Bitcoin Forum
November 11, 2024, 06:19:58 AM *
News: Latest Bitcoin Core release: 28.0 [Torrent]
 
   Home   Help Search Login Register More  
Warning: One or more bitcointalk.org users have reported that they strongly believe that the creator of this topic is a scammer. (Login to see the detailed trust ratings.) While the bitcointalk.org administration does not verify such claims, you should proceed with extreme caution.
Pages: « 1 2 3 4 5 6 [7] 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 ... 91 »
  Print  
Author Topic: [Announcement] Avalon ASIC Batch #1 Ships  (Read 238921 times)
aTg
Legendary
*
Offline Offline

Activity: 1358
Merit: 1000



View Profile
January 20, 2013, 08:42:05 PM
 #121

Quote from: Specifications
Number of Pads: 48
    8 Data
    40+1 Power

40 of the pads are GND and 1 is VCC only have 8 tracks of Data
MrTeal
Legendary
*
Offline Offline

Activity: 1274
Merit: 1004


View Profile
January 20, 2013, 08:49:00 PM
 #122

Quote from: Specifications
Number of Pads: 48
    8 Data
    40+1 Power

40 of the pads are GND and 1 is VCC only have 8 tracks of Data
What? That makes no sense.
There's 48 pins and the one thermal slug under the chip. Of the 48 pins, 40 are power (GND and Vcc), and the +1 refers to the thermal pad which is also ground.
aTg
Legendary
*
Offline Offline

Activity: 1358
Merit: 1000



View Profile
January 20, 2013, 08:54:08 PM
 #123

Quote from: Specifications
Number of Pads: 48
    8 Data
    40+1 Power

40 of the pads are GND and 1 is VCC only have 8 tracks of Data
What? That makes no sense.
There's 48 pins and the one thermal slug under the chip. Of the 48 pins, 40 are power (GND and Vcc), and the +1 refers to the thermal pad which is also ground.

Good, that really is not important, the story is that there are only 8 pads data, the remaining 40 are split between GND and VCC.

The important question is always lost in many posts, how many chips in every PCB?
Frizz23
Hero Member
*****
Offline Offline

Activity: 1162
Merit: 500


View Profile
January 20, 2013, 09:12:10 PM
 #124

How many units did you actually ship?

There are lots of posts on BFL forums where users write something like: "The shipping thread names two recipients. Therefore there are only two units shipping, QED."

Ξtherization⚡️First P2E 2016⚡️🏰💎🌈 etherization.org
smoothie
Legendary
*
Offline Offline

Activity: 2492
Merit: 1474


LEALANA Bitcoin Grim Reaper


View Profile
January 20, 2013, 09:19:36 PM
 #125

I guess BFL needs to start preordering through AVALON! LOL!!!!

███████████████████████████████████████

            ,╓p@@███████@╗╖,           
        ,p████████████████████N,       
      d█████████████████████████b     
    d██████████████████████████████æ   
  ,████²█████████████████████████████, 
 ,█████  ╙████████████████████╨  █████y
 ██████    `████████████████`    ██████
║██████       Ñ███████████`      ███████
███████         ╩██████Ñ         ███████
███████    ▐▄     ²██╩     a▌    ███████
╢██████    ▐▓█▄          ▄█▓▌    ███████
 ██████    ▐▓▓▓▓▌,     ▄█▓▓▓▌    ██████─
           ▐▓▓▓▓▓▓█,,▄▓▓▓▓▓▓▌          
           ▐▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▌          
    ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓─  
     ²▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓╩    
        ▀▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▀       
           ²▀▀▓▓▓▓▓▓▓▓▓▓▓▓▀▀`          
                   ²²²                 
███████████████████████████████████████

. ★☆ WWW.LEALANA.COM        My PGP fingerprint is A764D833.                  History of Monero development Visualization ★☆ .
LEALANA BITCOIN GRIM REAPER SILVER COINS.
 
klondike_bar
Legendary
*
Offline Offline

Activity: 2128
Merit: 1005

ASIC Wannabe


View Profile
January 20, 2013, 09:26:53 PM
 #126

Quote from: Specifications
Number of Pads: 48
    8 Data
    40+1 Power

40 of the pads are GND and 1 is VCC only have 8 tracks of Data
What? That makes no sense.
There's 48 pins and the one thermal slug under the chip. Of the 48 pins, 40 are power (GND and Vcc), and the +1 refers to the thermal pad which is also ground.

Good, that really is not important, the story is that there are only 8 pads data, the remaining 40 are split between GND and VCC.

my 2cents would be that its better form-factor or for cooling if the chip can be fed power from multiple lines where it is needed, rather then piping in through 2 lines and being forked within the chip to where power is consumed.

(Im not a hardware professional/manufacturer, thats just my logic for explaining the number of connections)

my other 2 conspiracy theories are:
1) more contacts simply help keep the chip sealed to the board better to prevent seperation in the case of thermal expansion/contraction or other physical abuse
2) some power connects are redundant in case errors or heat problems cause some pins to burn out or short. (though i dont think this is a very likely practise for mitigation of such an issue)

24" PCI-E cables with 16AWG wires and stripped ends - great for server PSU mods, best prices https://bitcointalk.org/index.php?topic=563461
No longer a wannabe - now an ASIC owner!
BR0KK
Hero Member
*****
Offline Offline

Activity: 784
Merit: 500



View Profile
January 20, 2013, 09:28:35 PM
 #127

How many units did you actually ship?

There are lots of posts on BFL forums where users write something like: "The shipping thread names two recipients. Therefore there are only two units shipping, QED."

Better than the leading Manufacturer in ASIC design Smiley

repentance
Hero Member
*****
Offline Offline

Activity: 868
Merit: 1000


View Profile
January 20, 2013, 09:31:07 PM
Last edit: January 20, 2013, 09:53:06 PM by repentance
 #128

How many units did you actually ship?

There are lots of posts on BFL forums where users write something like: "The shipping thread names two recipients. Therefore there are only two units shipping, QED."

Meh, I don't think it matters all that much if they did only ship two units last night.  The more relevant question for me would be about continuity of shipping and whether they are going to be keeping a steady supply of units going out the door daily now until all 300 units have shipped.  It would be annoying if there's now a break of a week before any more units ship, for instance, even though they've explicitly stated since October that they would ship over a two week period.

I guess it's also worth remembering that with some people ordering multiple units, the number of units shipped isn't going to match the number of orders fulfilled.  One order could be 5 or 10 units.

I think people overlook the fact that there are pros and cons for the way each vendor has planned launching their product and that initial shipping is only the beginning, not the end game.  I also expect that every single ASIC vendor which does deliver (and I believe that BFL will) will be doing a lot of post-delivery tweaking and that tweaking is what's going to determine who has the best product at the end of the day.



All I can say is that this is Bitcoin. I don't believe it until I see six confirmations.
jgarzik
Legendary
*
Offline Offline

Activity: 1596
Merit: 1100


View Profile
January 20, 2013, 09:32:27 PM
 #129

There are lots of posts on BFL forums where users write something like: "The shipping thread names two recipients. Therefore there are only two units shipping, QED."

Although I'm obviously biased...  BFL really should do that, too.  Most major manufacturers ship out demo and review units of hardware before the mass production units are shipped to vendors/stores/customers.

Even if BFL only shipped out two units, it would prove they have working hardware and start everyone salivating Smiley


Jeff Garzik, Bloq CEO, former bitcoin core dev team; opinions are my own.
Visit bloq.com / metronome.io
Donations / tip jar: 1BrufViLKnSWtuWGkryPsKsxonV2NQ7Tcj
klondike_bar
Legendary
*
Offline Offline

Activity: 2128
Merit: 1005

ASIC Wannabe


View Profile
January 20, 2013, 09:35:53 PM
 #130

another tip for the shipping speed debate: The faster they ship out 300 pre-paid units, the sooner they can start taking payments for more units.

it would seem silly for them to twiddle thier thumbs when the speed of delivering these units will determine how soon they can make more money/customers. (on the other side of the spectrum, BFL is still taking pre-orders)

ps: PLEASE dont turn my comment into a trollish BFL debate, this is an avalon thread for avalon devices and thats why im reading it.

24" PCI-E cables with 16AWG wires and stripped ends - great for server PSU mods, best prices https://bitcointalk.org/index.php?topic=563461
No longer a wannabe - now an ASIC owner!
Krakonos
Member
**
Offline Offline

Activity: 60
Merit: 10


View Profile
January 20, 2013, 09:37:41 PM
 #131

Hi,

I believe that connecting such many pins to ground is just "good engineering practice". If you need one power pin, one ground pin and 8 data pins, what do you do with the rest? I usually ground them if it is possible (if they are for example IO pins set for input). You never want to have floating pins for many reasons (in this case mechanical, but may be they'll act as an antenna and will introduce some noise into the circuit if not connected somewhere).

(Correct me if I'm wrong, I'm just an amateur engineer :-))

Tip jar: 1MWj8Etpt3ayLG5AvXwhtEU42szJD2m97z
PuertoLibre
Legendary
*
Offline Offline

Activity: 1890
Merit: 1003


View Profile
January 20, 2013, 09:39:42 PM
 #132

I'll post photos and a review when my unit arrives.

I was surprised that I was the only core developer who paid for a unit.  Followed the shotgun approach:  ordered units from all the people working on ASICs.


Good for you!

Most developers seem to be waiting for freebies simply because they are coders.

In most industries people pay for expensive Developer Kits (including a computer system).
MrTeal
Legendary
*
Offline Offline

Activity: 1274
Merit: 1004


View Profile
January 20, 2013, 09:54:00 PM
 #133

How many units did you actually ship?

There are lots of posts on BFL forums where users write something like: "The shipping thread names two recipients. Therefore there are only two units shipping, QED."

Meh, I don't think it matters all that much if they did only ship two units last night.  The more relevant question for me would be about continuity of shipping and whether they are going to be keeping a steady supply of units going out the door daily now until all 300 units have shipped.  It would be annoying if there's now a break of a week before any more units ship, for instance, even though they've explicitly stated since October that they would ship over a two week period.

I guess it's also worth remembering that with some people ordering multiple units, the number of units shipped isn't going to match the number of orders fulfilled.  One order could be 5 or 10 units.

I think people overlook the fact that there are pros and cons for the way each vendor has planned launching their product and that initial shipping is only the beginning, not the end game.

+1
Shipping two in the first day really doesn't mean much. It's only if they can only ship a few every day that they'll be in serious trouble.
repentance
Hero Member
*****
Offline Offline

Activity: 868
Merit: 1000


View Profile
January 20, 2013, 10:03:58 PM
 #134

another tip for the shipping speed debate: The faster they ship out 300 pre-paid units, the sooner they can start taking payments for more units.

it would seem silly for them to twiddle thier thumbs when the speed of delivering these units will determine how soon they can make more money/customers. (on the other side of the spectrum, BFL is still taking pre-orders)

ps: PLEASE dont turn my comment into a trollish BFL debate, this is an avalon thread for avalon devices and thats why im reading it.

Not really.  Their production capacity is affected by CNY.  If they shipped all 300 units today, they're not really going to be able to deliver a second batch sooner than if they wait a couple of weeks.

There's also the issue of how the first 300 perform in the wild.  If they start the production process for a second batch now, they don't have information on how the first batch has performed and risk having to scrap that run and start again if there are major issues which need addressing.  

All I can say is that this is Bitcoin. I don't believe it until I see six confirmations.
Bogart
Legendary
*
Offline Offline

Activity: 966
Merit: 1000


View Profile
January 20, 2013, 10:04:56 PM
 #135

Quote from: Specifications
Number of Pads: 48
    8 Data
    40+1 Power

40 of the pads are GND and 1 is VCC only have 8 tracks of Data
What? That makes no sense.
There's 48 pins and the one thermal slug under the chip. Of the 48 pins, 40 are power (GND and Vcc), and the +1 refers to the thermal pad which is also ground.

Good, that really is not important, the story is that there are only 8 pads data, the remaining 40 are split between GND and VCC.

my 2cents would be that its better form-factor or for cooling if the chip can be fed power from multiple lines where it is needed, rather then piping in through 2 lines and being forked within the chip to where power is consumed.

(Im not a hardware professional/manufacturer, thats just my logic for explaining the number of connections)

my other 2 conspiracy theories are:
1) more contacts simply help keep the chip sealed to the board better to prevent seperation in the case of thermal expansion/contraction or other physical abuse
2) some power connects are redundant in case errors or heat problems cause some pins to burn out or short. (though i dont think this is a very likely practise for mitigation of such an issue)

Suppose the underside of the board contains the ground plane, which is connected to the top layer with a bunch of vias to the solder pads that connect to the thermal pads on the bottoms of the QFN chips.  Then the top side of the board has the Vcc plane, which connects to the chips on all the pins that are not otherwise used for data.

That way they have the metal in place to support moving a lot of amps through the chips.  If they wanted to get more extreme they could even apply a DC bias to the data pins and use them for power too.

"All safe deposit boxes in banks or financial institutions have been sealed... and may only be opened in the presence of an agent of the I.R.S." - President F.D. Roosevelt, 1933
420
Hero Member
*****
Offline Offline

Activity: 756
Merit: 500



View Profile
January 20, 2013, 10:08:37 PM
 #136

just a theory. maybe everyone's waiting till after chinese new year.

it will help all of bitcoin if all these companies remain in business

Donations: 1JVhKjUKSjBd7fPXQJsBs5P3Yphk38AqPr - TIPS
the hacks, the hacks, secure your bits!
beekeeper
Sr. Member
****
Offline Offline

Activity: 406
Merit: 250


LTC


View Profile WWW
January 20, 2013, 10:10:39 PM
 #137

400W and only termal pads? That means a lot of chips on board, I guess.. Did they say how much MHs per chip they got?

25Khs at 5W Litecoin USB dongle (FPGA), 45kHs overclocked
https://bitcointalk.org/index.php?topic=310926
Litecoin FPGA shop -> http://ltcgear.com
PuertoLibre
Legendary
*
Offline Offline

Activity: 1890
Merit: 1003


View Profile
January 20, 2013, 10:12:56 PM
 #138

I wonder how much testing has gone into this.  Surely quite a bit.  But wouldn't it be a great upset if Avalon was first to ship and all of the units failed after a week or two?  
A prediction for BFL in March 2013?
abeaulieu
Sr. Member
****
Offline Offline

Activity: 295
Merit: 250



View Profile
January 20, 2013, 10:13:20 PM
 #139

Where did these pictures of their ASICs come from?
Bogart
Legendary
*
Offline Offline

Activity: 966
Merit: 1000


View Profile
January 20, 2013, 10:18:20 PM
 #140

400W and only termal pads?

Perhaps I used the wrong term.  I meant the large solder point in the center of the QFN packages.


"All safe deposit boxes in banks or financial institutions have been sealed... and may only be opened in the presence of an agent of the I.R.S." - President F.D. Roosevelt, 1933
Pages: « 1 2 3 4 5 6 [7] 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 ... 91 »
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!