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

[Bicknellski]

when do you think that they will ship a working miner?

most likely they will.  they sprung from gridseed hubs,blades & 5 pack blade.  So they did make very good LTC scrypt miners last year.

No reason they can not make one  of good quality.  Still at the same point when and how much. As always time keeps on ticking ticking into the future….

They have the money, they have the people and experience.  So they should be able to make a working miner.

What it ends up being, or what it will be on spec is unknown.  It does seem they are not in a hurry, they are taking it slow.  Likely focusing on profit not how fast to pump it out.

Who are they going to sell the majority of something that is 'cutting edge' to?

Not the home market that is for certain. They might just build and farm or like every other fabricator find a deal with someone who has the money to invest in petahash volume.

[notlist3d]

when do you think that they will ship a working miner?

most likely they will.  they sprung from gridseed hubs,blades & 5 pack blade.  So they did make very good LTC scrypt miners last year.

No reason they can not make one  of good quality.  Still at the same point when and how much. As always time keeps on ticking ticking into the future….

They have the money, they have the people and experience.  So they should be able to make a working miner.

What it ends up being, or what it will be on spec is unknown.  It does seem they are not in a hurry, they are taking it slow.  Likely focusing on profit not how fast to pump it out.

Who are they going to sell the majority of something that is 'cutting edge' to?

Not the home market that is for certain. They might just build and farm or like every other fabricator find a deal with someone who has the money to invest in petahash volume.

I still think there is a home market.  It is harder an tighter margins but it's still there.

And they already have the data centers in place to put them.  I would expect them to make quite a few of their own. They made TONS of dragon's and they plugged them in until they sold.  They would take what remaining A1's they have out and throw in new miners with new chip.  So investing does make sense for them.

[Epoch]

Much though I hate to be a party pooper, although it's very nice to see a new mining chip design, especially a dual mode one,
this design is really nothing new, SOI or not. Low w/(Gh/sec) figures can be achieved through simply lowering the clock speed
and/or chip voltage, if you look at the power specs you see that the device in SHA256 mode takes just under 78W at 0.8V at
870Mhz with a hashrate of 139GH/sec. ...

I like your lines of thought, bronto. But let's be careful not to get too far into the speculative world; it may be interesting and entertaining to do, but that's as far as it goes. At this point all we have is a datasheet. The rest is, to various levels, speculation.

I will admit that, based on the datasheet, my expectations for products based on this chip are tempered. Bitmain's BM1384 and Spondoolies's Rockerbox system designs approach ~0.4W/GH at the system level (never mind chip level) when underclocked/undervolted. For example an SP35 at 3400GH/s uses 1500W at the wall ... a system efficiency of 0.44W/GH. KNC's Neptune is also in that range.

If an Sfards system achieves 0.4W/GH at the system level it may not be a significant enough improvement over what's already available. It all depends on system cost, though, which is something we don't have a clue about. So I'm quite happy to remain patient and wait for an actual system demo, with measured specs, rather than speculating at this point.

[philipma1957]

Much though I hate to be a party pooper, although it's very nice to see a new mining chip design, especially a dual mode one,
this design is really nothing new, SOI or not. Low w/(Gh/sec) figures can be achieved through simply lowering the clock speed
and/or chip voltage, if you look at the power specs you see that the device in SHA256 mode takes just under 78W at 0.8V at
870Mhz with a hashrate of 139GH/sec. ...

I like your lines of thought, bronto. But let's be careful not to get too far into the speculative world; it may be interesting and entertaining to do, but that's as far as it goes. At this point all we have is a datasheet. The rest is, to various levels, speculation.

I will admit that, based on the datasheet, my expectations for products based on this chip are tempered. Bitmain's BM1384 and Spondoolies's Rockerbox system designs approach ~0.4W/GH at the system level (never mind chip level) when underclocked/undervolted. For example an SP35 at 3400GH/s uses 1500W at the wall ... a system efficiency of 0.44W/GH. KNC's Neptune is also in that range.

If an Sfards system achieves 0.4W/GH at the system level it may not be a significant enough improvement over what's already available. It all depends on system cost, though, which is something we don't have a clue about. So I'm quite happy to remain patient and wait for an actual system demo, with measured specs, rather than speculating at this point.

yeah pretty much agree we have 19 pages of what if.

And still waiting for a working demo of the gear they will sell.

 BTW I had 1 sp20 give me a steady .42 watts on a plat psu  with a solid underclock.

[J4bberwock]

.....
Really neat, but we still don't know if this is at the wall yet?

It's a chip datasheet. "At wall" depends on many factors. DC/DC converter and PSU efficiency. So, if thats not specified all data covers only chip.

up to 25% extra power draw is to be expected at the wall with a gold rated PSU, depending on the design:
10-12% for the PSU, and 10-15% for the power converter that will drop your +12v to 0.6-0.9v

When using string design, you'll save almost all the loss from the power converter.
That's what you can expect from the "24 chips design" they announced at the same time as the 1 chip design.
I still believe that 18-20 chips would be a better choice if they go the string design route.

If some of you are interested in kicad footprint and library for the SF3301/GC3328 chip, I finished mine a few minutes ago.
I'll double check everything tomorrow and share with those interested.

Could you create a  Github repo or DropBox folder to share your work ? Do you plan on making your designs available  to the community ?

Here is a dropbox link with the Kicad footprint and library.

https://www.dropbox.com/sh/cjjkrq1fv2u7m0l/AADU4n7cCJsl5N1E6Qtai5M5a?dl=0

First trial at dropbox, so I'm not 100% sure it'll work

My schematics for a 2 chips board will probably be public.
I still need to create a few components to finish it.

Summary:
BTC side:
6 phases buck controller for 200+ amps available
Voltage selectable with dip switches from 0.50 to 0.81v (can be unlocked to 1.21v)
LTC side:
Same power circuit, but with 2-3 phases used
Communication interface:
USB communication with CP2105/2108

I also kept the ability to daisy-chain boards like in the reference design and a more simple adjustable body bias circuit from +/-0.5v to +/-1.1v

[HyperMega]

I have to admit that I missed that, my estimate of the size of the chip is 12.5 -13.2 mm on a side, equating to a packaged device
cost (in volumes of 2 million pa) of $11.4. With the actual dies size of 10.5 mm on a side, that price comes down to $9.1, saving
$37 per 16 chip system. That does make a difference.

$9.1 production costs per packaged chip?

Assuming 90% yield they should have about 560 yielding dies per 28FDSOI wafer. The flip-chip substrate and assembly is most likely not below $1 per chip. Under this assumptions your price estimation would translate in a wafer price of less than $4500.
One gets maybe such prices if you talk with foundries about revenues in the range of 1 billion US$ but not for a "tiny" Bitcoin deal.

I would speculate that their chip production costs (including packaging) are more in the range of $15 and they can't sell below $20, because as you said they also have to get their NRE back. If they have yield issues the price could be even much higher. But we will see.

Overall I agree with your conclusion, this chip will be simply to expensive to be a game changer.

[Sfards]

SFARDS’ SF3301 sample chip is available on May 4th

SFARDS is now selling a few SF3301 development boards and sample chips to developers who wish to customize and build their own hardware.

http://i59.tinypic.com/xkyt4z.jpg

http://i57.tinypic.com/1693m7o.jpg

Sales Information:

Sample chip:                      200 USD per pack, include two chips (exclude shipping)
Development board set:      400 USD (exclude shipping)
Shipping info:                     Within 48 hours after full payment (Payment by Bitcoin or USD)

For more questions such as place order or large scale orders, please email sales@sfards.com or add our skype:sfardsservice.

Features:

Sample chip:
More details about SF3301 chips, please kindly find the datasheet.
https://github.com/sfards

Development board set (includes one dev board, two power supply boards, radiator and fan):
Main Board
SF3301 : 1 chip
Heatsink : heatsink & 12mmx12mm fan
FBB Voltage : +/- 0~1.0V
Communication Interface : RS232 / TTL

Power Supply Board:
Output Voltage Range : 0.5~1.6V
Voltage Step:6.25mV
Max Current : 120A

[dogie]

Hurrah. Appropriate pricing on the development boards as well, good job.

[adaseb]

WOW! That chip is pretty small, I assumed it would be the same size as KNCMiners or SP ASIC chips.

[Dexter770221]

Hurrah. Appropriate pricing on the development boards as well, good job.

Those are almost always quiet expensive. But price of sample chips is ridicolous. It will scare developers. In this industry (chips production) sample chips are very often given for free. They just want to charge 10 times of their price. If chip has so high price miner can't be cheap, simple logic. SFARDS is shooting to their own foot,or they just don't care and want to mine for themselfs. Then, why all this hassle with dev board and sample chips? I'm so confused...

[notlist3d]

Very nice looking dev kit!  Very impressive.  Cant wait to see some people put it though it's paces.

[ashleysly]

Now all we need is "How to build Minter with SF3301 Chip, for dummies"

[dogie]

WOW! That chip is pretty small, I assumed it would be the same size as KNCMiners or SP ASIC chips.

Hard to tell but looks to be about that of Blackarrow's chip, so we know that they could get at least 100W out of it. But yeah smaller than I expected.

Hurrah. Appropriate pricing on the development boards as well, good job.

Those are almost always quiet expensive. But price of sample chips is ridicolous. It will scare developers. In this industry (chips production) sample chips are very often given for free. They just want to charge 10 times of their price. If chip has so high price miner can't be cheap, simple logic. SFARDS is shooting to their own foot,or they just don't care and want to mine for themselfs. Then, why all this hassle with dev board and sample chips? I'm so confused...

The point of the pricing is that only those who will actually do something useful with them will buy them. Yes its high, but yes supply is extremely limited.

[pekatete]

I am thinking the price of the sample chips reflects, to an extent, the final retail price of chips to future buyers. That, though, would rule out making any market-worthy miner. So sfards, is this sample chip pricing only effective until you release your own rigs, or (assuming there are no further improvements in chip tech) is this the real price we'd also have expect after you release your own rigs?

The point of the pricing is that only those who will actually do something useful with them will buy them. Yes its high, but yes supply is extremely limited.

Waffle! What does it cost them if they sold the sample chips at a fair price and they are not put to good use? In any case, what is good use ...? And by the way, sfards is not bitmain where you get the privellege of taking pretty pictures of their rigs accompanied by waffle.

[philipma1957]

Okay here we go I will not and can not develop a miner from that kit.  I am fine with that.

Now for developers, here is the only: question do they want to spend 400 to make a design without knowing chip prices?

That is up to them. Does it encourage a lot of developers no it does not, but how many groups can take the kit and make a miner from it 20?

I have been on this site for years.  As I see it  right now   maybe 20 to 30  could take the kit and make something of it.

Should they pay 400 that is up to those that know they can build something from it.  They also need belive that they will get more chip when they need them at a fair price.

I am still waiting for a plug n play on this  and it still looks to be too little too late.

[TheRealSteve]

Hard to tell but looks to be about that of Blackarrow's chip, so we know that they could get at least 100W out of it. But yeah smaller than I expected.

Not that hard to tell - the chip silicon and substrate sizes are in the datasheet after all.  SF3301 substrate is 19x19mm.  Minion is 25x25mm - a fair bit larger, while KnC's public offerings were a massive 55x55mm.
It should be the same size as Spondoolies-Tech's RockerBox, which you're familiar with. 
( A bunch of this data is in the wiki )

[Meech]

Okay here we go I will not and can not develop a miner from that kit.  I am fine with that.

Now for developers, here is the only: question do they want to spend 400 to make a design without knowing chip prices?

That is up to them. Does it encourage a lot of developers no it does not, but how many groups can take the kit and make a miner from it 20?

I have been on this site for years.  As I see it  right now   maybe 20 to 30  could take the kit and make something of it.

Should they pay 400 that is up to those that know they can build something from it.  They also need belive that they will get more chip when they need them at a fair price.

I am still waiting for a plug n play on this  and it still looks to be too little too late.

Seems a reasonable expense for those qualified to do so especially if they receive them a month before Sfards first product ships.

[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?

[starsoccer9]

hmm added you guys on skype, but havent heard a word yet.

[TheRealSteve]

There's a bit more info + pictures in Bitcoin Magazine's feature: New Mining Chip Developed by SFARDS Becomes Most Efficient Chip Produced
( Presuming it's an implementation of the reference schematic, which looks likely, components are effectively listed there. )