[ANN] Sfards:SF100, the first 28nm Dual-Mode Miner is accepting pre-order now

[yslyung]

Hmm. My can do 0.51 Wt/Gh (freq. 400, 1.3 Th). At acceptable noise level.

May be it is not so "dead quiet" like Avalon 4.1, but S5 a) cheaper, b) easier in deploying.

Also, S5 is better in price/Gh if I am not mistaken. But you said it already. 

    The avalon 4.1 becomes really easy to setup.
  After the first one is setup on a rasp pi 12 more will just plug n play. Running very stable.
Plus you can stack them 2 or  3 high as the pcie's are on the back end.

If your s-5 is doing .51 at freq 400 1.3th  you did a fan mod.
BTW they can be fan modded and noise level is okay at 1.3th  somewhere I have a thread on it.

But price wise 800gh for 2.3 btc =   66 cents a gh
vs 1150gh  for               1.66 btc = 33 cents a gh



Which is what I said.  Believe it or not   all this is not off topic.

those 2 numbers   66 cents a gh vs 33 cents a gh  are my pricing guide for this new gear.

Too close to 66 cents and it is over priced.  Too close to 33 cents and price war with bitmaintech.

Should be very interesting to see what they do.

WAR is sometimes VERY good

[Bicknellski]

Only for the best equipped army.

Hmm. My can do 0.51 Wt/Gh (freq. 400, 1.3 Th). At acceptable noise level.

May be it is not so "dead quiet" like Avalon 4.1, but S5 a) cheaper, b) easier in deploying.

Also, S5 is better in price/Gh if I am not mistaken. But you said it already. 

    The avalon 4.1 becomes really easy to setup.
  After the first one is setup on a rasp pi 12 more will just plug n play. Running very stable.
Plus you can stack them 2 or  3 high as the pcie's are on the back end.

If your s-5 is doing .51 at freq 400 1.3th  you did a fan mod.
BTW they can be fan modded and noise level is okay at 1.3th  somewhere I have a thread on it.

But price wise 800gh for 2.3 btc =   66 cents a gh
vs 1150gh  for               1.66 btc = 33 cents a gh



Which is what I said.  Believe it or not   all this is not off topic.

those 2 numbers   66 cents a gh vs 33 cents a gh  are my pricing guide for this new gear.

Too close to 66 cents and it is over priced.  Too close to 33 cents and price war with bitmaintech.

Should be very interesting to see what they do.

WAR is sometimes VERY good

[philipma1957]

yeah war is for the winners.  I am pretty sure bitmaintech would win against  Sfards in a price war.

Only for the best equipped army.

Hmm. My can do 0.51 Wt/Gh (freq. 400, 1.3 Th). At acceptable noise level.

May be it is not so "dead quiet" like Avalon 4.1, but S5 a) cheaper, b) easier in deploying.

Also, S5 is better in price/Gh if I am not mistaken. But you said it already. 

    The avalon 4.1 becomes really easy to setup.
  After the first one is setup on a rasp pi 12 more will just plug n play. Running very stable.
Plus you can stack them 2 or  3 high as the pcie's are on the back end.

If your s-5 is doing .51 at freq 400 1.3th  you did a fan mod.
BTW they can be fan modded and noise level is okay at 1.3th  somewhere I have a thread on it.

But price wise 800gh for 2.3 btc =   66 cents a gh
vs 1150gh  for               1.66 btc = 33 cents a gh



Which is what I said.  Believe it or not   all this is not off topic.

those 2 numbers   66 cents a gh vs 33 cents a gh  are my pricing guide for this new gear.

Too close to 66 cents and it is over priced.  Too close to 33 cents and price war with bitmaintech.

Should be very interesting to see what they do.

WAR is sometimes VERY good

[Meech]

yeah war is for the winners.  I am pretty sure bitmaintech would win against  Sfards in a price war.

Only for the best equipped army.

Hmm. My can do 0.51 Wt/Gh (freq. 400, 1.3 Th). At acceptable noise level.

May be it is not so "dead quiet" like Avalon 4.1, but S5 a) cheaper, b) easier in deploying.

Also, S5 is better in price/Gh if I am not mistaken. But you said it already. 

    The avalon 4.1 becomes really easy to setup.
  After the first one is setup on a rasp pi 12 more will just plug n play. Running very stable.
Plus you can stack them 2 or  3 high as the pcie's are on the back end.

If your s-5 is doing .51 at freq 400 1.3th  you did a fan mod.
BTW they can be fan modded and noise level is okay at 1.3th  somewhere I have a thread on it.

But price wise 800gh for 2.3 btc =   66 cents a gh
vs 1150gh  for               1.66 btc = 33 cents a gh



Which is what I said.  Believe it or not   all this is not off topic.

those 2 numbers   66 cents a gh vs 33 cents a gh  are my pricing guide for this new gear.

Too close to 66 cents and it is over priced.  Too close to 33 cents and price war with bitmaintech.

Should be very interesting to see what they do.

WAR is sometimes VERY good

Do we really want a winner for the mining companies?  Look how the price battle went with SP and Bitmain, we definitely benefited.  Once SP was out of the picture Bitmain got greedy.  Fierce competition all the time is what we miners need, not a winner.

[alh]

We clearly don't need a winner. I wonder though if the price of Bitcoin won't keep ALL the ASIC vendors in check. By that I mean a  low BTC/USD price keeps them from trying to kill each other. A high BTC price may encourage a "winner take all" strategy, which may not benefit any miner long term (past a couple of difficulty adjustments).

A little competition is probably good, but not a WAR

[Bicknellski]

We clearly don't need a winner. I wonder though if the price of Bitcoin won't keep ALL the ASIC vendors in check. By that I mean a  low BTC/USD price keeps them from trying to kill each other. A high BTC price may encourage a "winner take all" strategy, which may not benefit any miner long term (past a couple of difficulty adjustments).

A little competition is probably good, but not a WAR

Don't kid yourself. The price war is not the issue it is the efficient and hashrate volume that is the battle ground.

The price of these chips and units for the small and ever shrinking home market is not what will drive the end of these companies. Guy at Spondoolies already laid it out numerous times on what will kill these wannabes competing with the those with a game plan, real engineering teams and the money.

The battle for $ per gh/s is silly and meaningless given the volume of hashpower that might come online before 2016. The nuclear weapons are ready to be unleashed. This will leave only 2 or 3 main superpowers.

---

As for wanting big centralized systems based on 2 or 3 companies. NO FUCKING WAY!

Here is how we stop it.

1. Open source ASIC chip.
2. Open source low cost miner design.
3. Low cost miners in every hand.
4. P2Pool with those miners.

If you can ship more Terahash Petahash in a small, quiet, unobtrusive low cost device to millions that is the only way to decentralize things and get those miners on a P2Pool. Counting on Avalon or Bitmain or SFards as your chip / miner supplier saviour is like asking the fox to watch the hen house.

Insanity. No one learned a lesson from KnC or Bitfury did they? If you have a chip and it makes economic sense to mine massively for your benefit as company you will do it.

Be realistic this won't happen so the best we can hope for is a few more months maybe a year of projects like Gekkoscience or J4bberwock's effort.

[vs3]

Here is how we stop it.

1. Open source ASIC chip.
2. Open source low cost miner design.
3. Low cost miners in every hand.
4. P2Pool with those miners.

well - for example all you have to do is buy Bitfury's chip design and open-source it.
I have already done #2, and plenty of people did #3. #4 will happen by itself.

[Bicknellski]

Here is how we stop it.

1. Open source ASIC chip.
2. Open source low cost miner design.
3. Low cost miners in every hand.
4. P2Pool with those miners.

well - for example all you have to do is buy Bitfury's chip design and open-source it.
I have already done #2, and plenty of people did #3. #4 will happen by itself.

Yup.

And what happened?
You broke even maybe?
How many Bitfury chips can you buy today?

Millions bought them and now we had decentralized mining right? 100's did that is great but that is not millions, not enough to do much more than keep some hobbyists happy.

There are things you can change. There are things you can't change. I don't see how you break a monopoly that is coming at all.

Expecting these companies to be like you vs3 in anyway is not the hope you should be waiting for. I don't see anyone going to put the time and the effort in to make an ASIC and then get miners in the hands of millions which is the only way you beat the big players. It is over unless that happens somehow. I submit that the end is now and the monopoly on mining is already here and won't be challenged.

[philipma1957]

Here is how we stop it.

1. Open source ASIC chip.
2. Open source low cost miner design.
3. Low cost miners in every hand.
4. P2Pool with those miners.

well - for example all you have to do is buy Bitfury's chip design and open-source it.
I have already done #2, and plenty of people did #3. #4 will happen by itself.

Yup.

And what happened?
You broke even maybe?
How many Bitfury chips can you buy today?

Millions bought them and now we had decentralized mining right? 100's did that is great but that is not millions, not enough to do much more than keep some hobbyists happy.

There are things you can change. There are things you can't change. I don't see how you break a monopoly that is coming at all.

Expecting these companies to be like you vs3 in anyway is not the hope you should be waiting for. I don't see anyone going to put the time and the effort in to make an ASIC and then get miners in the hands of millions which is the only way you beat the big players. It is over unless that happens somehow.

 okay bick I play this game. bitfury can run at 0.07 watts per gh and build cheaply using cheap power.
knc can run at 0.08 watts  per gh and build it cheaply using cheap power.

lets argue why expand hash risking downward btc price pressure.  how about swap  new gear for old gear and cut power cost.
thus the two of them keep a good share of the market hash wise but simply lower expenses with more efficient gear.

  I can simply show no big self building miner should ever expand hash power
 they should keep changing the older power hungry gear for new power sipping gear.
If knc bitfury and sp-tech  do ever expanding hash rate growth they only hurt each other. 
 If they self mine with more efficient gear the hurt does not happen.

But only time will tell.

[toptekk]

Here is how we stop it.

1. Open source ASIC chip.
2. Open source low cost miner design.
3. Low cost miners in every hand.
4. P2Pool with those miners.

well - for example all you have to do is buy Bitfury's chip design and open-source it.
I have already done #2, and plenty of people did #3. #4 will happen by itself.

Yup.

And what happened?
You broke even maybe?
How many Bitfury chips can you buy today?

Millions bought them and now we had decentralized mining right? 100's did that is great but that is not millions, not enough to do much more than keep some hobbyists happy.

There are things you can change. There are things you can't change. I don't see how you break a monopoly that is coming at all.

Expecting these companies to be like you vs3 in anyway is not the hope you should be waiting for. I don't see anyone going to put the time and the effort in to make an ASIC and then get miners in the hands of millions which is the only way you beat the big players. It is over unless that happens somehow.

 okay bick I play this game. bitfury can run at 0.07 watts per gh and build cheaply using cheap power.
knc can run at 0.08 watts  per gh and build it cheaply using cheap power.

lets argue why expand hash risking downward btc price pressure.  how about swap  new gear for old gear and cut power cost.
thus the two of them keep a good share of the market hash wise but simply lower expenses with more efficient gear.

  I can simply show no big self building miner should ever expand hash power
 they should keep changing the older power hungry gear for new power sipping gear.
If knc bitfury and sp-tech  do ever expanding hash rate growth they only hurt each other.  
 If they self mine with more efficient gear the hurt does not happen.

But only time will tell.

I only see KNC  doing that if at all but they are under more control now, after all they screwed us all .  SP-tech do seem to care can't say much about bitfury they seem to be more for volume buys but seem to be cool.

your right id say but time will tell.

[philipma1957]

...

I only see KNC doing that after all they screwed us all .  SP-tech do seem to care can't say much about bitfury they seem to be more for volume buys but seem to be cool .

We are in a very interesting time for BTC from now till the half-ing next year a lot of moves and double moves are to be made.

 Sfards looks like they are making a move in the next 60 days.

[toptekk]

...

I only see KNC doing that after all they screwed us all .  SP-tech do seem to care can't say much about bitfury they seem to be more for volume buys but seem to be cool .

We are in a very interesting time for BTC from now till the half-ing next year a lot of moves and double moves are to be made.

 Sfards looks like they are making a move in the next 60 days.

Cool man  i just hope there miners are affordable and every one benefits from them. and not Priced like they were way back when ..

good points . made here .

[alh]

While I can see some small advantages to a large and diverse mining base, it doesn't seem to be REQUIRED by any stretch of the imagination. Why would 10,000 30TH miners be better than 30 10PH mining farms for an aggregate of 300 PH let's say of hashrate? What's the threshold for "decentralized" and why must Bitcoin be decentralized to survive? It seems to me that in 50 years, long after I care, we'll want the most efficient "mining" (really transaction processing) setup to get by on the slender fees provided by transactions and the fraction of a new BTC in every block.

I understand the visceral reaction to a few large mining farms, but I haven't seen a compelling argument beyond the 51% attack scenario. It feels like the desire for a "miner in every home" is more driven by nostalgia of a few years ago than anything else.

[Meech]

We clearly don't need a winner. I wonder though if the price of Bitcoin won't keep ALL the ASIC vendors in check. By that I mean a  low BTC/USD price keeps them from trying to kill each other. A high BTC price may encourage a "winner take all" strategy, which may not benefit any miner long term (past a couple of difficulty adjustments).

A little competition is probably good, but not a WAR

Don't kid yourself. The price war is not the issue it is the efficient and hashrate volume that is the battle ground.

The price of these chips and units for the small and ever shrinking home market is not what will drive the end of these companies. Guy at Spondoolies already laid it out numerous times on what will kill these wannabes competing with the those with a game plan, real engineering teams and the money.

The battle for $ per gh/s is silly and meaningless given the volume of hashpower that might come online before 2016. The nuclear weapons are ready to be unleashed. This will leave only 2 or 3 main superpowers.

---

As for wanting big centralized systems based on 2 or 3 companies. NO FUCKING WAY!

Here is how we stop it.

1. Open source ASIC chip.
2. Open source low cost miner design.
3. Low cost miners in every hand.
4. P2Pool with those miners.

If you can ship more Terahash Petahash in a small, quiet, unobtrusive low cost device to millions that is the only way to decentralize things and get those miners on a P2Pool. Counting on Avalon or Bitmain or SFards as your chip / miner supplier saviour is like asking the fox to watch the hen house.

Insanity. No one learned a lesson from KnC or Bitfury did they? If you have a chip and it makes economic sense to mine massively for your benefit as company you will do it.

Be realistic this won't happen so the best we can hope for is a few more months maybe a year of projects like Gekkoscience or J4bberwock's effort.

The companies only need us to support they're own mining operations.  Once they're self sufficient well invested goodbye home mining.  The next year will definitely be the deciding factor in which way this goes.  If Btc falls below certain levels it could bring these companies to closing or downsizing as there investors fall back.  What a wonderful world it would be for open source asic designs and a strong company to do battle supplying decentralized Super miners to the world for little profit.  Greed consumes all of us at some point.

[philipma1957]

While I can see some small advantages to a large and diverse mining base, it doesn't seem to be REQUIRED by any stretch of the imagination. Why would 10,000 30TH miners be better than 30 10PH mining farms for an aggregate of 300 PH let's say of hashrate? What's the threshold for "decentralized" and why must Bitcoin be decentralized to survive? It seems to me that in 50 years, long after I care, we'll want the most efficient "mining" (really transaction processing) setup to get by on the slender fees provided by transactions and the fraction of a new BTC in every block.

I understand the visceral reaction to a few large mining farms, but I haven't seen a compelling argument beyond the 51% attack scenario. It feels like the desire for a "miner in every home" is more driven by nostalgia of a few years ago than anything else.

30 or 40 pools or farms may work.
The problems I see are why neglect the obvious : So many home owners that use electrical  heating in the forms of:

 hot water tanks
baseboard radiators
spot electric oil heaters.

BTW I use gas heat .
So I don't get much from the above.

I just don't see why potential  users of miners like these  above should be cut out. As they are spending the power anyway.

[Dabs]

Sfards, gimme a sample review unit, I will be needing the heat come winter time. (or however cold it gets here.)

[Sfards]

The power boards are  probably based on something like tps40140 or ltc3829.
They could have been ordered separately. worst case, they can be used to work on other chips later.
But I expected the power supply to be built in the dev board, and price for the dev board is almost twice what I expected, so sample chips will be fine and more fun.

Anyway, my 2 chips design only needs routing to be finished.
I've been away today and will be driving back home tomorrow (9hours trip at least), so I probably won't update until Wednesday.

Sample chips price is a little bit high too, I expected something closer to 50$ each, but if price for "small batches" is right, it doesn't really matters.
Before I buy 2 or 4 of them, I'll need some more technical informations than what can be found in the partial data sheet.
I'll check tomorrow or Wednesday for what's missing.
Can I assume the chips will be sent express (dhl/FedEx) within 1 or 2 days after payment will be made?

Yes, we will ship (DHL/FedEx) within 48 hours after full payment.

Thanks for the confirmation, here are some of my questions regarding the chips:
Can we supply non symetrical voltages to body bias?
Let's say +0.8v and -1.1v
This could be an option to improve a little bit the global efficiency.

Also, can you confirm that both pairs are supposed to receive the same voltages?
BTC_VDDS0/1
BTC_GNDS0/1
I won't mix voltages, but since there are 2 pairs of pins, it could happen on different designs

What do we do with the last chip downstream UART pins?
Left open, pulled up/down?

Your schematics, with the 4.7k resistors, seems to suggest pulled up, but even the pins in use are shown with the same 4.7k resistor.
So it's not really clear which ones do need what.

What do you call "internal oscillator"
Is it an oscillator built in the SF3301, or does it refer to the on board oscillator?

And last one for now:

Will you provide the software you used to test the dev board in your github?

Sorry for my delay to response.

As for your question, please kindly find the answers below.

         1.  for non symmetrical voltages to body bias, we don't recommend it.
         2. BTC_VDDS0/1 and BTC_GDNS0/1 are all for BTC, because the BTC has 160 cores and the area is so large, we need more pad to drive them.
         3. if you have only one chip,  for DNRX, please pullup. DNTX don't need any action.
         4. internal oscillator, the SF3301 has an internal OSCILLATOR, but it need an crystal to connect XCLKIN and ZO. so if you use it, please reference the following schematic. Please note “for internal OSC”, at the same time, please connect the 2/3 of J1, make sure the OSC_BP = 0.

http://i58.tinypic.com/2w20g7d.jpg

BTW, we will upload the software very quickly

[J4bberwock]

The power boards are  probably based on something like tps40140 or ltc3829.
They could have been ordered separately. worst case, they can be used to work on other chips later.
But I expected the power supply to be built in the dev board, and price for the dev board is almost twice what I expected, so sample chips will be fine and more fun.

Anyway, my 2 chips design only needs routing to be finished.
I've been away today and will be driving back home tomorrow (9hours trip at least), so I probably won't update until Wednesday.

Sample chips price is a little bit high too, I expected something closer to 50$ each, but if price for "small batches" is right, it doesn't really matters.
Before I buy 2 or 4 of them, I'll need some more technical informations than what can be found in the partial data sheet.
I'll check tomorrow or Wednesday for what's missing.
Can I assume the chips will be sent express (dhl/FedEx) within 1 or 2 days after payment will be made?

Yes, we will ship (DHL/FedEx) within 48 hours after full payment.

Thanks for the confirmation, here are some of my questions regarding the chips:
Can we supply non symetrical voltages to body bias?
Let's say +0.8v and -1.1v
This could be an option to improve a little bit the global efficiency.

Also, can you confirm that both pairs are supposed to receive the same voltages?
BTC_VDDS0/1
BTC_GNDS0/1
I won't mix voltages, but since there are 2 pairs of pins, it could happen on different designs

What do we do with the last chip downstream UART pins?
Left open, pulled up/down?

Your schematics, with the 4.7k resistors, seems to suggest pulled up, but even the pins in use are shown with the same 4.7k resistor.
So it's not really clear which ones do need what.

What do you call "internal oscillator"
Is it an oscillator built in the SF3301, or does it refer to the on board oscillator?

And last one for now:

Will you provide the software you used to test the dev board in your github?

Sorry for my delay to response.

As for your question, please kindly find the answers below.

         1.  for non symmetrical voltages to body bias, we don't recommend it.
         2. BTC_VDDS0/1 and BTC_GDNS0/1 are all for BTC, because the BTC has 160 cores and the area is so large, we need more pad to drive them.
         3. if you have only one chip,  for DNRX, please pullup. DNTX don't need any action.
         4. internal oscillator, the SF3301 has an internal OSCILLATOR, but it need an crystal to connect XCLKIN and ZO. so if you use it, please reference the following schematic. Please note “for internal OSC”, at the same time, please connect the 2/3 of J1, make sure the OSC_BP = 0.



BTW, we will upload the software very quickly

Thanks, you got my shipping informations, I'm now back to routing the board

[jstefanop]

Here is how we stop it.

1. Open source ASIC chip.
2. Open source low cost miner design.
3. Low cost miners in every hand.
4. P2Pool with those miners.

well - for example all you have to do is buy Bitfury's chip design and open-source it.
I have already done #2, and plenty of people did #3. #4 will happen by itself.

Yup.

And what happened?
You broke even maybe?
How many Bitfury chips can you buy today?

Millions bought them and now we had decentralized mining right? 100's did that is great but that is not millions, not enough to do much more than keep some hobbyists happy.

There are things you can change. There are things you can't change. I don't see how you break a monopoly that is coming at all.

Expecting these companies to be like you vs3 in anyway is not the hope you should be waiting for. I don't see anyone going to put the time and the effort in to make an ASIC and then get miners in the hands of millions which is the only way you beat the big players. It is over unless that happens somehow. I submit that the end is now and the monopoly on mining is already here and won't be challenged.

Well at least these guys seem to be promoting the whole open source thing, and are mainly interested in selling chips. People are already working with their designs and could have an open sourced miner pretty quickly.

[Dexter770221]

Well at least these guys seem to be promoting the whole open source thing, and are mainly interested in selling chips. People are already working with their designs and could have an open sourced miner pretty quickly.

And what that change if chips will be priced like 50$? No one will make open sourced miners becuse it will not to be economical...