Pages:
Author

Topic: who bought bitmain Z9 will face a big issues now (Read 4134 times)

sr. member
Activity: 857
Merit: 262
what if an algo or new stuff (whatever it may be called) is created that it hashes by rendering-> the kind of work where GPUs excel over ASICs and FPGAs...

Rendering is really not that much different from hashing different algos.

Even if there was a similiar algo like that, it would still be possible to create an ASIC.

After Sha256 got ASICs, Scrypt was created by the Litecoin dev as a way to keep it GPU mineable only.

I think it was more intensive than Sha256 but a few years later ASICs came out for that too.

Same with Dashcoin which was based on the X11 algo which was also suppose to be ASIC resistant.

So Equihash and DaggerHash are no different.

I mean like it has to be processed with a "certain" graphics work (like a secret sauce) involved in the computation..not exactly rendering..

doing stuff where GPU's excel...

a) can be merged - mix mix

b) can be separate - a security measure, like "do this stuff first" before hashing....that "stuff" is like something that may or may not be relevant to hashing but only serves as a security like a lock, where the graphics card have to produce a "key" ->the key is something that a GPU can produce but FPGAs and ASICs cannot...or something only GPUs can produce easily but others cannot.

.......what do GPUs do when playing games? maybe that's where we can find the secret sauce..

Please keep in mind that all of the following are ASICs(with different capabilities): Intel, AMD and ARM CPUs, Nvidia and AMD GPUs, FPGAs, DSPs... anything actually.
And there's nothing special about GPU - it just allows for some (massive comparing to CPU) parallel calculations(SIMD/MIMD) has some fast RAM nearby and is more flexible than dedicated mining ASICs. However ASIC can be built as flexible as you desire providing there's a justified need.

The current modus operandi of ASIC makers is not "highly efficient and sustainable", which is generally understandable as it's hard to operate planing long term while the silicon process keep shrinking, and blockchain is exploding at the same time.
Hopefully they will do better now when more competition arrives... First we'll need to see if it will actually make the mining hw generally more available.
Interesting thing though: it probably will only seemingly reduce "exclusiveness" of mining and improve decentralization ... as it will reduce mining profits and thus force only the most optimized operations to be profitable...
newbie
Activity: 48
Merit: 0
Seems like the x16r variable algorithm chain should make it pretty difficult to make an asic for it.

Still possible but unlikely.

Makes raven seem like a decent project. ZEC is about the same as far as profitability goes right now. Seems the manipulators pulled all of their sell walls on Kraken for ZEC. It will be interesting to se how far it climbs before they crash it again.

Seems like it will be a good time to sell ZEC if you have it just before it lists on Gemini
Bro kraken isnt even on top 15 places with how many zec are being trade on some exchange,that wall you saw doesnt make any difference.If you want to do ta do it on exchange with most volume
https://coinmarketcap.com/currencies/zcash/#markets

I imagine that they look similar across most exchanges. The people moving big enough money to manipulate are on all of them simultaneously. The only way it works.

I don’t have time to be on all of them but they only deviate so much from one another to collectively determine market price at a given point in time.

The consensus seems to be anticipation of Gemini. ZEC has been hammered over the past couple of months after being relatively consistent with the change in percent value from one day to the next relative to the other major coins.

I still have some hashing power on ZEC. Once this z9 comes in I will have more. I will run it until I cannot anymore. Hopefully it does fork though. Has been a great GPU coin.
full member
Activity: 1148
Merit: 132
Seems like the x16r variable algorithm chain should make it pretty difficult to make an asic for it.

Still possible but unlikely.

Makes raven seem like a decent project. ZEC is about the same as far as profitability goes right now. Seems the manipulators pulled all of their sell walls on Kraken for ZEC. It will be interesting to se how far it climbs before they crash it again.

Seems like it will be a good time to sell ZEC if you have it just before it lists on Gemini
Bro kraken isnt even on top 15 places with how many zec are being trade on some exchange,that wall you saw doesnt make any difference.If you want to do ta do it on exchange with most volume
https://coinmarketcap.com/currencies/zcash/#markets
zec is only pumping bc its added to gemini
full member
Activity: 294
Merit: 103
Seems like the x16r variable algorithm chain should make it pretty difficult to make an asic for it.

Still possible but unlikely.

Makes raven seem like a decent project. ZEC is about the same as far as profitability goes right now. Seems the manipulators pulled all of their sell walls on Kraken for ZEC. It will be interesting to se how far it climbs before they crash it again.

Seems like it will be a good time to sell ZEC if you have it just before it lists on Gemini
Bro kraken isnt even on top 15 places with how many zec are being trade on some exchange,that wall you saw doesnt make any difference.If you want to do ta do it on exchange with most volume
https://coinmarketcap.com/currencies/zcash/#markets
newbie
Activity: 48
Merit: 0
Seems like the x16r variable algorithm chain should make it pretty difficult to make an asic for it.

Still possible but unlikely.

Makes raven seem like a decent project. ZEC is about the same as far as profitability goes right now. Seems the manipulators pulled all of their sell walls on Kraken for ZEC. It will be interesting to se how far it climbs before they crash it again.

Seems like it will be a good time to sell ZEC if you have it just before it lists on Gemini
full member
Activity: 216
Merit: 100
what if an algo or new stuff (whatever it may be called) is created that it hashes by rendering-> the kind of work where GPUs excel over ASICs and FPGAs...

Rendering is really not that much different from hashing different algos.

Even if there was a similiar algo like that, it would still be possible to create an ASIC.

After Sha256 got ASICs, Scrypt was created by the Litecoin dev as a way to keep it GPU mineable only.

I think it was more intensive than Sha256 but a few years later ASICs came out for that too.

Same with Dashcoin which was based on the X11 algo which was also suppose to be ASIC resistant.

So Equihash and DaggerHash are no different.

The only difference is that the DaggerHash is more GPU resistant.
member
Activity: 378
Merit: 11
Decentralized Digital Billboards
Quote
GPU is decentralized because pretty much anyone can get into the game with 1 GPU and scale up over time.  With ASIC's....you literally have all your coins in 1 basket and if that coin is forked, well you are forked too.
Decentralizing is when every miner download full blockchain and store it on their machine.
How we know more then 99% miner not doing that, more then 90 mine to the exchange directly
legendary
Activity: 3444
Merit: 1061
what if an algo or new stuff (whatever it may be called) is created that it hashes by rendering-> the kind of work where GPUs excel over ASICs and FPGAs...

Rendering is really not that much different from hashing different algos.

Even if there was a similiar algo like that, it would still be possible to create an ASIC.

After Sha256 got ASICs, Scrypt was created by the Litecoin dev as a way to keep it GPU mineable only.

I think it was more intensive than Sha256 but a few years later ASICs came out for that too.

Same with Dashcoin which was based on the X11 algo which was also suppose to be ASIC resistant.

So Equihash and DaggerHash are no different.

I mean like it has to be processed with a "certain" graphics work (like a secret sauce) involved in the computation..not exactly rendering..

doing stuff where GPU's excel...

a) can be merged - mix mix

b) can be separate - a security measure, like "do this stuff first" before hashing....that "stuff" is like something that may or may not be relevant to hashing but only serves as a security like a lock, where the graphics card have to produce a "key" ->the key is something that a GPU can produce but FPGAs and ASICs cannot...or something only GPUs can produce easily but others cannot.

.......what do GPUs do when playing games? maybe that's where we can find the secret sauce..
legendary
Activity: 3808
Merit: 1723
what if an algo or new stuff (whatever it may be called) is created that it hashes by rendering-> the kind of work where GPUs excel over ASICs and FPGAs...

Rendering is really not that much different from hashing different algos.

Even if there was a similiar algo like that, it would still be possible to create an ASIC.

After Sha256 got ASICs, Scrypt was created by the Litecoin dev as a way to keep it GPU mineable only.

I think it was more intensive than Sha256 but a few years later ASICs came out for that too.

Same with Dashcoin which was based on the X11 algo which was also suppose to be ASIC resistant.

So Equihash and DaggerHash are no different.
legendary
Activity: 3444
Merit: 1061
what if an algo or new stuff (whatever it may be called) is created that it hashes by rendering-> the kind of work where GPUs excel over ASICs and FPGAs...
jr. member
Activity: 147
Merit: 1
newbie
Activity: 45
Merit: 0
GPU miners are like the banks and corps that need govt bailout because their business model is trash

whine whine welfare, protect only my marxist interests and block others from joining the markets..
 Tongue

Lol ... You have no concept of what decentralization means.

decentralizing isn't people hoarding hundreds of video cards - -this can be the solution to the centralization of hoarding GPUs

You have a weird idea of a solution to hoarding GPU's....by making ASIC's made by 1 - 3 companies, who also own the largest mining pools, of which those who can afford the high entry point $1500-$2000 minimum for one machine.  That leaves a small number of individuals/corporations.

I have 2 ASIC's and 2 GPU rigs and the ROI on each is night and day.   You need volume with ASIC's to be profitable.  ASIC's are like the amazon and walmarts.   It is about quantity.  I'm not saying you can't be profitable with ASIC's and those who by 1st generations are generally those who ROI on the ASIC's.

GPU is decentralized because pretty much anyone can get into the game with 1 GPU and scale up over time.  With ASIC's....you literally have all your coins in 1 basket and if that coin is forked, well you are forked too.

newbie
Activity: 83
Merit: 0
GPU miners are like the banks and corps that need govt bailout because their business model is trash

whine whine welfare, protect only my marxist interests and block others from joining the markets..
 Tongue

Lol ... You have no concept of what decentralization means.

decentralizing isn't people hoarding hundreds of video cards - -this can be the solution to the centralization of hoarding GPUs
member
Activity: 644
Merit: 24
GPU miners are like the banks and corps that need govt bailout because their business model is trash

whine whine welfare, protect only my marxist interests and block others from joining the markets..
 Tongue

Lol ... You have no concept of what decentralization means.
legendary
Activity: 1848
Merit: 1166
My AR-15 ID's itself as a toaster. Want breakfast?
they would just write the software so one 10,000 sol asic simulates 14 miners at 700 sols and connects to 14 pools. kinda like a 14 threaded cpu.

or is this limited sols per ip or something.. unless you get dozens of internet addys/lines that would make farms useless and gpu sales would tank anyway as each person will only buy one gpu per algo. a few 8 or whatever gpu rigs to cover a couple dozen algos and youre done.

maybe im interpreting this wrong.

since mining was designed around being able to do things as fast as possible..... no way to really limit mining speed I can think of except for a time check/counter being mixed in with the mining code as to change the value of the result obtained from said hash based on the time to calculate;  plus or minus a percentage.  Before you hash, you salt the input with a timestamp.  Then, you salt it with a ratio of the outbound result with another timestamp.    Apply the "salt" as you would a TLS handshake....  and verify the validity of a result the same way as any other; via the stratum and network nodes verifying and sharing the same clock specs and +/- error.

If the network detects too high of a result coming from a certain node;  that node will get penalized on the network somehow.... TBD....

There are methods though... but flat out putting a "speedlimit" on people will only convince them to find ways to break up their hash and have it seemingly spread across multiple sources....  IF you make anti-pool code;  people will just set up a million smaller miners.... etc.  Gotta think past all that.
legendary
Activity: 4354
Merit: 3614
what is this "brake pedal" you speak of?
I have a viable ideas that will stop any big miner.

Ie my new coin is nvidia gpu coin   you can only send 1 miner capped at say  650sols     or like 1x  1080ti

So the benefit of a 10000 sol asic  is worthless as it needs to clock to  650 sols.

the asic = worthless
the fpga = worthless.

in fact  a new coin  such as gpuZcash  can use  the asic al-gore-rythm of equihash  but on the blockchain  limit the block winners to 700 sols. of hash

so you 10000 sol asic has to downclock to 700 sols  not a brick  but fucked none the less.

btw at zcash founders  you could do this with zcash.  and it would work.  pools would need more servers  since more workers  would be used.

program your pos asics of fpga's then  just down clock them to 7 percent 

so the 10000 sols drops to 700 sols and it saves some watts   maybe you do 50- 100 watts, but it does not look that good anymore.


so asic resistance is easy.


they would just write the software so one 10,000 sol asic simulates 14 miners at 700 sols and connects to 14 pools. or 14 different workers at the same pool.

kinda like a 14 core cpu. each string of asic chips works as a different worker on its own.

or is this limited sols per ip or something.. unless you get dozens of internet addys/lines that would make farms useless and gpu sales would tank anyway as each person will only buy one gpu per algo. a few 8 or whatever gpu rigs to cover a couple dozen algos and youre done.

maybe im interpreting this wrong.
newbie
Activity: 83
Merit: 0
GPU miners are like the banks and corps that need govt bailout because their business model is trash

whine whine welfare, protect only my marxist interests and block others from joining the markets..
 Tongue
full member
Activity: 312
Merit: 104
Don't worry guys, you wont be using your GPU's for mining much longer anyway.

I'm going all in as a supplier in the sonm.io network. Livenet launches in june, but only with crypto mining IaaS to begin with.
IaaS crypto mining is basically automated mining with profit switching, aka a contract that will be generated to mine the most profitable coin when your rig is idle/has no other contracts/deals.

Later on customers will be able to rent your GPU/CPU/HDD or internet bandwidth as well. Your rig(s) will always sell to the highest bidder, which means profit will always match or be better than regular GPU mining.



Honestly, i can`t wait for this to become a reality.
full member
Activity: 462
Merit: 118
many frequent forks stop all asics dead.
hash growth limits stop super farms of gpus.

I figure 400,000,000 gpus are out there right now that can mine a coin at a profit.


if ½ are big player and ¼ are small players ¼ or 100,000,000 are tiny one or 2 card machines.  This is a board broad adoption base to be nurtured not killed off.  

If you don't believe this you are the type of person that will always be dissatisfied with life and generally always be somewhat annoyed and unhappy.

So anyone reading above and feeling pissed off  about that sentence I just wrote in bold  I am talking to you. Grin
I know that sentence will anger quite a few people that want an asic takeover of all coins.

My gut tells me asic manufacturers will come up with a cheap FPGA instead if this persist. GPU miners need to do some soundproofing and they can run asics actually.


Just my 2 cents.

mining asics are not the issue...  have plenty of asics but none that will mine and kill off gpu algos

There is no way there will be 100,000,000  or more individual asic miners.
Since the chip can only mine.

You simply  refused to talk that you want to kill off 100,000,000 possible gpu miners and replace  100,000 or less asic miners.

You refuse to talk about 2017  the best year in cryptocoins history being driven by gpus,

And you don.t mention the loss of quality gpu warranty service being replaced by shit asic warranties

I disagree. All they need is one programmable asic aka FPGA and gpu is dead. No more forking will occur because the FPGA will just program itself. When no more forks occur, new asics can be created to target individual algo. The FPGA is insurance in case algos fork. Another way is drop E3 price back to 800 USD. Gpu dies then lol.


I am not refusing anything. GPU mining may have surprisingly better than asics for 2017. But owning the coin wins all actually because difficulty climbs too much. U make more by owning the underlying gpu coin as well, in general.

The ball to destroy GPU is in asic manufacturers court. They want to or they dont want to only.

Keep that thought.  BTW  I am now down to 16x 1080ti  and asic for LTC and BTC

It is not that asics can win or lose if mining becomes all asic  or asic + asic /fpga  this industry will eat its own self alive.

It needs the broad base of gpus in smaller setups.  As you still refuse to see that 100,000,000 small guys getting a piece of the pie is why 2017  was a record year.

Asics only or fpga/asics  will kill it.

And I can guarantee amd or nvidia backers will create a new way to earn  that shuts out asics and fpga miners.

They simply don't want to lose the sales.

I have a viable ideas that will stop any big miner.

Ie my new coin is nvidia gpu coin   you can only send 1 miner capped at say  650sols     or like 1x  1080ti

So the benefit of a 10000 sol asic  is worthless as it needs to clock to  650 sols.

the asic = worthless
the fpga = worthless.

in fact  a new coin  such as gpuZcash  can use  the asic al-gore-rythm of equihash  but on the blockchain  limit the block winners to 700 sols. of hash

so you 10000 sol asic has to downclock to 700 sols  not a brick  but fucked none the less.

btw at zcash founders  you could do this with zcash.  and it would work.  pools would need more servers  since more workers  would be used.

program your pos asics of fpga's then  just down clock them to 7 percent  

so the 10000 sols drops to 700 sols and it saves some watts   maybe you do 50- 100 watts, but it does not look that good anymore.


so asic resistance is easy.

Lets try again.  there are two fully mature asic al-gore-rhythms   scrypt and sha 256

some al-gore-rhythms  are taking on asics and resisting
some figure we should go asic.

I am fine with  this.


I am not fine with gpus  being fully knocked out of the game.

Here is why:

1 every year gpus sucked to mine since 2014   crypto-coins had bad times.  but when gpus were good to mine crypto-coins has good times.
2 only avalon has a long asic warranty
3 all gpus have at least a 1 year warranty.



i dont think asics only is wrong. Bitcoin, bch, litecoin, dash is asics only and it does fine. It didnt affect their coin negatively.

The main threat is the bitmain E3 can mine Eth and it can cost as little as 800 USD. If bitmain sold it at 800 USD. it is over gpu. Because eth is 60+ % of gpu mining. GPU being forced out of eth into other coins will kill gpu mining due to super high difficulty.

asic resistance and forking is always a gamble
SIA coin didnt fork. Etn has issues forking,etc.
Anyways, I suspect bitmain is too smart and will get fpga or something if this keeps up.
As for Nvidia, they are more interested in AI and datacenters, and self driving tech,etc.
They had so much time and cant come up with a better gpu miner. Bitmain and others have done so.

The price is not $800 anymore. Last I checked it was like $2000.

Keep in mind this. If they were willing to sell at $800, imagine how cheap it was for them to produce them in the first place. Cost to manufacture was maybe $200-$400 or so.

Since it's so cheap and uses little power. They can have millions of them mining in some private farm somewhere.

Exactly. They can sell it at 800 usd if they wanted to and there is profit built in it. They just decided not to. Look at how cheap L3+ once was, 450 USD. Last year it was as high as 2k usd. And usually around 1.3k-1.6k usd. They can possible sell E3 for 500-600 USD if they wanted to and skyrocket eth difficulty. But they are going for high margin profits on the E3 at the moment it seems.
legendary
Activity: 3808
Merit: 1723
many frequent forks stop all asics dead.
hash growth limits stop super farms of gpus.

I figure 400,000,000 gpus are out there right now that can mine a coin at a profit.


if ½ are big player and ¼ are small players ¼ or 100,000,000 are tiny one or 2 card machines.  This is a board broad adoption base to be nurtured not killed off.  

If you don't believe this you are the type of person that will always be dissatisfied with life and generally always be somewhat annoyed and unhappy.

So anyone reading above and feeling pissed off  about that sentence I just wrote in bold  I am talking to you. Grin
I know that sentence will anger quite a few people that want an asic takeover of all coins.

My gut tells me asic manufacturers will come up with a cheap FPGA instead if this persist. GPU miners need to do some soundproofing and they can run asics actually.


Just my 2 cents.

mining asics are not the issue...  have plenty of asics but none that will mine and kill off gpu algos

There is no way there will be 100,000,000  or more individual asic miners.
Since the chip can only mine.

You simply  refused to talk that you want to kill off 100,000,000 possible gpu miners and replace  100,000 or less asic miners.

You refuse to talk about 2017  the best year in cryptocoins history being driven by gpus,

And you don.t mention the loss of quality gpu warranty service being replaced by shit asic warranties

I disagree. All they need is one programmable asic aka FPGA and gpu is dead. No more forking will occur because the FPGA will just program itself. When no more forks occur, new asics can be created to target individual algo. The FPGA is insurance in case algos fork. Another way is drop E3 price back to 800 USD. Gpu dies then lol.


I am not refusing anything. GPU mining may have surprisingly better than asics for 2017. But owning the coin wins all actually because difficulty climbs too much. U make more by owning the underlying gpu coin as well, in general.

The ball to destroy GPU is in asic manufacturers court. They want to or they dont want to only.

Keep that thought.  BTW  I am now down to 16x 1080ti  and asic for LTC and BTC

It is not that asics can win or lose if mining becomes all asic  or asic + asic /fpga  this industry will eat its own self alive.

It needs the broad base of gpus in smaller setups.  As you still refuse to see that 100,000,000 small guys getting a piece of the pie is why 2017  was a record year.

Asics only or fpga/asics  will kill it.

And I can guarantee amd or nvidia backers will create a new way to earn  that shuts out asics and fpga miners.

They simply don't want to lose the sales.

I have a viable ideas that will stop any big miner.

Ie my new coin is nvidia gpu coin   you can only send 1 miner capped at say  650sols     or like 1x  1080ti

So the benefit of a 10000 sol asic  is worthless as it needs to clock to  650 sols.

the asic = worthless
the fpga = worthless.

in fact  a new coin  such as gpuZcash  can use  the asic al-gore-rythm of equihash  but on the blockchain  limit the block winners to 700 sols. of hash

so you 10000 sol asic has to downclock to 700 sols  not a brick  but fucked none the less.

btw at zcash founders  you could do this with zcash.  and it would work.  pools would need more servers  since more workers  would be used.

program your pos asics of fpga's then  just down clock them to 7 percent  

so the 10000 sols drops to 700 sols and it saves some watts   maybe you do 50- 100 watts, but it does not look that good anymore.


so asic resistance is easy.

Lets try again.  there are two fully mature asic al-gore-rhythms   scrypt and sha 256

some al-gore-rhythms  are taking on asics and resisting
some figure we should go asic.

I am fine with  this.


I am not fine with gpus  being fully knocked out of the game.

Here is why:

1 every year gpus sucked to mine since 2014   crypto-coins had bad times.  but when gpus were good to mine crypto-coins has good times.
2 only avalon has a long asic warranty
3 all gpus have at least a 1 year warranty.



i dont think asics only is wrong. Bitcoin, bch, litecoin, dash is asics only and it does fine. It didnt affect their coin negatively.

The main threat is the bitmain E3 can mine Eth and it can cost as little as 800 USD. If bitmain sold it at 800 USD. it is over gpu. Because eth is 60+ % of gpu mining. GPU being forced out of eth into other coins will kill gpu mining due to super high difficulty.

asic resistance and forking is always a gamble
SIA coin didnt fork. Etn has issues forking,etc.
Anyways, I suspect bitmain is too smart and will get fpga or something if this keeps up.
As for Nvidia, they are more interested in AI and datacenters, and self driving tech,etc.
They had so much time and cant come up with a better gpu miner. Bitmain and others have done so.

The price is not $800 anymore. Last I checked it was like $2000.

Keep in mind this. If they were willing to sell at $800, imagine how cheap it was for them to produce them in the first place. Cost to manufacture was maybe $200-$400 or so.

Since it's so cheap and uses little power. They can have millions of them mining in some private farm somewhere.
Pages:
Jump to: