin world over the last year + you cannot die my love and passion for Bitcoin and the Bitcoin community.

I am passionate about fighting for Bitcoin and if I offended anyone I am sorry, ASIC has the potential to be very good for Bitcoin but the potential is greater for it to be very very bad for Bitcoin

I have outlined these things in other threads so I am not going to beat a dead horse here, I just hope everyone really does a lot of research and looks at the motives of the people involved before they make a decision on where they stand and just please before you completely embrace this thing without caution take a look at the potential for abuse.

for me its back to work

Have a nice day Gentleman its been nice debating with you.

Your such a clown you have no idea what your talking about

ASICS arent even out yet.

no one can dispute FPGA is #1 right now, the debate is what will happen when ASICS DO come out.

Myself and the other FPGA manufacturers are enjoying sales in the many tens of thousands, you really need to get a clue your just making yourself look foolish at this point.

Who said they could?

Neither.  Outside of Bitcoin there is absolutely nothing which needs this level of performance.  Even a mere 1 GH/s is a staggering amount of hashing power.  Say you are building a board which will do PBKDF2 in hardware.  At 10,000 iterations of SHA-256 a 1 GH/s chip could handle 100,000 simultaneous logins.

Know any webservers which need 100K simultaneous logins?  If you are hashing anything other than a incrementing nonce the bandwidth requirements also become equally insane.  At 1 GH/s (2 billion hashes per second) your I/O is 512 bit in and 256 bit out so that under a Tbps.  To put it into perspective it is nearly 6x as much as PCIe 3.0 with full 16 lanes.

There simply is no reason to design a chip which can clock any higher.  Why?  How many entire libraries of congress per second do you need to hash?

Designing the chip to run at lower performance levels means lower power & cooling requirements, simpler deployment, and higher longevity/stability.

ASIC's cant adapt! Once you make that chip, you cant change it man!

One change the Bitcoin code and all current ASICs are useless and cannot be fixed, end of story

FPGA's and GPU's could easily adapt to the change with a software upgrade

do your homework before you come at me jack

LMAO watching Cablepair guy fight and cling for the last few days of his business is so entertaining. HAHA!

Your FPGA days are over son !

I believe ASICs will cause an initial centralization, but later on a widening decentralization. My reasoning is that once volume gets high enough, the marginal cost of a chip will be < $20 USD (based on existing costs for 1 GHz ARM chips), and at that time probably < 1 BTC. If that ever happens, someone will eventually offer a mass-market, 8 chip personal miner for < $100 USD.


-bgc