Announcement - ASIC mining processor by Butterfly Labs - page 17.

kano

legendary

Activity: 4592

Merit: 1851

Linux since 1997 RedHat 4

Quote from: smoothie on February 07, 2013, 11:35:45 PM

Quote from: Korbman on February 07, 2013, 09:35:26 AM

Quote from: kano on February 07, 2013, 07:51:23 AM

FYI

I've got my plane tickets to BFL - Kansas City today (email booked digital ticket details from the travel agent) arranged by Josh
So my flight to the USA, leaving Sydney on the 17th Feb (my time) - to be back here in Aus again on the 23rd.

... so I going from Oz (Aus) to Kansas (City) ... isn't that back to front? Cheesy

That's quite the trip! How many days do they have you touring the facilities? Anything specific on the agenda?

Agenda: To get their stories straight so when he reports back on the forums it is exactly how BFL wants him to report things lol.

There's a reason why your ignore is so dark ...

You are so far off the mark and you've made a comment based on zero knowledge of the situation ... or is your statement simply to mean that you can easily be bought and just assume others are like you?

Here's most of the opening and closing of the recent PM I sent Josh to find out where they were at and then got the visit finally organised.

Code:

Feb-2

Hey Josh,

you got any update on where your guys are at?

I'm certainly not someone to promote your guys unsubstantiated claims coz
of your crap history (even with the ASIC), and you know full well I've never been
a fan of BFL, but the Avalon people are really pissing me off, Tom's gone down
the drain, and I would certainly be happy to be able to run a BFL ASIC on an rpi
and show it leaving Avalon's shit in the dust.

...
...

So ... when is anything happening at BFL?

P.S. if anything I've said on the forum lately has pissed your guys off - oh well,
damn shame :)

Askit2

hero member

Activity: 981

Merit: 500

DIV - Your "Virtual Life" Secured and Decentralize

Quote from: Gomeler on February 08, 2013, 01:14:56 AM

Quote from: Askit2 on February 07, 2013, 03:48:30 PM

Quote from: Gomeler on February 07, 2013, 02:28:30 PM

Quote from: makomk on February 07, 2013, 01:30:52 PM

Quote from: Gomeler on February 05, 2013, 10:57:21 PM

Avalon went with a very simple chip on a large(110nm) process. I imagine this strategy gave them very high yields out of the wafers because of the small chip size. This came at the cost of a large PCB to route to all the chips.

Well, that's certainly going to be part of the reason. However, take a look at the distance between chips on the BFL board and compare it to the distance between chips on the Avalon board. Now remember Josh's comment about the PCB ground plane not being able to dissipate enough heat due to the thermal density of their design, and how having to redesign the chips to fix this delayed shipping whilst Avalon shipped on schedue using similar QFN chips.

Excellent point. I have minimal knowledge of using the the ground pad and vias to transmit heat through the PCB but it makes logical sense that Avalon would use a larger than electrically necessary PCB to give each chip some buffer from neighboring chips.

Wouldn't this require also the chips be physically farther apart to avoid heat soaking the inner ones? Plus wouldn't moving them farther apart increase the delay while communicating?

I guess a person could just make a huge board and run copper out to the edges but the thermal transfer isn't as effective as a nearer solution. The size of the piece of whatever metal would dictate the maximum amount of heat that would flow at a given thermal difference. It seems like going with a flip chip (ala P3) rather then older QFN would save materials and allow a more efficient board design. By efficient I mean all data would transfer faster going over a shorter connection. Yes short trips nearly speed of light etc but small changes in location on other devices have caused >10% increases in speed.

I think the more compact design would do better if it can be cooled. Thus far I see no reason it can't be cooled.

The communication between the chips and the controller is minimal. I am also of the opinion that the more compact design would be better but for compute density reasons. BFL excites me for the possibility of watercooling their chips and ending up with an insanely dense mining cluster. I hope in a few generations BTC ASICs are something that we could order like FPGAs now and assemble custom PCBs to suit our own desires.

I agree I guess my point was at each speed increase the unit either needs to roll work (more complicated design per chip) or get new work more often making the latency a larger issue as even from on board ram takes time (not much). I don't mean that the latency is huge per se. Just that say a current single has a latency of about .01 second I think it was with the proper bus speed (mostly USB delay and fixed in asics). now the Jalapeno will hash 5x as fast and if the latency stayed the same instead of .01 per 5.1 seconds it would be 0.19% but .01 of about 1 second becomes 1%. The percentage is small enough at low rates. As speed increases latency becomes a speed limiter. Again current to new isn't apples to apples. Current gets work, works, finsihes, submits, gets new work. New will they said cache next work so get work, work, get more work, finish work, work, get more work. This will remove a ton of latency.

I like the sound of what BFL said they changed. I like the higher density board. I like the heatsink getting less thermal resistance to the core. All these things will make for a better end product as long as they are accurate and work as expected. Not that I am saying a FCBGA isn't a viable chip, just that caching and FC will help a lot.

Gomeler

hero member

Activity: 697

Merit: 500

Quote from: Askit2 on February 07, 2013, 03:48:30 PM

Quote from: Gomeler on February 07, 2013, 02:28:30 PM

Quote from: makomk on February 07, 2013, 01:30:52 PM

Quote from: Gomeler on February 05, 2013, 10:57:21 PM

Avalon went with a very simple chip on a large(110nm) process. I imagine this strategy gave them very high yields out of the wafers because of the small chip size. This came at the cost of a large PCB to route to all the chips.

Well, that's certainly going to be part of the reason. However, take a look at the distance between chips on the BFL board and compare it to the distance between chips on the Avalon board. Now remember Josh's comment about the PCB ground plane not being able to dissipate enough heat due to the thermal density of their design, and how having to redesign the chips to fix this delayed shipping whilst Avalon shipped on schedue using similar QFN chips.

Excellent point. I have minimal knowledge of using the the ground pad and vias to transmit heat through the PCB but it makes logical sense that Avalon would use a larger than electrically necessary PCB to give each chip some buffer from neighboring chips.

Wouldn't this require also the chips be physically farther apart to avoid heat soaking the inner ones? Plus wouldn't moving them farther apart increase the delay while communicating?

I guess a person could just make a huge board and run copper out to the edges but the thermal transfer isn't as effective as a nearer solution. The size of the piece of whatever metal would dictate the maximum amount of heat that would flow at a given thermal difference. It seems like going with a flip chip (ala P3) rather then older QFN would save materials and allow a more efficient board design. By efficient I mean all data would transfer faster going over a shorter connection. Yes short trips nearly speed of light etc but small changes in location on other devices have caused >10% increases in speed.

I think the more compact design would do better if it can be cooled. Thus far I see no reason it can't be cooled.

The communication between the chips and the controller is minimal. I am also of the opinion that the more compact design would be better but for compute density reasons. BFL excites me for the possibility of watercooling their chips and ending up with an insanely dense mining cluster. I hope in a few generations BTC ASICs are something that we could order like FPGAs now and assemble custom PCBs to suit our own desires.

smoothie

legendary

Activity: 2492

Merit: 1473

LEALANA Bitcoin Grim Reaper

Quote from: Korbman on February 07, 2013, 09:35:26 AM

Quote from: kano on February 07, 2013, 07:51:23 AM

FYI

I've got my plane tickets to BFL - Kansas City today (email booked digital ticket details from the travel agent) arranged by Josh
So my flight to the USA, leaving Sydney on the 17th Feb (my time) - to be back here in Aus again on the 23rd.

... so I going from Oz (Aus) to Kansas (City) ... isn't that back to front? Cheesy

That's quite the trip! How many days do they have you touring the facilities? Anything specific on the agenda?

Agenda: To get their stories straight so when he reports back on the forums it is exactly how BFL wants him to report things lol.

PuertoLibre

legendary

Activity: 1890

Merit: 1003

Quote from: nbtcminer on February 07, 2013, 04:43:57 PM

Quote from: bitcoindaddy on February 07, 2013, 04:23:55 PM

Quote from: kano on February 07, 2013, 07:51:23 AM

FYI

I've got my plane tickets to BFL - Kansas City today (email booked digital ticket details from the travel agent) arranged by Josh
So my flight to the USA, leaving Sydney on the 17th Feb (my time) - to be back here in Aus again on the 23rd.

... so I going from Oz (Aus) to Kansas (City) ... isn't that back to front? Cheesy

Whoo-hoo! Kano is coming to America! Everyone get your daughters locked up safe!

https://www.youtube.com/watch?v=hMtZfW2z9dw

Play it!

Phinnaeus Gage

legendary

Activity: 1918

Merit: 1570

Bitcoin: An Idea Worth Spending

Quote from: bitcoindaddy on February 07, 2013, 04:23:55 PM

Quote from: kano on February 07, 2013, 07:51:23 AM

FYI

I've got my plane tickets to BFL - Kansas City today (email booked digital ticket details from the travel agent) arranged by Josh
So my flight to the USA, leaving Sydney on the 17th Feb (my time) - to be back here in Aus again on the 23rd.

... so I going from Oz (Aus) to Kansas (City) ... isn't that back to front? Cheesy

Whoo-hoo! Kano is coming to America! Everyone get your daughters locked up safe!

Guess it before clicking: https://www.youtube.com/watch?v=T3S7mlRYL-8

hotdealdave

jr. member

Activity: 61

Merit: 1

Crikey!

nbtcminer

sr. member

Activity: 249

Merit: 250

Quote from: bitcoindaddy on February 07, 2013, 04:23:55 PM

Quote from: kano on February 07, 2013, 07:51:23 AM

FYI

I've got my plane tickets to BFL - Kansas City today (email booked digital ticket details from the travel agent) arranged by Josh
So my flight to the USA, leaving Sydney on the 17th Feb (my time) - to be back here in Aus again on the 23rd.

... so I going from Oz (Aus) to Kansas (City) ... isn't that back to front? Cheesy

Whoo-hoo! Kano is coming to America! Everyone get your daughters locked up safe!

bitcoindaddy

hero member

Activity: 481

Merit: 500

Quote from: kano on February 07, 2013, 07:51:23 AM

FYI

I've got my plane tickets to BFL - Kansas City today (email booked digital ticket details from the travel agent) arranged by Josh
So my flight to the USA, leaving Sydney on the 17th Feb (my time) - to be back here in Aus again on the 23rd.

... so I going from Oz (Aus) to Kansas (City) ... isn't that back to front? Cheesy

Whoo-hoo! Kano is coming to America! Everyone get your daughters locked up safe!

Askit2

hero member

Activity: 981

Merit: 500

DIV - Your "Virtual Life" Secured and Decentralize

Quote from: Gomeler on February 07, 2013, 02:28:30 PM

Quote from: makomk on February 07, 2013, 01:30:52 PM

Quote from: Gomeler on February 05, 2013, 10:57:21 PM

Avalon went with a very simple chip on a large(110nm) process. I imagine this strategy gave them very high yields out of the wafers because of the small chip size. This came at the cost of a large PCB to route to all the chips.

Well, that's certainly going to be part of the reason. However, take a look at the distance between chips on the BFL board and compare it to the distance between chips on the Avalon board. Now remember Josh's comment about the PCB ground plane not being able to dissipate enough heat due to the thermal density of their design, and how having to redesign the chips to fix this delayed shipping whilst Avalon shipped on schedue using similar QFN chips.

Excellent point. I have minimal knowledge of using the the ground pad and vias to transmit heat through the PCB but it makes logical sense that Avalon would use a larger than electrically necessary PCB to give each chip some buffer from neighboring chips.

Wouldn't this require also the chips be physically farther apart to avoid heat soaking the inner ones? Plus wouldn't moving them farther apart increase the delay while communicating?

I guess a person could just make a huge board and run copper out to the edges but the thermal transfer isn't as effective as a nearer solution. The size of the piece of whatever metal would dictate the maximum amount of heat that would flow at a given thermal difference. It seems like going with a flip chip (ala P3) rather then older QFN would save materials and allow a more efficient board design. By efficient I mean all data would transfer faster going over a shorter connection. Yes short trips nearly speed of light etc but small changes in location on other devices have caused >10% increases in speed.

I think the more compact design would do better if it can be cooled. Thus far I see no reason it can't be cooled.

Gomeler

hero member

Activity: 697

Merit: 500

Quote from: makomk on February 07, 2013, 01:30:52 PM

Quote from: Gomeler on February 05, 2013, 10:57:21 PM

Avalon went with a very simple chip on a large(110nm) process. I imagine this strategy gave them very high yields out of the wafers because of the small chip size. This came at the cost of a large PCB to route to all the chips.

Well, that's certainly going to be part of the reason. However, take a look at the distance between chips on the BFL board and compare it to the distance between chips on the Avalon board. Now remember Josh's comment about the PCB ground plane not being able to dissipate enough heat due to the thermal density of their design, and how having to redesign the chips to fix this delayed shipping whilst Avalon shipped on schedue using similar QFN chips.

Excellent point. I have minimal knowledge of using the the ground pad and vias to transmit heat through the PCB but it makes logical sense that Avalon would use a larger than electrically necessary PCB to give each chip some buffer from neighboring chips.

makomk

hero member

Activity: 686

Merit: 564

Quote from: Gomeler on February 05, 2013, 10:57:21 PM

Avalon went with a very simple chip on a large(110nm) process. I imagine this strategy gave them very high yields out of the wafers because of the small chip size. This came at the cost of a large PCB to route to all the chips.

Well, that's certainly going to be part of the reason. However, take a look at the distance between chips on the BFL board and compare it to the distance between chips on the Avalon board. Now remember Josh's comment about the PCB ground plane not being able to dissipate enough heat due to the thermal density of their design, and how having to redesign the chips to fix this delayed shipping whilst Avalon shipped on schedue using similar QFN chips.

Fiyasko

legendary

Activity: 1428

Merit: 1001

Okey Dokey Lokey

Quote from: John (John K.) on February 05, 2013, 09:34:32 PM

Thread has been split yet again.

https://bitcointalksearch.org/topic/bfl-announcement-trolling-museum-2-141496

Guys come on! Stay on topic!, Stop talking about the wizard of Oz and Kansas and how Newegg sells shoes!
Not only is this one of the most heavily moderated threads i can think of, But its already been "cleaned and split" before, And you've made him do it again!
And by the looks of it, He may soon have to do it a Third time!

Also, Very happy to see new photos

Korbman

legendary

Activity: 1064

Merit: 1001

Quote from: kano on February 07, 2013, 07:51:23 AM

FYI

I've got my plane tickets to BFL - Kansas City today (email booked digital ticket details from the travel agent) arranged by Josh
So my flight to the USA, leaving Sydney on the 17th Feb (my time) - to be back here in Aus again on the 23rd.

... so I going from Oz (Aus) to Kansas (City) ... isn't that back to front? Cheesy

That's quite the trip! How many days do they have you touring the facilities? Anything specific on the agenda?

Soros Shorts

donator

Activity: 1617

Merit: 1012

Quote from: beekeeper on February 07, 2013, 09:21:07 AM

Quote from: kano on February 07, 2013, 07:51:23 AM

... so I going from Oz (Aus) to Kansas (City) ... isn't that back to front? Cheesy

LOL, true... what you need plane tickets, just clicking your heels three times..

Shoes are avaialble on Newegg http://www.newegg.com/Product/Product.aspx?Item=019-0004-00Y07

beekeeper

sr. member

Activity: 406

Merit: 250

LTC

Quote from: kano on February 07, 2013, 07:51:23 AM

... so I going from Oz (Aus) to Kansas (City) ... isn't that back to front? Cheesy

LOL, true... what you need plane tickets, just clicking your heels three times..

ewibit

legendary

Activity: 2955

Merit: 1049

Quote from: kano on February 07, 2013, 07:51:23 AM

FYI

I've got my plane tickets to BFL - Kansas City today (email booked digital ticket details from the travel agent) arranged by Josh
So my flight to the USA, leaving Sydney on the 17th Feb (my time) - to be back here in Aus again on the 23rd.

nice

hope you make some pictures...

JWU42

legendary

Activity: 1666

Merit: 1000

Quote from: kano on February 07, 2013, 07:51:23 AM

FYI

I've got my plane tickets to BFL - Kansas City today (email booked digital ticket details from the travel agent) arranged by Josh
So my flight to the USA, leaving Sydney on the 17th Feb (my time) - to be back here in Aus again on the 23rd.

... so I going from Oz (Aus) to Kansas (City) ... isn't that back to front? Cheesy

kano

legendary

Activity: 4592

Merit: 1851

Linux since 1997 RedHat 4

FYI

I've got my plane tickets to BFL - Kansas City today (email booked digital ticket details from the travel agent) arranged by Josh
So my flight to the USA, leaving Sydney on the 17th Feb (my time) - to be back here in Aus again on the 23rd.

... so I going from Oz (Aus) to Kansas (City) ... isn't that back to front? Cheesy

Tomatocage

legendary

Activity: 1554

Merit: 1222

brb keeping up with the Kardashians

Quote from: Gomeler on February 05, 2013, 10:57:21 PM

Avalon went with a very simple chip on a large(110nm) process. I imagine this strategy gave them very high yields out of the wafers because of the small chip size. This came at the cost of a large PCB to route to all the chips. BFL appears to be going the other way with a few very large chips based on a smaller(65nm) process. This should give them better power efficiency and more hashing power per chip at the risk of random manufacturing defects giving them fewer of these chips per wafer. Of course one hopes for 100% yields but that isn't always the case. We'll see how this all plays out very soon.

Ah, thanks for the explanation. I think I'll opt for the BFL unit then.

Topic: Announcement - ASIC mining processor by Butterfly Labs - page 17. (Read 227559 times)