|
rudi
|
 |
November 28, 2014, 09:44:29 AM |
|
Update
Some details of BE300:
Process: tsmc 28nm hpc
Package: fclga (5mm x 5mm)
Normal Mode:
0.7v vdd
6gh/s per chip
8gh/s-12gh/s per chip for mass production
0.343w/g on chip
~0.3w/g on chip for mass production
Low Power Mode:
0.55v vdd
4.5gh/s per chip
6gh/s-9gh/s per chip for mass production
0.225w/g on chip
~0.2w/g on chip for mass production
The schedule of BE300 producing: First batch production will be done next Feb.
So these chips are smaller than the BE200, and also have fewer GH/s compared to the BE200, but are expected to have much better power efficiency. Does this open up the possibility of putting these chips on USB sticks again? It seems that even two chips could be built into a usb stick and still have low enough power consumption. On the flip slide, I expect this also means that the BE300 chips are not compatible with the BE200 chips with respect to pinouts and package, as friedcat stated would be likely, back in june? Finally, I'm also confused regarding the difference in GH/s per chip and power consumption before and during mass production. |
|
|
|
|
|
|
|
|
|
|
|
|
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
|
|
|
|
|
|
|
|
Caesium
|
|
November 28, 2014, 10:24:52 AM |
|
|
|
|
|
Dexter770221
Legendary
 Offline
Activity: 1029
Merit: 1000
|
|
November 28, 2014, 11:15:09 AM |
|
6gh/s per chip
8gh/s-12gh/s per chip for mass production
0.343w/g on chip
~0.3w/g on chip for mass production
I don't understand the numbers. So it's 6GH/s per chip at 0.343W/GH, but for mass production the chips get better speed and lower power consumption. How will they do this? New mask? I'm quessing 4(2?) dies on one substrate (bigger package) lowered voltage and clock and... voila. |
Under development Modular UPGRADEABLE Miner (MUM). Looking for investors.
Changing one PCB with screwdriver and you have brand new miner in hand... Plug&Play, scalable from one module to thousands.
|
|
|
|
arnuschky
|
|
November 28, 2014, 12:09:51 PM |
|
That Laxo Trade spammer is sending 1000 satoshis all the time, all over the place, and none of their tx failed. Some fountains send even lesser amounts.
I don't think that threshold thing happened here - more likely not all shares are connected to the right addresses yet. That's why this div was - to clear out the shareholding status. So, check your address, and if the div is not there, clear that out with FC.
Laxo is sending 10000 Satoshis as far as I know. What you remember is maybe from before the introduction of the dust threshold in 0.8.2? Faucets and fountains were mostly killed by this, see for example http://bitcoin.lift-institute.com/killing-the-dust/0.00001000 btc, september 2014, long after v0.8.2 many outputs like this, put in a single tx - same way AM sends out divs. Interesting. I have no idea how they are able to do that - they shouldn't be. Thanks for this, I will investigate. Maybe they use their own miner? @all others: sorry for going OT so much, will stop now.  |
|
|
|
|
|
arnuschky
|
|
November 28, 2014, 12:28:23 PM |
|
Package: fclga (5mm x 5mm)
Quite interesting. If I am not mistaken, FCLGA is flip-chip land grid array. What this means is basically a package as we know from Intel and AMD processors: FC means that die is soldered and then glued on a mini-PCB (called "substrate"). LGA means that this mini-PCB has pins that go into a socket on the main electronics board. As said by Dexter770221, that it's flip-chip might explain the difference in testing vs. mass production: they simply stick multiple dies on one substrate. (This is, by the way, what Hashfast has been doing, but please don't tell Icebreaker  ) Fun fact about LGA: it gives you the opportunity to make the substrates swappable, similar to what we know from Intel and AMD. A reaction to their problems with chips burning? A way for future upgrade which allows you to keep the base miner and just buy a new generation of chips from AM every few months? Who knows.  Intel Core I3 FC-LGA with two dies on a substrate. This thing is of course swappable, but might be also soldered onto the board (ie, in notebooks). Some refs: http://en.wikipedia.org/wiki/Flip_chiphttp://en.wikipedia.org/wiki/Land_grid_array |
|
|
|
|
nycgoat
Member
Offline
Activity: 117
Merit: 10
|
|
November 28, 2014, 01:50:28 PM |
|
So the companies name is "Basic Miner" now? Haha, jk Keep up the good work guys! |
|
|
|
|
mrlupin
Newbie
Offline
Activity: 30
Merit: 0
|
|
November 28, 2014, 04:22:04 PM |
|
I'm glad to see FC hasn't completely forgotten about shareholders. Does anyone know if he still reads/accepts PM requests here to change dividend payment addresses? I have two addresses I'd like to consolidate into one.
I did a transfer from havelock to direct and it was acked within two weeks by email by friedcat |
|
|
|
|
hdbuck
Legendary
Offline
Activity: 1260
Merit: 1002
|
|
November 28, 2014, 06:06:55 PM |
|
Update
Some details of BE300:
Process: tsmc 28nm hpc
Package: fclga (5mm x 5mm)
Normal Mode:
0.7v vdd
6gh/s per chip
8gh/s-12gh/s per chip for mass production
0.343w/g on chip
~0.3w/g on chip for mass production
Low Power Mode:
0.55v vdd
4.5gh/s per chip
6gh/s-9gh/s per chip for mass production
0.225w/g on chip
~0.2w/g on chip for mass production
The schedule of BE300 producing: First batch production will be done next Feb.
So these chips are smaller than the BE200, and also have fewer GH/s compared to the BE200, but are expected to have much better power efficiency. Does this open up the possibility of putting these chips on USB sticks again? It seems that even two chips could be built into a usb stick and still have low enough power consumption. On the flip slide, I expect this also means that the BE300 chips are not compatible with the BE200 chips with respect to pinouts and package, as friedcat stated would be likely, back in june? Finally, I'm also confused regarding the difference in GH/s per chip and power consumption before and during mass production. I would love to see usb miners again... Those things sold like hotcakes... meh, does the bitmain U3s sold good? could it even compete with them? https://bitcointalk.org/index.php?topic=827356.0 |
|
|
|
|
|
stompysteve
|
|
November 28, 2014, 06:34:08 PM |
|
6gh per chip? why so low
|
|
|
|
|
|
jimmothy
|
|
November 28, 2014, 07:15:12 PM |
|
Those require a power supply. A usb miner would definitely be more appealing for beginners/hobbyists. |
|
|
|
|
|
armedmilitia
|
|
November 28, 2014, 07:29:56 PM |
|
What email did you use to contact friedcat? I tried a PM via bitcointalk but I got no response. |
|
|
|
|
Franktank
|
|
November 28, 2014, 08:34:33 PM |
|
Update
Some details of BE300:
Process: tsmc 28nm hpc
Package: fclga (5mm x 5mm)
Normal Mode:
0.7v vdd
6gh/s per chip
8gh/s-12gh/s per chip for mass production
0.343w/g on chip
~0.3w/g on chip for mass production
Low Power Mode:
0.55v vdd
4.5gh/s per chip
6gh/s-9gh/s per chip for mass production
0.225w/g on chip
~0.2w/g on chip for mass production
The schedule of BE300 producing: First batch production will be done next Feb.
Are there folks that happen to know more about this elaborate on this further? Other than just being a little more efficient than BE200, it doesn't seem like that great of an improvement for a die shrink (40nm -> 28nm). Please excuse my ignorance, I'm more of a biochemist and not familiar with the specifics of chip R&D. |
|
|
|
|
|
Caesium
|
|
November 28, 2014, 09:43:21 PM |
|
What email did you use to contact friedcat? I tried a PM via bitcointalk but I got no response. - I didn't get a reply, still haven't even now, but apparently it does get read. |
|
|
|
|
jimmothy
|
|
November 28, 2014, 09:46:18 PM |
|
Update
Some details of BE300:
Process: tsmc 28nm hpc
Package: fclga (5mm x 5mm)
Normal Mode:
0.7v vdd
6gh/s per chip
8gh/s-12gh/s per chip for mass production
0.343w/g on chip
~0.3w/g on chip for mass production
Low Power Mode:
0.55v vdd
4.5gh/s per chip
6gh/s-9gh/s per chip for mass production
0.225w/g on chip
~0.2w/g on chip for mass production
The schedule of BE300 producing: First batch production will be done next Feb.
Are there folks that happen to know more about this elaborate on this further? Other than just being a little more efficient than BE200, it doesn't seem like that great of an improvement for a die shrink (40nm -> 28nm). Please excuse my ignorance, I'm more of a biochemist and not familiar with the specifics of chip R&D. 0.2-0.343 w/gh is not "a little more efficient". |
|
|
|
|
hdbuck
Legendary
Offline
Activity: 1260
Merit: 1002
|
|
November 28, 2014, 11:18:46 PM |
|
Those require a power supply. A usb miner would definitely be more appealing for beginners/hobbyists. well then, im not so sure its a good idea anyway. we are a little over beginners and usbminers. |
|
|
|
|
mrlupin
Newbie
Offline
Activity: 30
Merit: 0
|
|
November 29, 2014, 03:34:06 AM |
|
What email did you use to contact friedcat? I tried a PM via bitcointalk but I got no response. - I didn't get a reply, still haven't even now, but apparently it does get read. Yes that's the email address that notified me of share transfers too. |
|
|
|
|
|
mithrandi
|
|
November 29, 2014, 08:19:51 AM |
|
Those require a power supply. A usb miner would definitely be more appealing for beginners/hobbyists.
Frankly, beginners/hobbyists are just wasting their time if they're messing around with miners at this point. Of course, a market may exist for such a novelty item nonetheless, but... |
|
|
|
|
Rival
|
|
November 29, 2014, 01:08:20 PM |
|
Those require a power supply. A usb miner would definitely be more appealing for beginners/hobbyists.
Frankly, beginners/hobbyists are just wasting their time if they're messing around with miners at this point. Of course, a market may exist for such a novelty item nonetheless, but... The fact that thousands of 333 Mhs USB miners still change hands on ebay each month should indicate that there is a large demand for usb miners. |
|
|
|
|
|
jdany
|
|
November 29, 2014, 01:26:19 PM |
|
Those require a power supply. A usb miner would definitely be more appealing for beginners/hobbyists.
Frankly, beginners/hobbyists are just wasting their time if they're messing around with miners at this point. Of course, a market may exist for such a novelty item nonetheless, but... The fact that thousands of 333 Mhs USB miners still change hands on ebay each month should indicate that there is a large demand for usb miners. I think someone said 30 billion USB slots will go unused, with nothing plugged into them, in 2015. That's almost 4 USB slots for every man, woman, and child in the world. Don't let this happen. Think of the children. |
|
|
|
|
|
drasted
|
|
November 29, 2014, 02:02:08 PM |
|
Those require a power supply. A usb miner would definitely be more appealing for beginners/hobbyists.
Frankly, beginners/hobbyists are just wasting their time if they're messing around with miners at this point. Of course, a market may exist for such a novelty item nonetheless, but... Same things were said when the asicminers old USB miners were being sold. And it was probably true. Didn't stop them from selling like crazy. |
|
|
|
|
|
