More and more coins moving away from scrypt algos
The smart thing to do would be to release affordable asics before people get fed up and deal with the problem with an algo change
Topic: [ANN] Sfards:SF100, the first 28nm Dual-Mode Miner is accepting pre-order now - page 8. (Read 129863 times)
September 27, 2015, 08:31:27 AM
September 16, 2015, 04:49:50 PM
It seems that you need to plan to sell your miner in the next 3-6 months, and hope that it's value hasn't declined too much. If for some reason a newer/different miner arrives, then that plan blows up big time. And also assume BTC price doesn't decline during those 3-6 months.
You really have to stretch your assumptions to make things work these days, particularly with new gear, expensive, gear.
You really have to stretch your assumptions to make things work these days, particularly with new gear, expensive, gear.
September 16, 2015, 12:08:08 PM
It's possible to RoI a S7 - if you get one early enough and you have cheap enough electric. Might even be possible in the US, if the diff rate increases flatten out sometime late this year back down from the recent 5% ballpark to 2-3%ish AND if you live in one of 3 counties in Washington State (2 if the diff rate does NOT flatten out this year).
While it may be possible, in theory, I see it as incredibly unlikely to ROI before next halving, at least for "normal" miners like me, if you have a peta-hash farm it is different.
At 4c per kWh if difficulty grows more than 2% per retarget you won't; at 6c you need to be around 1% diff. increase per jump or you won't.
A few more 5% jumps and you're out of luck.
spiccioli
September 15, 2015, 07:24:29 PM
It's possible to RoI a S7 - if you get one early enough and you have cheap enough electric. Might even be possible in the US, if the diff rate increases flatten out sometime late this year back down from the recent 5% ballpark to 2-3%ish AND if you live in one of 3 counties in Washington State (2 if the diff rate does NOT flatten out this year).
September 15, 2015, 01:02:15 PM
I think they are only selling big order to large private farms right now they have the technology edge for the moment and they are exploiting it
Their chips are having huge yield issues, I doubt they have sold any apart from the couple they made from their sample chips. The whole "batch 1 sold out" is a marketing gimmick.
The 16/14nm chip process node itself is the problem for EVERYONE going to those sizes regardless of what the chips are used for. Intel, nVidia, Samsung et al ALL have the same poor yield issues because the manufacturing process is still far from stable.
This simple fact is why Bitmain took the safer route to just optimize their 28nm chips vs rolling the dice on the smaller nodes. Give another year or so and yields for 16/14nm *should* get better. Of course researchers/devs at the chip foundries have been saying that for a couple years now...
I would expect bitmain is looking at R/D on lower nm chips. They have lots of money and people. They choose to release another 28nm which was the most profitable for them at the moment. They likely will do it until a competitor forces them to move to a lower NM.
But I think they have the best of both worlds. Sell 28nm and work on r/d on lower nm for later on.... win win for them.
Exactly, and all those buying S7s (which will never ROI) are footing the bill
spiccioli
September 15, 2015, 12:22:59 PM
I think they are only selling big order to large private farms right now they have the technology edge for the moment and they are exploiting it
Their chips are having huge yield issues, I doubt they have sold any apart from the couple they made from their sample chips. The whole "batch 1 sold out" is a marketing gimmick.
The 16/14nm chip process node itself is the problem for EVERYONE going to those sizes regardless of what the chips are used for. Intel, nVidia, Samsung et al ALL have the same poor yield issues because the manufacturing process is still far from stable.
This simple fact is why Bitmain took the safer route to just optimize their 28nm chips vs rolling the dice on the smaller nodes. Give another year or so and yields for 16/14nm *should* get better. Of course researchers/devs at the chip foundries have been saying that for a couple years now...
got thread confused with the LK groups one... https://bitcointalksearch.org/topic/m.12324684 you be right - these are touted as 28nm process, questions then would be is it 2 28nm process dies per-chip or one big one.On the correct note... 28nm node is stable and easy enough to be good for even some commodity-level quantities. If there is chip problems then highly doubt is is in the mfg end unless they contracted with a company in India (2 20nm fabs, better at 28nm, opened there in the last couple years) or something like that...
September 15, 2015, 12:02:35 PM
I think they are only selling big order to large private farms right now they have the technology edge for the moment and they are exploiting it
Their chips are having huge yield issues, I doubt they have sold any apart from the couple they made from their sample chips. The whole "batch 1 sold out" is a marketing gimmick.
The 16/14nm chip process node itself is the problem for EVERYONE going to those sizes regardless of what the chips are used for. Intel, nVidia, Samsung et al ALL have the same poor yield issues because the manufacturing process is still far from stable.
This simple fact is why Bitmain took the safer route to just optimize their 28nm chips vs rolling the dice on the smaller nodes. Give another year or so and yields for 16/14nm *should* get better. Of course researchers/devs at the chip foundries have been saying that for a couple years now...
I would expect bitmain is looking at R/D on lower nm chips. They have lots of money and people. They choose to release another 28nm which was the most profitable for them at the moment. They likely will do it until a competitor forces them to move to a lower NM.
But I think they have the best of both worlds. Sell 28nm and work on r/d on lower nm for later on.... win win for them.
September 15, 2015, 11:34:50 AM
I think they are only selling big order to large private farms right now they have the technology edge for the moment and they are exploiting it
Their chips are having huge yield issues, I doubt they have sold any apart from the couple they made from their sample chips. The whole "batch 1 sold out" is a marketing gimmick.
The 16/14nm chip process node itself is the problem for EVERYONE going to those sizes regardless of what the chips are used for. Intel, nVidia, Samsung et al ALL have the same poor yield issues because the manufacturing process is still far from stable.
This simple fact is why Bitmain took the safer route to just optimize their 28nm chips vs rolling the dice on the smaller nodes. Give another year or so and yields for 16/14nm *should* get better. Of course researchers/devs at the chip foundries have been saying that for a couple years now...
I would expect bitmain is looking at R/D on lower nm chips. They have lots of money and people. They choose to release another 28nm which was the most profitable for them at the moment. They likely will do it until a competitor forces them to move to a lower NM.
But I think they have the best of both worlds. Sell 28nm and work on r/d on lower nm for later on.... win win for them.
September 15, 2015, 11:00:35 AM
I think they are only selling big order to large private farms right now they have the technology edge for the moment and they are exploiting it
Their chips are having huge yield issues, I doubt they have sold any apart from the couple they made from their sample chips. The whole "batch 1 sold out" is a marketing gimmick.
The 16/14nm chip process node itself is the problem for EVERYONE going to those sizes regardless of what the chips are used for. Intel, nVidia, Samsung et al ALL have the same poor yield issues because the manufacturing process is still far from stable.
This simple fact is why Bitmain took the safer route to just optimize their 28nm chips vs rolling the dice on the smaller nodes. Give another year or so and yields for 16/14nm *should* get better. Of course researchers/devs at the chip foundries have been saying that for a couple years now...
September 15, 2015, 07:27:39 AM
Maybe the S7 from Bitmain has forced them back to the drawing board???
The S7 is quite a step forward in anyones books... Even if they have sorted how to mine both Sha & script at the same time...
The S7 is quite a step forward in anyones books... Even if they have sorted how to mine both Sha & script at the same time...
They have a special market with the dual miner aspect. They are competing against old zeus/hashra/etc, and A2 for the most part. So it's not new gear they are competing with but old gear.
S7 might bring down the worth of the SHA mining side. But scrypt the S7 does not compete with this miner on.
September 15, 2015, 07:23:41 AM
Maybe the S7 from Bitmain has forced them back to the drawing board???
The S7 is quite a step forward in anyones books... Even if they have sorted how to mine both Sha & script at the same time...
The S7 is quite a step forward in anyones books... Even if they have sorted how to mine both Sha & script at the same time...
September 15, 2015, 06:16:24 AM
I heard from reliable source, that somebody will start producing new dual miner with Sfards chips. It should be available by end of October. But I have not more technical info about it
I heard this too..but I think they mean the current SFARDS units....I thought that was when THEY shipped
I don't know it still would not make sense to me. If they have "Their chips are having huge yield issues" and don't have enough for their own miners, why would they sell chips to others to make miners before them?
I just think they would use them on SFards miner first before selling enough chips that someone makes miners with.
September 15, 2015, 06:11:59 AM
Nearly 3 months since sfards posted any info on their product. 3 months of nothing...... 
73 pages of speculation & guesswork while they mine with their units on their "could" site.
This is a non-product thread.
73 pages of speculation & guesswork while they mine with their units on their "could" site.
This is a non-product thread.
September 15, 2015, 04:33:45 AM
I heard from reliable source, that somebody will start producing new dual miner with Sfards chips. It should be available by end of October. But I have not more technical info about it
I heard this too..but I think they mean the current SFARDS units....I thought that was when THEY shipped
September 15, 2015, 04:16:05 AM
I heard from reliable source, that somebody will start producing new dual miner with Sfards chips. It should be available by end of October. But I have not more technical info about it
September 15, 2015, 03:35:08 AM
Looks like the SF100 second batch will be mid-August - at the SAME insanely high 8000 yuan price.
I suspect the next batch doing "pre-sale mining, errr testing" might explain the recent Litecoin hash increases and perhaps a small amount of the recent Bitcoin hashrate bump.
I suspect the next batch doing "pre-sale mining, errr testing" might explain the recent Litecoin hash increases and perhaps a small amount of the recent Bitcoin hashrate bump.
I'm confused...mid-august? 8000 rmb is terrible...more like SF-TARDS if you ask me!
What month is this? Lol.
I dunno anymore! It was trippin me out reading this thread when I woke up on the sofa hung over..thought I had slipped into a worm hole and woken up 2 months into the past
I think they are only selling big order to large private farms right now they have the technology edge for the moment and they are exploiting it
Their chips are having huge yield issues, I doubt they have sold any apart from the couple they made from their sample chips. The whole "batch 1 sold out" is a marketing gimmick.
Where did you get this information? This could explain a lot if true, but they have not indicated this.
We have seen what 20 maybe or so go to resellers. I still think decent chance of a big buyer bought first batch.
It definitely would...also why are they waiting so long before the mass productions? Are there really any big buyers left in the scrypt market? Surely they would make more money selling them off in small quantities to resellers. What scrypt coin is worth mining now since ltc has gone tits up?
September 14, 2015, 03:41:27 PM
I think they are only selling big order to large private farms right now they have the technology edge for the moment and they are exploiting it
Their chips are having huge yield issues, I doubt they have sold any apart from the couple they made from their sample chips. The whole "batch 1 sold out" is a marketing gimmick.
Where did you get this information? This could explain a lot if true, but they have not indicated this.
We have seen what 20 maybe or so go to resellers. I still think decent chance of a big buyer bought first batch.
September 14, 2015, 11:17:10 AM
I think they are only selling big order to large private farms right now they have the technology edge for the moment and they are exploiting it
Their chips are having huge yield issues, I doubt they have sold any apart from the couple they made from their sample chips. The whole "batch 1 sold out" is a marketing gimmick.
September 11, 2015, 09:59:47 PM
the average life span of a mining company seems to be about 12 months...whatever happened to gridseed?
GridSeed + WiiBox + Diginforce = SFARDS
don't forget to add Zeus at the end, Zeus and grindseed got together then it slowly faded away if i find that I'll link it, it all happen about the time or at the time Zeus did the ripe off thing .
Here we go
https://zeusminer.com/forums/topic/yahoo-finnance-zeushash-announces-gridseed-partnership-thanksgiving-promotion/
I still say Zeus had some thing to do with stopping L1's which left a bad taste in bitmains month and one reason we don't get much more from anyone, they really did a number .
sadly like some one said why even bro posting it here unless it is for the free advertisement they get ?
the price is beyond what most here are willing to pay for etc, they know it .
what would be a better place then here ?.
it's free all most any one or every one into bitcoins is here at some point in there day week month etc that includes the Big data centers and knc etc .
or does Sfards aka grindseed pay to post here ? no idea !!! .
September 11, 2015, 09:45:06 PM
I think they are only selling big order to large private farms right now they have the technology edge for the moment and they are exploiting it
Jump to: