Swedish ASIC miner company kncminer.com

[Phoenix1969]

We need to ask support for a hosting plan for the Saturns.

Depending upon the price of bitcoin and difficultly, it's very likely that a hosting plan won't be inexpensive enough for KnC or profitable enough for the customer. You can ask, but KnC is doing this because it's not in Saturn owners' interest to pay hosting fees.

Possibly even $350 for Jupiter is alot, when if you host yourself electricity will be about $90/mo on the high end.

For me it's $150/month for electricity, but yes - it's a question about how profitable it will end up being. I did the calculation and if I factor in shipping + power + possible downtime it comes out to about a wash for 6 months.

350 alot...ok, assuming the $90 estimate is accurate, that's only 260 for the service, with the best tweaks from the developers themselves.... sounds great to me... not o mention the fact you wont have to wait on shipping, so the first few days will pay for alot of hosting, hah

[JohnyBigs]

We need to ask support for a hosting plan for the Saturns.

says right in news release that KNC will only host Jupiters

They take feedback just like they have with hardware im sure they can figure a solution if enough people ask for it. Doesnt seem to hard the same space is going to be taken up with less power.

[soniq]

We need to ask support for a hosting plan for the Saturns.

says right in news release that KNC will only host Jupiters

They take feedback just like they have with hardware im sure they can figure a solution if enough people ask for it. Doesnt seem to hard the same space is going to be taken up with less power.

Cant hurt to ask, but that is their published response right now

[JohnyBigs]

We need to ask support for a hosting plan for the Saturns.

Depending upon the price of bitcoin and difficultly, it's very likely that a hosting plan won't be inexpensive enough for KnC or profitable enough for the customer. You can ask, but KnC is doing this because it's not in Saturn owners' interest to pay hosting fees.

Possibly even $350 for Jupiter is alot, when if you host yourself electricity will be about $90/mo on the high end.

For me it's $150/month for electricity, but yes - it's a question about how profitable it will end up being. I did the calculation and if I factor in shipping + power + possible downtime it comes out to about a wash for 6 months.

350 alot...ok, assuming the $90 estimate is accurate, that's only 260 for the service, with the best tweaks from the developers themselves.... sounds great to me... not o mention the fact you wont have to wait on shipping, so the first few days will pay for alot of hosting, hah

It can be a double edged sword, say a forum member comes up with a tweak or solution and the devs want to take their time on an official solution. Yeah saving the shipping time is the best.

[JohnyBigs]

We need to ask support for a hosting plan for the Saturns.

says right in news release that KNC will only host Jupiters

They take feedback just like they have with hardware im sure they can figure a solution if enough people ask for it. Doesnt seem to hard the same space is going to be taken up with less power.

Cant hurt to ask, but that is their published response right now

I know i asked them anyway, i plan on upgrading my saturn to a jupiter anyway within the first month, dont see the logic behind not hosting saturns or mercuries if people want to pay for it.

[Loredo]

3d transistors, energy efficiency, memristors, graphene whatever...
Do you think chip evolution will stop with some Orsoc innovating? That would be bad!

This sounds like stuff you pay millions of bucks to human and physical capital for over years, and when you get it right, you better have a market that has room of lots of millions of units.

Any anyway, why would anyone here do that?  AM has the cash flow to get all they need even though their chip is the equivalent of a 1994 Buick.  Same with Avalon. 

And bitfury: why would they do that?  If the comparison table above is right, bitfury has already won with what they have.  They have a stable process, a tiny die, a network of volunteers who have helped bring the chip to marketable reality, a hungry retail base that will pay all of their development NRE and other incurred pre-launch costs, and they can get 9 TH per wafer. 

If you had that cash cow, why would you entertain an idea of engaging in a new dairy breeding program?

[Phoenix1969]

3d transistors, energy efficiency, memristors, graphene whatever...
Do you think chip evolution will stop with some Orsoc innovating? That would be bad!

This sounds like stuff you pay millions of bucks to human and physical capital for over years, and when you get it right, you better have a market that has room of lots of millions of units.

Any anyway, why would anyone here do that?  AM has the cash flow to get all they need even though their chip is the equivalent of a 1994 Buick.  Same with Avalon.  

And bitfury: why would they do that?  If the comparison table above is right, bitfury has already won with what they have.  They have a stable process, a tiny die, a network of volunteers who have helped bring the chip to marketable reality, a hungry retail base that will pay all of their development NRE and other incurred pre-launch costs, and they can get 9 TH per wafer.  

If you had that cash cow, why would you entertain an idea of engaging in a new dairy breeding program?

WHAT?   9 Terrahash per wafer??   where? plz post link, or I'm calling BS on that right now!

[Loredo]

If the comparison table above is right, bitfury ... can get 9 TH per wafer. 

WHAT?   9 Terrahash per wafer??   where? plz post link

https://bitcointalk.org/index.php?topic=170332.msg2721253#msg2721253

[Phoenix1969]

Just as I thought... 2 gh/s per chip, about 50x slower, you had me there

[bkpduke]

This is the best piece of info in this entire discussion.  Unfortunately Fabs almost never release their defect density and yields, unless you are a customer under a Non-Disclosure Agreement.  In almost two decades following CPU/IC production I have never willingly seen a fab release this info.  Very rarely you can get it leaked from a former employee, but then it is always tainted and uncertain if that person has an axe to grind with his former employer or not.  Alternatively, sometimes, you will get vague impressions from the executives during financial disclosures, but never the detailed info you want.

Now, the only thing apparently known about KNC is that the package size is about 3000 mm2.  Let's assume I am wrong, would not be the first time, and that they have a relatively small chip for the package size.  Say 15-20%.  That would be a die size of 450-600 mm2.  That would still be BELOW the lowest curve on the above diagram.  That's really scary for yields, but not impossible to obtain working product by history.  But any customer with a die size that large would have to expect very low yields, and they probably had to sign a waiver from the foundry about guaranteed yield due to "design constraints".

Here are some historical data points for 3rd party fab yield problems with some of the larger GPU dies, for bigger customers:
http://www.geek.com/games/ati-and-nvidia-have-troubling-40nm-yields-from-tsmc-815501/  which were "fixed" a year later:
http://www.dailytech.com/TSMC+Says+40nm+Problems+Resolved+Preparing+28nm+Fab+Production+/article17355.htm
http://www.eetimes.com/document.asp?doc_id=1261006  which were "fixed" about 10 months later:
http://www.pcmag.com/article2/0,2817,2411985,00.asp

Now, a TSMC/UMC is going to work with an AMD/NVidia to improve yields because they are producing 10-100s of millions of chips.  For a smaller companies, they will not go to great lengths to help them revise designs and processes to improve yields.  You get what you get.

Traditionally GPU dies have been some of the largest dies, with the high-end (low yield) parts topping out around mid-500 mm2 in size.  And the parallel structure of a GPU is what most closely resembles the SHA256 ASICs being designed and produced by companies today (Avalon, BFL, KNC, BitFury, etc.).  So from a manufacturing standpoint, we can probably infer a lot from large ASIC production issues by looking at issues that GPU manufacturers have had in the past.

The bigger the dies, the lower the yield, the more you have to charge to cover your costs on the wasted dies from a wafer.  Therefore it is always good to make your dies as small as possible.  Just 10 mm2 per die, when factored over an entire wafer, can mean tens of thousands of dollars in improve yield per wafer.

For comparison, here is a good illustration from the 45 nm days, with Intel - who historically have much higher yields than the rest of the industry (the invest more in their process and refinement).



Again, I hope KNC brings to market a good product, and in a timely fashion.  The more competition, the better we all do in terms of up-front pricing for our miners.  I just am concerned that we haven't see any substantial data about the processor (number of SHA256 cores, clock frequency, die size, heat production, etc.).

Considering the 10+ million dollars of orders they have had, you would hope they would publish this info to keep their investors/customers happy.

[Phoenix1969]

Who cares about that? Not me: it's about end user performance & efficiency, period.
By that I mean: He with the most efficient machine(s) will be the last man standing.
To me, nothing else matters.

[bkpduke]

who cares about that...not me, it's about end user performance & efficiency, period.
By that I mean: He with the most efficient machine will be the last man standing.

The why you should care is because low chip yields = fewer chips = slower production = delays in getting to customers.

You can have the most efficient design in the world, but if you cannot reliably produce it or make enough to sell in any significant quantity, it's worthless.

[bebfoo]

Well researched post bkpduke! 

[Phoenix1969]

who cares about that...not me, it's about end user performance & efficiency, period.
By that I mean: He with the most efficient machine will be the last man standing.

The why you should care is because low chip yields = fewer chips = slower production = delays in getting to customers.

You can have the most efficient design in the world, but if you cannot reliably produce it or make enough to sell in any significant quantity, it's worthless.

You think they can't do what they say?    I believe in them (KNC)100%
Sir Isaac Asimov: "Mathemiticians use data from the past to predict the future"
Orsoc is about the only asic developer with a past.

[bkpduke]

who cares about that...not me, it's about end user performance & efficiency, period.
By that I mean: He with the most efficient machine will be the last man standing.

The why you should care is because low chip yields = fewer chips = slower production = delays in getting to customers.

You can have the most efficient design in the world, but if you cannot reliably produce it or make enough to sell in any significant quantity, it's worthless.

You think they can't do what they say?    I believe in them (KNC)100%

Good, continue to drink the Koolaid.  I'll continue to crunch numbers and look at data.

[canth]

who cares about that...not me, it's about end user performance & efficiency, period.
By that I mean: He with the most efficient machine will be the last man standing.

The why you should care is because low chip yields = fewer chips = slower production = delays in getting to customers.

You can have the most efficient design in the world, but if you cannot reliably produce it or make enough to sell in any significant quantity, it's worthless.

Die size, defects, power efficiency - who cares about *blah, blah* 40 years of microprocessor production history? I like the magic money production factory description better too.

Keep up on the education, bkpduke - some will care.

[Phoenix1969]

Please do, We're tired of hearing you whine about nonsense. no other company can compare. I like the coolaid, it's a winning recipe, thank you.

[bkpduke]

Please do, We're tired of hearing you whine about nonsense. no other company can compare. I like the coolaid, it's a winning recipe, thank you.

Funny that KNC's "more advanced" miner that isn't out yet will consume "up to" 1kW to produce 400GH/s.

Bitfury, which is demo'ed at least, does the same 400GH/s with "up to" 400W.

Hrm . . . they are going to need to improve their "efficiency" by 2.5 X just to get even with the competition.

[2112]

The bigger the dies, the lower the yield, the more you have to charge to cover your costs on the wasted dies from a wafer.

Your analysis is 100% correct and 100% inapplicable to the SHA-256 hashing chip.

You cannot compare yield of a supremely-complex OoO CPU like x86 with a dumb repetitive SHA-256 hasher.

Nearly any defect on a CPU is a chip-killer, probably with an exception of defect in a cache line or virtualization support when then chip can still be sold as a cheaper model.

If the KnC designed their chip with redundand I/O (multiple clock pins, etc) the probabiity that the particular defect is chip killing is very close to zero. The defective chips will simply have lower performance.

All in all, you've made a competent but nearly nonsensical post.

Edit: Included a full quote to insure against modifications.

This is the best piece of info in this entire discussion.  Unfortunately Fabs almost never release their defect density and yields, unless you are a customer under a Non-Disclosure Agreement.  In almost two decades following CPU/IC production I have never willingly seen a fab release this info.  Very rarely you can get it leaked from a former employee, but then it is always tainted and uncertain if that person has an axe to grind with his former employer or not.  Alternatively, sometimes, you will get vague impressions from the executives during financial disclosures, but never the detailed info you want.

Now, the only thing apparently known about KNC is that the package size is about 3000 mm2.  Let's assume I am wrong, would not be the first time, and that they have a relatively small chip for the package size.  Say 15-20%.  That would be a die size of 450-600 mm2.  That would still be BELOW the lowest curve on the above diagram.  That's really scary for yields, but not impossible to obtain working product by history.  But any customer with a die size that large would have to expect very low yields, and they probably had to sign a waiver from the foundry about guaranteed yield due to "design constraints".

Here are some historical data points for 3rd party fab yield problems with some of the larger GPU dies, for bigger customers:
http://www.geek.com/games/ati-and-nvidia-have-troubling-40nm-yields-from-tsmc-815501/  which were "fixed" a year later:
http://www.dailytech.com/TSMC+Says+40nm+Problems+Resolved+Preparing+28nm+Fab+Production+/article17355.htm
http://www.eetimes.com/document.asp?doc_id=1261006  which were "fixed" about 10 months later:
http://www.pcmag.com/article2/0,2817,2411985,00.asp

Now, a TSMC/UMC is going to work with an AMD/NVidia to improve yields because they are producing 10-100s of millions of chips.  For a smaller companies, they will not go to great lengths to help them revise designs and processes to improve yields.  You get what you get.

Traditionally GPU dies have been some of the largest dies, with the high-end (low yield) parts topping out around mid-500 mm2 in size.  And the parallel structure of a GPU is what most closely resembles the SHA256 ASICs being designed and produced by companies today (Avalon, BFL, KNC, BitFury, etc.).  So from a manufacturing standpoint, we can probably infer a lot from large ASIC production issues by looking at issues that GPU manufacturers have had in the past.

The bigger the dies, the lower the yield, the more you have to charge to cover your costs on the wasted dies from a wafer.  Therefore it is always good to make your dies as small as possible.  Just 10 mm2 per die, when factored over an entire wafer, can mean tens of thousands of dollars in improve yield per wafer.

For comparison, here is a good illustration from the 45 nm days, with Intel - who historically have much higher yields than the rest of the industry (the invest more in their process and refinement).



Again, I hope KNC brings to market a good product, and in a timely fashion.  The more competition, the better we all do in terms of up-front pricing for our miners.  I just am concerned that we haven't see any substantial data about the processor (number of SHA256 cores, clock frequency, die size, heat production, etc.).

Considering the 10+ million dollars of orders they have had, you would hope they would publish this info to keep their investors/customers happy.

[Phoenix1969]

BS. You are totally lying now. Welcome to the list of trolls.(bdpuke)
Bitfury's 110gh rack is HUGE, and consumes 10 kilowatts!!!!
http://www.bitfury.org/bitfury110.html