[ANN] Spondoolies-Tech - carrier grade, data center ready mining rigs

[RichBC]

I am far from certain that there is such a thing as a GPU Coin. Yes that is the case up to a point but as soon as a coin makes it into the mainstream, worthy of investment, then I suspect there will always be an ASIC of some sort that could be designed & fabricated?


Rich

[AleScamHole]

They have also stated they will end the POW phase in a year or so as well. It will be interesting to see how they fork goes when they do.

What's the critera for Eth to jump algorithms? Is there an automated conversion of Eth1 to Eth2 in the protocol? Is there a "central authority" handling that jump and those conversions?

a few months ago they were saying PoW ends sometime in summer or fall, ten tey pushed it to winter/2017, and now it sounds like summer 2017?

that's what i was thinkin.  lots of people invested in GPU mining ETH lately, maybe their "pot committed" to the community to keep pow now?

WE ARE NOW "POW COMMITTED"
like Btc
only with Eth
lolz             

[ManeBjorn]

Yup they are leaving it vague for sure.
No real reason to even give a time frame. They should just say the POW will end when they have the next step of the software ready to handle it.

They have also stated they will end the POW phase in a year or so as well. It will be interesting to see how they fork goes when they do.

What's the critera for Eth to jump algorithms? Is there an automated conversion of Eth1 to Eth2 in the protocol? Is there a "central authority" handling that jump and those conversions?

a few months ago they were saying PoW ends sometime in summer or fall, ten tey pushed it to winter/2017, and now it sounds like summer 2017?

[AleScamHole]

Yup they are leaving it vague for sure.
No real reason to even give a time frame. They should just say the POW will end when they have the next step of the software ready to handle it.

They have also stated they will end the POW phase in a year or so as well. It will be interesting to see how they fork goes when they do.

What's the critera for Eth to jump algorithms? Is there an automated conversion of Eth1 to Eth2 in the protocol? Is there a "central authority" handling that jump and those conversions?

a few months ago they were saying PoW ends sometime in summer or fall, ten tey pushed it to winter/2017, and now it sounds like summer 2017?

but let's play that out, do you not think they would have a RIOT, the day they announced if at all this year even, that pow will end for ETH?
does it mean anything that the "big" companies have invested in ETH mining?
does that tell us maybe it WONT end? (for a while atleast)
genesis ,hashcoins, zeushash all building huge eth farms, they MUST anticipate atleast 12 more months of mining?
imo why would they invest now if less than 12 months remain? (and that was SUPPOSED to be a conservative number, casper was intended for 2016 if i remember right initially)
inside info i bet is that we have atleast 12 more months on eth mining, that's my guess

[Searing]

Yup they are leaving it vague for sure.
No real reason to even give a time frame. They should just say the POW will end when they have the next step of the software ready to handle it.

They have also stated they will end the POW phase in a year or so as well. It will be interesting to see how they fork goes when they do.

What's the critera for Eth to jump algorithms? Is there an automated conversion of Eth1 to Eth2 in the protocol? Is there a "central authority" handling that jump and those conversions?

a few months ago they were saying PoW ends sometime in summer or fall, ten tey pushed it to winter/2017, and now it sounds like summer 2017?

but let's play that out, do you not think they would have a RIOT, the day they announced if at all this year even, that pow will end for ETH?
does it mean anything that the "big" companies have invested in ETH mining?
does that tell us maybe it WONT end? (for a while atleast)
genesis ,hashcoins, zeushash all building huge eth farms, they MUST anticipate atleast 12 more months of mining?
imo why would they invest now if less than 12 months remain? (and that was SUPPOSED to be a conservative number, casper was intended for 2016 if i remember right initially)
inside info i bet is that we have atleast 12 more months on eth mining, that's my guess

yeah never did get this...if he'd have done Ethereum with sha-256 everybody with a usb stick would have jumped in..more cheerleaders...more hype etc etc

but again why not keep the pow going for as long as possible....more invested in the coin working (miners) more hype etc

even if say the difficulty gets too large for current GPU's you could say..hey......sorry ....use the new gpu stuff coming out and extend it again...for another round of hype /pump

but imho ethereum would have been even bigger using asic and sha-256 .....I woulda jumped on that bandwagon just to watch my 'baby' hash again for fun from day 1

(can't sell my knc jupiter 550gh...it made me 53 btc.......a forlorn hope it will ever run again) so hey better odds

then me hooking up with Natalie Portman ..but not by much.....)

then again I know a couple guys who 'streaked' back to the family farm and dug the GPU rigs out of the attic because they could not let go either ( They are happily mining either with a dozen or more cards from say week one ...were just looking

for an excuse to fire up the toys again..so it paid off big time for them

anyway in 20/20 hindsight with what Ethereum has done....this would have been the better play...hell think of the large farms that would have dumped out of BTC

no we are talking war then

(I have no Ethereum...missed boat..just wool gathering on what if's ...likely wrong)

[Unacceptable]

Yup they are leaving it vague for sure.
No real reason to even give a time frame. They should just say the POW will end when they have the next step of the software ready to handle it.

They have also stated they will end the POW phase in a year or so as well. It will be interesting to see how they fork goes when they do.

What's the critera for Eth to jump algorithms? Is there an automated conversion of Eth1 to Eth2 in the protocol? Is there a "central authority" handling that jump and those conversions?

a few months ago they were saying PoW ends sometime in summer or fall, ten tey pushed it to winter/2017, and now it sounds like summer 2017?

but let's play that out, do you not think they would have a RIOT, the day they announced if at all this year even, that pow will end for ETH?
does it mean anything that the "big" companies have invested in ETH mining?
does that tell us maybe it WONT end? (for a while atleast)
genesis ,hashcoins, zeushash all building huge eth farms, they MUST anticipate atleast 12 more months of mining?
imo why would they invest now if less than 12 months remain? (and that was SUPPOSED to be a conservative number, casper was intended for 2016 if i remember right initially)
inside info i bet is that we have atleast 12 more months on eth mining, that's my guess

yeah never did get this...if he'd have done Ethereum with sha-256 everybody with a usb stick would have jumped in..more cheerleaders...more hype etc etc

but again why not keep the pow going for as long as possible....more invested in the coin working (miners) more hype etc

even if say the difficulty gets too large for current GPU's you could say..hey......sorry ....use the new gpu stuff coming out and extend it again...for another round of hype /pump

but imho ethereum would have been even bigger using asic and sha-256 .....I woulda jumped on that bandwagon just to watch my 'baby' hash again for fun from day 1

(can't sell my knc jupiter 550gh...it made me 53 btc.......a forlorn hope it will ever run again) so hey better odds

then me hooking up with Natalie Portman ..but not by much.....)

then again I know a couple guys who 'streaked' back to the family farm and dug the GPU rigs out of the attic because they could not let go either ( They are happily mining either with a dozen or more cards from say week one ...were just looking

for an excuse to fire up the toys again..so it paid off big time for them

anyway in 20/20 hindsight with what Ethereum has done....this would have been the better play...hell think of the large farms that would have dumped out of BTC

no we are talking war then

(I have no Ethereum...missed boat..just wool gathering on what if's ...likely wrong)

You weren't here WAY back when,when BTC was GPU mined  It was awesome,EVERYONE had a chance to mine,big farms & little guys at home  

ASIC's are the bane of ANY coin....since "we" cannot make them at home or have normal access to them for instant purchase like GPU's,also we have no control how the corps use them(sell or self mine).

So,I'll keep looking for another coin to mine & use,to me BTC is dead.

Only use I have for it now is localbtc to sell my converted earnings,if there were a localnameofaltcoin to sell what I mine,I wouldn't need BTC at all  

Oh & at least one more "competitor" is out of the race... CYA SPTech!!!!!!   

[Biffa]

Surely all someone has to do is come up with a way to build a ASIC miner packaged multi gpu box. i.e. buy the GPU's and mount multiples of them on boards like many of the current multi-small-asic miners (antminer, avalon, etc) clock them low enough to be kept cool by heatsinks and large fan combos and you are back in the mining race.

Yes it depends on being able to get raw gpu cores from AMD or whoever, or develop their own programmable asic but given the right incentives (money) someone will come up with a method of concentrating the crunching power into "boxes".

Maybe.  Here is what I have been looking at from an economic viewpoint.

 AMD and Nvidia  are bigger by far then any asic company.
 They have a vested interest in finding ways to market GPU's.
 Eth coin is responsible for about 100,000 r9 390's in hash power  24/7/365
 When you breakdown that 100,000 number   we all know that some gamers will mine while they sleep and turn the mining off while they game.
 Also some mine with r9 380's  so Eth coin has driven far more sales then  100,000 r9's
 So both Nvidia and AMD   can prop up Eth coin as an advertising cost.
 Asrock builds a gpu mining board so  they can prop up Eth coin as an advertising cost.
 Intel sells cpus they can prop up eth coin as an advertising cost
 Amd sell cpus   they can prop up eth coin as an advertising cost.


All of the above companies have zero interest or gain from asic chips.

So from an economical viewpoint   GPU mining is going to stick around.

I see huge issues for Asic companies down the road. 

A simple this ' tape will self destruct ' method for Eth coin.  Ie turn it into a pos only coin once an asic builder makes a chip to mine it.

We already have  Eth coin 2 in place with a new algorithm  to fend off asics.  You can keep your Eth 1 coin as pos or convert it to Eth 2 coin as pow


  If I was an asic company  I would be shitting myself as this could mean death to asic mining.

Note bitmaintech now has come out with a s-7 lite version in my opinion due to eth coin explosion.

You sort of missed the point, my bad I didn't explain it well.

Sure with ASIC resistant coins such as ETH with their willingness to modify and prevent a fixed function processor being able to crunch the algo getting an box with a bunch of ASIC boards in it isn't going to be a success and ends up in the same situation BTC is now.

However, and remember a the term GPU mining is a bit of a misnomer, it really is Graphics card mining. Yes the GFX card does contain the GPU, just like a PC contains a CPU and a Bitcoin Miner contains ASIC's, but a GFX card also has a lot of stuff on it that isn't needed on a miner.

The point I was making was that its possible given the right set of circumstances and money for someone to purchase the GPU cores directly from the manufacturer (NV/AMD/IMG/INTEL/etc) in large numbers and instead of mounting them one or two on a GFX card board, designs a board and transport which houses 10/20/30 GPU cores on the board, just like a ASIC miner design does. Yes its not easy, yes you need memory and other stuff that the GPU cores need to access, but its not impossible if the market is there.

You won't get up to the TH level in a single box doing it this way, but you could get up to multi MH and maybe low GH in a single box. The technology may or may not be worth investing and building but it would be an interesting extension to just buying GFX cards and being limited by the PC form factor and motherboard designs.

[sidehack]

Given that, if you did it right, the box could be used for variety of parallelizable computationally intensive tasks, there'd be a (limited, but still extant) market outside cryptocurrency mining.

[Biodom]

The point I was making was that its possible given the right set of circumstances and money for someone to purchase the GPU cores directly from the manufacturer (NV/AMD/IMG/INTEL/etc) in large numbers and instead of mounting them one or two on a GFX card board, designs a board and transport which houses 10/20/30 GPU cores on the board, just like a ASIC miner design does. Yes its not easy, yes you need memory and other stuff that the GPU cores need to access, but its not impossible if the market is there.

You won't get up to the TH level in a single box doing it this way, but you could get up to multi MH and maybe low GH in a single box. The technology may or may not be worth investing and building but it would be an interesting extension to just buying GFX cards and being limited by the PC form factor and motherboard designs.

Considering that AMD/NVIDIA sells billion(s) worth of cards yearly and now with Polaris/Pascal will be pushing average PC toward VR capability at a reasonable price, I don't think that it would be a viable idea, but who knows. I might get Polaris just for kicks on one PC first and see how it does as far as mining is concerned.

BTW, we will have 75% coins already mined at halving, so there will be not much for the taking afterwards.

[RoadStress]

I am with @smooth on this one. How could you measure this?
Also, 1 mil GPUs could be just as good if not better if it meant 1mil nodes.
Nodes are disappearing on bitcoin side and growing somewhere else.

If we consider that a well funded and bad intended malicious party would want to take over the bitcoin network hashing power I think that there are at least 3 things that should be taken into consideration.
1) First and the most important is the the production cost of GH/s hashing power;
2) Second would be the requirements of infrastructure required for deploying enough hashpower to take over the network. I think we can include here things like how many total MW are required, what kind of cooling, GH/W ratio and so on;
3) The deployment speed of the total hashpower.

Maybe there are more things that can affect this, but the ASICS surely got these 3 points covered way more better than 1 mil GPUs. I can think of many downsides of having 1 mil GPUs which would make it way more easier for the malicious party to do nasty things to the network. This is very obvious for me and I can't wait to hear your motivation for not thinking so.

Also I totally agree with Biffa's above post

BTC is going to die due to Asic's and Chinese Asic building tactics combine with self-mining.

I am in my late 50's and have seen a lot of patterns of commerce unfold.

Don't compare bitcoin to anything! It's an all-in-one-one-of-a-kind thing and will not follow any legacy pattern! BTC

Isn't the SHA256 algo at their absolute limits with the 16nm chip technology? We might get a 0.1W/GHS miner but I don't think we will get 0.001W/GHS anytime soon.

Don't underestimate brilliant minds! Later generations are always better!

[s1gs3gv]

BTC is going to die due to Asic's and Chinese Asic building tactics combine with self-mining.

I am in my late 50's and have seen a lot of patterns of commerce unfold.

Don't underestimate brilliant minds! Later generations are always better!

I was born in the first half of the last century and realize i don't know very much about anything, but the older i get the more it seems that
"What has been will be again, what has been done will be done again; there is nothing new under the sun."

Maybe the new generation is an exception.

~LOL~

[Biodom]

I am with @smooth on this one. How could you measure this?
Also, 1 mil GPUs could be just as good if not better if it meant 1mil nodes.
Nodes are disappearing on bitcoin side and growing somewhere else.

If we consider that a well funded and bad intended malicious party would want to take over the bitcoin network hashing power I think that there are at least 3 things that should be taken into consideration.
1) First and the most important is the the production cost of GH/s hashing power;
2) Second would be the requirements of infrastructure required for deploying enough hashpower to take over the network. I think we can include here things like how many total MW are required, what kind of cooling, GH/W ratio and so on;
3) The deployment speed of the total hashpower.

Maybe there are more things that can affect this, but the ASICS surely got these 3 points covered way more better than 1 mil GPUs. I can think of many downsides of having 1 mil GPUs which would make it way more easier for the malicious party to do nasty things to the network. This is very obvious for me and I can't wait to hear your motivation for not thinking so.

Also I totally agree with Biffa's above post

as far as Biffa's remark is concerned-I don't see why would AMD/Nvidia sell cores to someone at a discount when they can sell a nice card with premium.
so, i disagree with that and with points 1-3 as well-these are not the most important. Why? Because a few lines of code can take care of this, like switching from ASIC to non-asic POW or from POW to POS. lightning network is basically POS.

Kind of tired of the notion that you need a loud machine running 24/7 and consuming 1.4 kw/h of electricity to process much less than ONE transaction a day.

[RoadStress]

I was born in the first half of the last century and realize i don't know very much about anything, but the older i get the more it seems that
"What has been will be again, what has been done will be done again; there is nothing new under the sun."

Maybe the new generation is an exception.

~LOL~

While you are true, I was talking about the chips generation.

[smooth]

I don't see why would AMD/Nvidia sell cores to someone at a discount when they can sell a nice card with premium.

Because GPUs are not a special purpose device i.e. magic money machine. AMD and Nvidia price their chips to maximize their return on investment across a wide range of applications, not based on how many BTC they can mine.

It conceivably could become the case that if Bitcoin were GPU mined and became much larger, mining would be the primary use for GPUs in which case they too would become in effect application specific integrated circuits. That's a long way off if ever.

[mjosephs]

That doesn't answer any of my questions. Who decides? Is it consensus, or is it central? Is the coin designed to be replaced, or is the transition seamless requiring no consumer intervention?

No, it's not seamless at all.  Ethereum has what they call the "difficulty bomb".

https://github.com/ethereum/go-ethereum/pull/1588

There's an extra factor added to the difficulty (regardless of actual network hashrate) and that factor rises exponentially.  It's estimated to overtake the hashrate-computed difficulty sometime around August.  By about November the network will basically grind to a halt unless somebody wants to devote infinite amounts of compute power to ethash.

So users are forced to pick some hard fork or abandon their wallets.

It remains to be seen how well this will work, or whether it will work at all.

To their credit, what they did was somewhat smarter than simply hardwiring a "halt at block X" since the most likely resolving hardfork would have been just deleting that line sometime before block X.  Deleting the difficulty bomb at this point would require actual coordination regarding exactly how it is wound down (i.e. at what rate, or if instantly at what block?).

Personally this malfeature, and the fact that most people buying ethereum are unaware of it, is the main reason I've avoided ETH both as an asset and a mining target.  It's one thing to accomplish a non-contentious hardfork -- bitcoin has done that too.  But pre-planning hardforks basically guarantees that the coin will get infected with nasty political influences.  If this scheme does work, I suspect that the resolving hardfork will itself have another "bomb" planted 16 months or so into the future -- the infection will prove to be lifelong and uncurable.

[tvasbn]

Guy, any inside infos about whos "behind" current hashrate bump?

[Guy Corem]

Guy, any inside infos about whos "behind" current hashrate bump?

 

I have a bet with someone it will be over 2 EHs on July 1st.

[alh]

Guy, any inside infos about whos "behind" current hashrate bump?

 

I have a bet with someone it will be over 2 EHs on July 1st.

Looks likely you will win that bet, possibly even by JUNE 1st.

So where is it coming from? Bitmain S9 or something else?

[klondike_bar]

Guy, any inside infos about whos "behind" current hashrate bump?

lady luck, also known as "statistical variance"

[klondike_bar]

Guy, any inside infos about whos "behind" current hashrate bump?

 

I have a bet with someone it will be over 2 EHs on July 1st.

thats pretty ballsy - with recent TMSC delays of 1-2 months it looks like chips may just be getting into the right hands now, or during the next few weeks. How quickly that becomes fabricated hardware and plugged in is the big question.

personally i think August 1st is a more realistic 2EH target (assuming you take a larger moving average, and not day-to-day spike)