bitcoinarnold
Member
Offline
Activity: 63
Merit: 10
|
|
July 12, 2013, 05:57:21 PM |
|
bpduke knows what he's talking about, butthurt kncminer fanboys who blindly gave away their money beware.
|
|
|
|
bkpduke
|
|
July 12, 2013, 06:04:27 PM |
|
bpduke knows what he's talking about, butthurt kncminer fanboys who blindly gave away their money beware.
I want to be clear that I WANT KNC to come to market. The more manufacturers the merrier and the better for all of us. Just when we dig into the nuts and bolts, there are some things that really bug me. 1) Chip design appears to be huge - my main concern. 2) Chip has never been shown (ok, it's in manuf, but they have to have prototypes by now - you don't order up 1 mil chips unless you have validated the hardware) 3) Chip details like number of SHA256 "cores" has never been released. All we have are some demos with FGPA hardware, nothing here. Until I see more, my money goes behind chips we have actually seen: Avalon and Bitfury. BFL's business practices are so bad I'm marking them off my (personal) list.
|
|
|
|
|
titomane
|
|
July 12, 2013, 06:36:23 PM |
|
BKPDUKE 28nm die size, wrapped in a 55mm package, with a 2048 ball count for thermal characteristics. Wtf is so hard to understand? If you don't believe them, then don't buy.
Do you even realize what you said? You just said you are going to have the equivalent of a FOOTBALL FIELD to mount a CPU the size of a lunchbox. WTF is so hard to understand is 28nm is the LINE WIDTH not the AREA. Please don't comment when it is clear you have @#$% for brains. Sorry but your before you too have denied yourself. https://bitcointalk.org/index.php?topic=170332.msg2714375#msg2714375You say you never use a big package for a small chip. Instead here says that the intel chips have a die (80 & 130mm2) https://bitcointalk.org/index.php?topic=170332.msg2713678#msg2713678When we know that the package of these chips is much higher Pentium 3 socket 370 Pentium 4 socket 478 I have not found the exact measurements but are between 50x50mm and 60x60mm This package is a 2500-3600mm2. This is package 21 times biggest than Die By this mistake I will not say " Please don't comment when it is clear you have @#$% for brains", a lapsus what has anyone.
|
|
|
|
jimmy3dita
|
|
July 12, 2013, 07:23:03 PM |
|
More FUD please, so the hashrate keeps low
|
Acquista il mio libro "Investire Bitcoin": clicca qui
|
|
|
bkpduke
|
|
July 12, 2013, 09:01:52 PM |
|
BKPDUKE 28nm die size, wrapped in a 55mm package, with a 2048 ball count for thermal characteristics. Wtf is so hard to understand? If you don't believe them, then don't buy.
Do you even realize what you said? You just said you are going to have the equivalent of a FOOTBALL FIELD to mount a CPU the size of a lunchbox. WTF is so hard to understand is 28nm is the LINE WIDTH not the AREA. Please don't comment when it is clear you have @#$% for brains. Sorry but your before you too have denied yourself. https://bitcointalk.org/index.php?topic=170332.msg2714375#msg2714375You say you never use a big package for a small chip. Instead here says that the intel chips have a die (80 & 130mm2) https://bitcointalk.org/index.php?topic=170332.msg2713678#msg2713678When we know that the package of these chips is much higher Pentium 3 socket 370 Pentium 4 socket 478 I have not found the exact measurements but are between 50x50mm and 60x60mm This package is a 2500-3600mm2. This is package 21 times biggest than Die By this mistake I will not say " Please don't comment when it is clear you have @#$% for brains", a lapsus what has anyone. Those a PGA SOCKETED chips. KNC is using BGA UNSOCKETED chips. PGA vs. BGA is not so big as the Socketed part. You create a socket so that you give people an upgrade path or choice of different CPUs to plug into a universal board that can be used among different CPUs. THIS IS AN EXTRA EXPENSE, so you only do it if you want to cover the market in different tiers of performance (and price) for your product. KNC is soldering their chips directly to the mainboard - NO SOCKET. Why does this matter? If you don't use a Socket, then your "package" does not need to be nearly as big. Plus P3/P4 is over a decade old. You don't don't make packages that big anymore. It costs more money over a smaller, compact package. The fanboys here keep missing the point: EVEN IF 75% of this package is just wasted space, that leaves a die size of 750mm2. That would make this one of, if not the largest, mass-produced dies ever. If true, that's just asking for trouble. Would love for KNC to post some screens of the wafers or dies and more technical info. IMO you only stay this quiet when you are up a creek without a paddle.
|
|
|
|
blazin8s
|
|
July 12, 2013, 09:28:57 PM Last edit: July 12, 2013, 09:49:21 PM by blazin8s |
|
Has anyone made profitability calculations based on projected difficulty by the time these come out? These look like a fantastic deal now, but will they be in September, October, November. If they are released by then that is. I am also curious about this quoted from their website. Why are you selling the Miners and don’t mine yourself (will you use the miners for mining yourself)?
We are mining ourselves, but we believe in the bitcoin project and a diversified market is the best for all parties. We will continue to mine but we do not believe in a monopoly of miners. They say here that they are currently mining, yet they only show 3d sketches of their product which leads to the assumption that it is not even produced yet. I do get why people that make ASICs would want to spread them around, especially since one can only cash out so many bitcoins per day/month, but it only makes sense that anyone creating these would want to mine for at least a couple months before releasing them, and if they are not even made yet is it really safe to believe these will be released this fall as they say? And what are they mining with now, FPGAs, XFX cards? Or maybe, they are mining with them right now (this being the couple months prior) and not releasing a photo because then they would have to come up with a reason for why they are not shipping yet.
|
|
|
|
erk
|
|
July 12, 2013, 09:43:49 PM |
|
Plus P3/P4 is over a decade old. You don't don't make packages that big anymore. It costs more money over a smaller, compact package.
Still more nonsense, have you looked at the Intel LGA 2011 CPU specifications? They are bigger than the P3? Please make an effort to check your claims before you post.
|
|
|
|
notme
Legendary
Offline
Activity: 1904
Merit: 1002
|
|
July 12, 2013, 09:52:52 PM |
|
Hmm.... let's see if kncminer has said anything lately.... nope, just people with too much time on their hands arguing about off topic shit.
|
|
|
|
nightengale
|
|
July 12, 2013, 09:56:42 PM |
|
Hmm.... let's see if kncminer has said anything lately.... nope, just people with too much time on their hands arguing about off topic shit.
+1
|
|
|
|
RHA
|
|
July 12, 2013, 10:03:16 PM Last edit: July 12, 2013, 10:27:42 PM by RHA |
|
Bkpduke and erk, view this short video from open day at KnC: https://www.youtube.com/watch?v=by-je8XRCdYThey (KnC) stated quite clearly, the chip (the package) would be around the size shown (about 60 mm x 60 mm) not the die, which will be much smaller. (That worries me, because while the yield will be higher, the heat density will be quite higher too.)
|
|
|
|
titomane
|
|
July 12, 2013, 10:29:12 PM Last edit: July 12, 2013, 10:47:37 PM by titomane |
|
BKPDUKE 28nm die size, wrapped in a 55mm package, with a 2048 ball count for thermal characteristics. Wtf is so hard to understand? If you don't believe them, then don't buy.
Do you even realize what you said? You just said you are going to have the equivalent of a FOOTBALL FIELD to mount a CPU the size of a lunchbox. WTF is so hard to understand is 28nm is the LINE WIDTH not the AREA. Please don't comment when it is clear you have @#$% for brains. Sorry but your before you too have denied yourself. https://bitcointalk.org/index.php?topic=170332.msg2714375#msg2714375You say you never use a big package for a small chip. Instead here says that the intel chips have a die (80 & 130mm2) https://bitcointalk.org/index.php?topic=170332.msg2713678#msg2713678When we know that the package of these chips is much higher Pentium 3 socket 370 Pentium 4 socket 478 I have not found the exact measurements but are between 50x50mm and 60x60mm This package is a 2500-3600mm2. This is package 21 times biggest than Die By this mistake I will not say " Please don't comment when it is clear you have @#$% for brains", a lapsus what has anyone. Those a PGA SOCKETED chips. KNC is using BGA UNSOCKETED chips. PGA vs. BGA is not so big as the Socketed part. You create a socket so that you give people an upgrade path or choice of different CPUs to plug into a universal board that can be used among different CPUs. THIS IS AN EXTRA EXPENSE, so you only do it if you want to cover the market in different tiers of performance (and price) for your product. KNC is soldering their chips directly to the mainboard - NO SOCKET. Why does this matter? If you don't use a Socket, then your "package" does not need to be nearly as big. Plus P3/P4 is over a decade old. You don't don't make packages that big anymore. It costs more money over a smaller, compact package. The fanboys here keep missing the point: EVEN IF 75% of this package is just wasted space, that leaves a die size of 750mm2. That would make this one of, if not the largest, mass-produced dies ever. If true, that's just asking for trouble. Would love for KNC to post some screens of the wafers or dies and more technical info. IMO you only stay this quiet when you are up a creek without a paddle. XBOX-WII & PS3 CPU's are BGA soldered. Their DIE size are less than 15 & 17% of package size. XBOX http://pictures.xbox-scene.com/xbox360/cpu/CPU.jpgPS3 http://techon.nikkeibp.co.jp/english/NEWS_EN/20061127/124495/?SS=imgview_e&FD=-2047582866&ad_qWII http://megagames.com/news/nintendo-rd-director-defends-wii-u-cpu-speed
|
|
|
|
ASIC-K
Sr. Member
Offline
Activity: 280
Merit: 250
Hell?
|
|
July 12, 2013, 10:34:08 PM |
|
please please please can we stop talking about die and package size. lets wait until we see some more specs or a picture first....
ANYWAY
im really hoping they are sandbagging us and making the units now. that would be a huge surprise if they delivered early.
|
|
|
|
erk
|
|
July 12, 2013, 10:46:22 PM |
|
please please please can we stop talking about die and package size. lets wait until we see some more specs or a picture first....
ANYWAY
im really hoping they are sandbagging us and making the units now. that would be a huge surprise if they delivered early.
It takes two months to make a wafer, no sandbagging involved, and that's on top of how long it takes you to raise the funds first.
|
|
|
|
ASIC-K
Sr. Member
Offline
Activity: 280
Merit: 250
Hell?
|
|
July 12, 2013, 10:47:32 PM |
|
well i mean in general. im hoping they are staying tight lipped because they are on schedule or slightly ahead with the whole production.
|
|
|
|
Phoenix1969
Legendary
Offline
Activity: 938
Merit: 1000
LIR DEV
|
|
July 12, 2013, 11:24:50 PM |
|
bpduke knows what he's talking about, butthurt kncminer fanboys who blindly gave away their money beware.
I want to be clear that I WANT KNC to come to market. The more manufacturers the merrier and the better for all of us. Just when we dig into the nuts and bolts, there are some things that really bug me. 1) Chip design appears to be huge - my main concern. 2) Chip has never been shown (ok, it's in manuf, but they have to have prototypes by now - you don't order up 1 mil chips unless you have validated the hardware) 3) Chip details like number of SHA256 "cores" has never been released. All we have are some demos with FGPA hardware, nothing here. Until I see more, my money goes behind chips we have actually seen: Avalon and Bitfury. BFL's business practices are so bad I'm marking them off my (personal) list. 28nm die with a 55mm wrapper is what we were told by Marcus Erlanderson himself at the openday event. I have no reason to doubt his 10+ years of experience in ASICs. You barking about chip dimensions is quite irrelevant as well. The only thing that matters is how they perform. That's good news for me anyway, less in line before my next purchase.
|
|
|
|
bkpduke
|
|
July 12, 2013, 11:42:33 PM |
|
That's a 30-35% die vs. package size right there (the RAISED board is the "package" - the rectangle on it is the DIE).
|
|
|
|
bkpduke
|
|
July 12, 2013, 11:45:38 PM |
|
Anyway, by the time these guys get this to market the ROI is going to be non-existant based upon current estimates for difficulty.
Oct delivery would be when difficulty is in the 150-250 million range. ANYTHING at all that delays these puppies, and you can forget any kind of return on your investment.
|
|
|
|
androz
|
|
July 13, 2013, 12:11:25 AM |
|
Oct delivery would be when difficulty is in the 150-250 million range. ANYTHING at all that delays these puppies, and you can forget any kind of return on your investment.
that is really a scary point if for some reason the shipping will be delayed (x example 1 month) how deep could be the impact?
|
|
|
|
erk
|
|
July 13, 2013, 12:16:27 AM |
|
Anyway, by the time these guys get this to market the ROI is going to be non-existant based upon current estimates for difficulty.
Oct delivery would be when difficulty is in the 150-250 million range. ANYTHING at all that delays these puppies, and you can forget any kind of return on your investment.
BS you are just trying to spread FUD in almost everyone of your posts in this thread, for starters the ASIC vendors have been rolling out far slower than they said they would be, so the difficulty projections are skewed, secondly a large volume of GPU miners may leave BTC, influenced by the BTC price which you can't predict.
|
|
|
|
|