Slush and bitparking are real pools. They INVENTED both pool mining and merged mining respectively. I wouldn't touch knc with a 10 foot pole. They have no experience.
Yet, you purchased multiple devices from them??!
Topic: Swedish ASIC miner company kncminer.com - page 1895. (Read 3049501 times)
Yet, you purchased multiple devices from them??!
Slush and bitparking are real pools. They INVENTED both pool mining and merged mining respectively. I wouldn't touch knc with a 10 foot pole. They have no experience.
and 3 months ago, KnC was an idea to "make lots of money".
No, three months ago you found out about their plan when they went public. We don't in fact know how long this has been in planning behind the scenes, but probably at least since January, and even more likely into 2012. By the time we heard about it, things were already at a very advanced stage we must assume, for they already by then had a deal with Orsoc, which we heard about almost immediately after, and such deals take time to organize, communicate and get done, and they also had in place their design which was nearly ready to tape out, and they'd already planned their Mars miner which they executed in less than 6 weeks into a running miner, which is fairly impressive.I have to balk at your assertion that 3 months ago KNC was a twinkle in someone's eye. This company has done an unknown amount of planning and work prior to their public debut, and we probably won't hear the retrospective on that until after they ship the first gen and have breathing room to do interviews and throw parties and the like.
+3000TH. That sounds very scary for anyone wanting to meet ROI. Anyway...
Would they need a 55mmx55mm package for a 11mmx11mm chip??
Thus why I got a refund on my Avalon :\ Would they need a 55mmx55mm package for a 11mmx11mm chip??
All they need is Dave from Bitfury and Asicminer and BOOM.....CARTEL !!!!
paranoid thinking, now we saw an avalon - bfl "delay'n mine "alliance
let's wait... september is coming
Why would you even use their pool?
Use a real pool or run your own p2pool node.
Use a real pool or run your own p2pool node.
The first reply from KNC regarding enabling NMC merge mining within the KNC pool was that the miners only support BTC. I did my best to explain where this configuration was set and tried to further explain that this is not a hardware change.
I asked them to forward the question to the operator of the pool if they knew that at this time-- if not we can revisit this later this month.
No response from byteminr yet.
I asked them to forward the question to the operator of the pool if they knew that at this time-- if not we can revisit this later this month.
No response from byteminr yet.
use a "real" pool? how do you define "real"? KNC will have a huge chunk of hashing power.
I believe the level of trust for a pool run and managed by KNC would be higher than a pool managed by an anonymous person. It's possible that the pool operator could skim coins. I would hope the KNC pool would be protected against DDOS attacks and other breach attempts. The fees are 1.5% which are lower than other available pools. The pool will should also be monitored by the hosting facility for 99.9% up-time.
The previous question about merge mining not available on stratum, I believe this is not the case. BitMinter uses merge mining and statum but yes, they also have getwork.
If the devices are even 100GH/s, using getwork would be bad since ... well this has been discussed over and over, go read up about it
But simply, once you get around 500GH/s and more, it's a risk to lose blocks on the network using getwork - which gets higher the more people who are doing this (read up about roll-n-time)
Stratum does not stop merged mining, it's simply a question of implementing it.
... and who are these so called "anonymous person" managing pools?
Yes there are a lot of small anonymous pools, and yes avoid them, but unless you're a noob around bitcoin, you would know well that there are indeed quite a few pools with VERY well known operators ... however ...
Is it some pool run by one of the well known names in KnC (well there are only a few well known names in KnC) or is it run by some anonymous person no one has heard of ...
If KnC says officially: We back this pool and we will cover the losses if the person running the pool does a runner, then yep that's OK, but until then, it is exactly as you describe "a pool managed by an anonymous person"
Why would you even use their pool?
Use a real pool or run your own p2pool node.
Use a real pool or run your own p2pool node.
The first reply from KNC regarding enabling NMC merge mining within the KNC pool was that the miners only support BTC. I did my best to explain where this configuration was set and tried to further explain that this is not a hardware change.
I asked them to forward the question to the operator of the pool if they knew that at this time-- if not we can revisit this later this month.
No response from byteminr yet.
I asked them to forward the question to the operator of the pool if they knew that at this time-- if not we can revisit this later this month.
No response from byteminr yet.
use a "real" pool? how do you define "real"? KNC will have a huge chunk of hashing power.
I believe the level of trust for a pool run and managed by KNC would be higher than a pool managed by an anonymous person. It's possible that the pool operator could skim coins. I would hope the KNC pool would be protected against DDOS attacks and other breach attempts. The fees are 1.5% which are lower than other available pools. The pool will should also be monitored by the hosting facility for 99.9% up-time.
The previous question about merge mining not available on stratum, I believe this is not the case. BitMinter uses merge mining and statum but yes, they also have getwork.
Why would you even use their pool?
Use a real pool or run your own p2pool node.
Use a real pool or run your own p2pool node.
The first reply from KNC regarding enabling NMC merge mining within the KNC pool was that the miners only support BTC. I did my best to explain where this configuration was set and tried to further explain that this is not a hardware change.
I asked them to forward the question to the operator of the pool if they knew that at this time-- if not we can revisit this later this month.
No response from byteminr yet.
I asked them to forward the question to the operator of the pool if they knew that at this time-- if not we can revisit this later this month.
No response from byteminr yet.
The first reply from KNC regarding enabling NMC merge mining within the KNC pool was that the miners only support BTC.
That's reassuring. (Sarcasm) Maybe he meant that they only accept Bitcoin for their non-existent miners. Since they are just imaginary, I suppose you could assume they mine whatever you want them to.
The first reply from KNC regarding enabling NMC merge mining within the KNC pool was that the miners only support BTC. I did my best to explain where this configuration was set and tried to further explain that this is not a hardware change.
I asked them to forward the question to the operator of the pool if they knew that at this time-- if not we can revisit this later this month.
No response from byteminr yet.
It's my understanding that NMC still uses Getwork protocall, not stratum.... ? Could do both at same time? I asked them to forward the question to the operator of the pool if they knew that at this time-- if not we can revisit this later this month.
No response from byteminr yet.
$800k for a 28nm full-mask set? This is an amazing offer, at which foundry will I get it that cheap?
That I do not believe KnC is producing a fully designed asic should be obvious by now.
That only works if the foundry is convinced that you are worthy of credit. Normally this is not the case for start-ups and SMEs in this price category. For them it is cash in advance or no tape-out. Yes, I know what I'm talking about, it's hard to deal directly with companies like GF or TSMC if your are only a small customer.
I doubt that KnC or OrSOC is seen as a big customer for this project, because they probably will not order more than 50 wafers in the first year (which would be already about 3 PH). 1000 wafers a year would be a big customer. Why should a foundry take any risk for a bitcoin mining ASIC?
So I think they most likely have to pay the complete $1.5M to $2M in advance of the tape-out. And because they said tape-out done "some time ago", they probably paid it also some time ago.
From the commercial point of view I see no reason why KnC should not be able to produce a fully designed 28nm ASIC. They collected enough money by pre-orders to pay the complete 3rd party NREs in advance. But it should be clear, that if they are on track as they claimed, a great deal of the pre-order money isn't in their hands any more.
4. My guess, which people have posted that it is not true, is they are going with eAsic.
Yes, I'm not that far away, but do not agree with all of your points:
1. OrSOC and KnC need a channel partner of one of the two 28nm foundries to get access at all. eASIC would be a choice (also for a standard cell ASIC!), eSilicon or GUC are others. There are many and they love to take your money for that kind of projects. In this way there is no background check by the foundry required.
2. From my own experience I know that not all 28nm "giga" fabs running currently at there capacity limit. If you bring the money (as cash in advance
) through an accepted channel partner they will do a tape-out for you. They will for sure make you not the best price for masks and wafers, but they will do it, just to get workload for their fabs.3. Physical implementation and sign-off of a >100mm2 28nm die in about 6 weeks (because mid of June they didn't even have the foundry finally selected and they did tape-out "some time ago") is really a challenge, but not completely impossible. The related risk to fail for such a turbo implementation is huge. I hope they had a really experienced layout team for this job, which already did some tape-out in the concrete 28nm node.
eASIC is GF, Europe.
Altera (my guess) is TSMC, Taiwan, Asia.
ORSoC has a working relationship with both, both do standard cell and take the RTL code. I guess they just had to find the right balance between cost and delivery time. Since Asia won, that very much means they didn't choose eASIC. And they're not dealing with the foundry directly either (hoping they wouldn't take the risk of dealing with a non established partner which would really risk screwing up their timetable).
Also, don't forget they promised a September delivery, NOT a 1st of September delivery. That's 4 weeks to play with.
As far as I know, Altera offers FPGA-Hardcopy only. KnC said several times that they are doing definitely a standard cell ASIC design, no FPGA-Hardcopy.
Their released R&D slides are looking as a typical standard cell ASIC too.
So it's for sure not Altera and hopefully not eASIC, because they have no proven 28nm experience so far. I guess there is a new partner.
Their ambitious delivery dates I rather do not comment.
You're right, HardCopy is not standard cell. Well, if it isn't eASIC or Altera, then it's someone else we don't know about yet. Or it's ON Semi (they have an agreement with Flextronics, whose Flextronics Semi subsidiary eventually ended up at ON Semi - this fits, unless it's the "wrong" Flextronics I'm talking about).
edit: it could also be Synopsys (through the Magma acquisition), since they have/had a tripartite agreement with Flextronics/eASIC.
The first reply from KNC regarding enabling NMC merge mining within the KNC pool was that the miners only support BTC. I did my best to explain where this configuration was set and tried to further explain that this is not a hardware change.
I asked them to forward the question to the operator of the pool if they knew that at this time-- if not we can revisit this later this month.
No response from byteminr yet.
I asked them to forward the question to the operator of the pool if they knew that at this time-- if not we can revisit this later this month.
No response from byteminr yet.
CMP offers access to STMs Multi-Project-Wafer runs for SMEs and academia. STM is probably happy about some money coming from this source to finance the MPW mask costs, because they have not enough own content. MPWs are for prototyping runs only. They are not suitable for volume production.
I'm not sure if STM will do a full-mask (not MPW) tape-out for everybody too. For customers ordering 1000 wafers of course, but for one who orders only 50?
Anyway, from the definition STM is an IDM (Integrated Device Manufacturer) not an "pure play" foundry without own chip products. Did you know that STM is a big customer of GF? The FDSOI technology mentioned above will finally also run in Dresden at the GF fab 1 (under STM licence) for volume production.
STM is apparently making them in Crolles, France; not in any facility that you've enumerated.I'm not sure if STM will do a full-mask (not MPW) tape-out for everybody too. For customers ordering 1000 wafers of course, but for one who orders only 50?
Anyway, from the definition STM is an IDM (Integrated Device Manufacturer) not an "pure play" foundry without own chip products. Did you know that STM is a big customer of GF? The FDSOI technology mentioned above will finally also run in Dresden at the GF fab 1 (under STM licence) for volume production.
Quote from: STMicroelectronics
Following the Company’s December announcement of the successful manufacturing of System on Chip (SoC) integrated circuits, ST today announced that application-processor engine devices manufactured at the Company’s Crolles, France fab, were capable of operating at 3GHz with even greater power efficiency at a given operating frequency than alternate technologies.
http://www.st.com/web/en/press/en/t3405Who cares about "volume" production in the way "volume" is used in the semiconductor industry. KnC isn't and isn't going to reach that size of the orders. "pure play" vs. IDM distinction also doesn't really matter. What does matter is that the list of fabs capable of making KnC chips is much longer that you've proposed. Because the Bitcoin miner is essentially free from intelectual property encumberances it is also free from many other manufacturing constraints typical in the industry.
The question is not how many IDMs/"pure play" foundries are capable of 28nm, the question is to how many of thme a small customer gets access for a 50 wafer project.
I don't like to argue if STM belongs to this list or not. Maybe, then my list has 3 entries instead of 2. Would be a much longer list!
Regarding the STM/GF relations and their future plans you should have a look at this press release (slide 6):
http://www.globalfoundries.com/technology/pdf/GF-14XM-Press-FINAL.pdf
CMP offers access to STMs Multi-Project-Wafer runs for SMEs and academia. STM is probably happy about some money coming from this source to finance the MPW mask costs, because they have not enough own content. MPWs are for prototyping runs only. They are not suitable for volume production.
I'm not sure if STM will do a full-mask (not MPW) tape-out for everybody too. For customers ordering 1000 wafers of course, but for one who orders only 50?
Anyway, from the definition STM is an IDM (Integrated Device Manufacturer) not an "pure play" foundry without own chip products. Did you know that STM is a big customer of GF? The FDSOI technology mentioned above will finally also run in Dresden at the GF fab 1 (under STM licence) for volume production.
STM is apparently making them in Crolles, France; not in any facility that you've enumerated.I'm not sure if STM will do a full-mask (not MPW) tape-out for everybody too. For customers ordering 1000 wafers of course, but for one who orders only 50?
Anyway, from the definition STM is an IDM (Integrated Device Manufacturer) not an "pure play" foundry without own chip products. Did you know that STM is a big customer of GF? The FDSOI technology mentioned above will finally also run in Dresden at the GF fab 1 (under STM licence) for volume production.
Quote from: STMicroelectronics
Following the Company’s December announcement of the successful manufacturing of System on Chip (SoC) integrated circuits, ST today announced that application-processor engine devices manufactured at the Company’s Crolles, France fab, were capable of operating at 3GHz with even greater power efficiency at a given operating frequency than alternate technologies.
http://www.st.com/web/en/press/en/t3405Who cares about "volume" production in the way "volume" is used in the semiconductor industry. KnC isn't and isn't going to reach that size of the orders. "pure play" vs. IDM distinction also doesn't really matter. What does matter is that the list of fabs capable of making KnC chips is much longer that you've proposed. Because the Bitcoin miner is essentially free from intelectual property encumberances it is also free from many other manufacturing constraints typical in the industry.
Isn't bitfury's foundry UMC?
Could be, Europractice offers access to UMC, TSMC and some others. But the smallest features size avaiable through this channel are currently 40nm.
Oh ok, was wondering since ive seen bitfury say that. Wasnt sure of the tie in with europractice.
definitely, they have at least pictures of all components.
https://bitcointalksearch.org/topic/ann-usnorth-american-bitfury-sales-new-stock-now-shipping-251966
Uh... there's no case on that thing. It's just a bunch of chips and PCBs sitting on a cardboard box
Well fab will be in People's Rep. of China obv., same as all Bitcoin ASIC vendors, either at Global Foundries, or TSMC, assembly will be Sweden.
I'm pretty sure that TAIWAN Semiconductor Manufacturing is in Taiwan, A.K.A. the Republic of China, as opposed to the People's republic of China.
Taiwanese tend to be quite picky on that point.
Which only shows you haven't read up on anything. There is nothing to show. They aim to have chips in hand by September, that's all they've ever promised. See they could show you other components, but with the earlier similie I gave with respect to this being a formulae one race; McLaren aren't about to show Ferrari their upcoming seasons blueprints for their car, or for that matter reveal anything about it until the press release before the race starts. Fact is they have been spending pre-order money and are committed.
That's my reasoning, but you need to find your own.
Well, they said they'd ship the first two days worth of orders in September, and the rest so far in October. That means, obviously they'll need to not only have the chips, but also the PCBs, cases, and assembly factory all set up and ready to go in the next 60 days.
I don't really think showing cases/PCBs. would be that big of a deal. Remember, BFL has had their chips on hand for months, and they got strung up on the simple logistics of getting them all in cases and mailed out.
Same thing with Avalon - even when they had their chips, they still got hung up on the logistics of sending out just a few hundred units at a time.
It's not at all unreasonable to worry that a company that's never had to package and ship a consumer product might run into hiccups, even if they have their chips.
Isn't bitfury's foundry UMC?
Could be, Europractice offers access to UMC, TSMC and some others. But the smallest features size avaiable through this channel are currently 40nm.
Which makes sense as Bitfury is 55nm.
Isn't bitfury's foundry UMC?
Could be, Europractice offers access to UMC, TSMC and some others. But the smallest features size avaiable through this channel are currently 40nm.
$800k for a 28nm full-mask set? This is an amazing offer, at which foundry will I get it that cheap?
That I do not believe KnC is producing a fully designed asic should be obvious by now.
That only works if the foundry is convinced that you are worthy of credit. Normally this is not the case for start-ups and SMEs in this price category. For them it is cash in advance or no tape-out. Yes, I know what I'm talking about, it's hard to deal directly with companies like GF or TSMC if your are only a small customer.
I doubt that KnC or OrSOC is seen as a big customer for this project, because they probably will not order more than 50 wafers in the first year (which would be already about 3 PH). 1000 wafers a year would be a big customer. Why should a foundry take any risk for a bitcoin mining ASIC?
So I think they most likely have to pay the complete $1.5M to $2M in advance of the tape-out. And because they said tape-out done "some time ago", they probably paid it also some time ago.
From the commercial point of view I see no reason why KnC should not be able to produce a fully designed 28nm ASIC. They collected enough money by pre-orders to pay the complete 3rd party NREs in advance. But it should be clear, that if they are on track as they claimed, a great deal of the pre-order money isn't in their hands any more.
4. My guess, which people have posted that it is not true, is they are going with eAsic.
Yes, I'm not that far away, but do not agree with all of your points:
1. OrSOC and KnC need a channel partner of one of the two 28nm foundries to get access at all. eASIC would be a choice (also for a standard cell ASIC!), eSilicon or GUC are others. There are many and they love to take your money for that kind of projects. In this way there is no background check by the foundry required.
2. From my own experience I know that not all 28nm "giga" fabs running currently at there capacity limit. If you bring the money (as cash in advance
) through an accepted channel partner they will do a tape-out for you. They will for sure make you not the best price for masks and wafers, but they will do it, just to get workload for their fabs.3. Physical implementation and sign-off of a >100mm2 28nm die in about 6 weeks (because mid of June they didn't even have the foundry finally selected and they did tape-out "some time ago") is really a challenge, but not completely impossible. The related risk to fail for such a turbo implementation is huge. I hope they had a really experienced layout team for this job, which already did some tape-out in the concrete 28nm node.
eASIC is GF, Europe.
Altera (my guess) is TSMC, Taiwan, Asia.
ORSoC has a working relationship with both, both do standard cell and take the RTL code. I guess they just had to find the right balance between cost and delivery time. Since Asia won, that very much means they didn't choose eASIC. And they're not dealing with the foundry directly either (hoping they wouldn't take the risk of dealing with a non established partner which would really risk screwing up their timetable).
Also, don't forget they promised a September delivery, NOT a 1st of September delivery. That's 4 weeks to play with.
As far as I know, Altera offers FPGA-Hardcopy only. KnC said several times that they are doing definitely a standard cell ASIC design, no FPGA-Hardcopy.
Their released R&D slides are looking as a typical standard cell ASIC too.
So it's for sure not Altera and hopefully not eASIC, because they have no proven 28nm experience so far. I guess there is a new partner.
Their ambitious delivery dates I rather do not comment.
Isn't bitfury's foundry UMC?
Jump to: