[Announcement] Avalon ASIC Batch #1 Ships

[nathanrees19]

What happens if one of those unreleased features makes them an non-competitive lemon?

You're going out on a very long limb here if you think they have a hidden feature that can negate the dramatic difference in power costs.

[1714002675]

1714002675

[1714002675]

1714002675

[1714002675]

1714002675

[Syke]

Wow.. the self denial is incredible.  Yes, enjoy your 4U, 400w hashing monster, lol. Size a fucking refrigerator for 66 GH/s.

Profit per day at 66 GH/s @ 400w = $190
Profit per day at 60 GH/s @ 100w = $174

Hashing monster > vaporware product.

[niko]

What happens if one of those unreleased features makes them an non-competitive lemon?

You're going out on a very long limb here if you think they have a hidden feature that can negate the dramatic difference in power costs.

My speculation: They will be selling series of ever slightly more powerful chips, which you just plug into sockets. Easy upgrades.

[PuertoLibre]

BFL has the 40Gh/s system.

BFL takes them seriously, makes a promise to overclock their ASICs by 30%. Raising it to 60 Gh/s.

Last time I checked, overclocking from 40GH/s -> 60GH/s was a 50% overclock, not 30%. I could be wrong tho. Those damn maths are sneaky little bastards.

Your right. I was thinking of the 60Gh/s instead of the old 40Gh/s. I'll go back and change it.

[PuertoLibre]

I hope they keep their tricks a secret until BFL starts building the PCB's and commits hard money to completing their project.

I doubt at that point any BFL customer will get their cash back.

Josh can correct me if I am wrong about this and tell his customers that they can pull their cash out of this venture at any point (without limitations) prior to shipping.

[MooC Tals]

So I guess in 4 days we should see signs of Avalons asics runnin....

[sounds]

My speculation: the board supports dynamic frequency and voltage changes, a software upgrade makes overclocking and undervolting automatic (pick your target hash rate)

Since the chips internal structure is very repetitive, binning them into frequency ranges won't produce a large spread of chip speeds.

Aftermarket heatsink and fan arrangements could be very profitable for the first month or so.

[bbb8]

interesting

[jl2012]

I hope they keep their tricks a secret until BFL starts building the PCB's and commits hard money to completing their project.

I doubt at that point any BFL customer will get their cash back.

Josh can correct me if I am wrong about this and tell his customers that they can pull their cash out of this venture at any point (without limitations) prior to shipping.

Even they have legitimate reasons to keep the spec secret, there is absolutely no reason to keep the tracking number secret.

[YipYip]

What happens if one of those unreleased features makes them an non-competitive lemon?

You're going out on a very long limb here if you think they have a hidden feature that can negate the dramatic difference in power costs.

My speculation: They will be selling series of ever slightly more powerful chips, which you just plug into sockets. Easy upgrades.

+1

Agreed...having a simple upgrade path using the existing chasiss

Also some way of linking them together in a borg type deal..(thats a me dreaming ) ...lol

[bitmar]

I hope they keep their tricks a secret until BFL starts building the PCB's and commits hard money to completing their project.

I doubt at that point any BFL customer will get their cash back.

Josh can correct me if I am wrong about this and tell his customers that they can pull their cash out of this venture at any point (without limitations) prior to shipping.

Even they have legitimate reasons to keep the spec secret, there is absolutely no reason to keep the tracking number secret.

You are wrong
no ASIC = no shipment = no tracking number = reason to keep the tracking number (fictitious) secret

[ElectricMucus]

What happens if one of those unreleased features makes them an non-competitive lemon?

You're going out on a very long limb here if you think they have a hidden feature that can negate the dramatic difference in power costs.

My speculation: They will be selling series of ever slightly more powerful chips, which you just plug into sockets. Easy upgrades.

+1

Agreed...having a simple upgrade path using the existing chasiss

Also some way of linking them together in a borg type deal..(thats a me dreaming ) ...lol

I doubt it, if anything it won't be just sockets since common QFP sockets are freaking expensive. (I researched that option once for a home project)
Added to that even if they have a concept for upwards compatible pin layout the buffer caps would depend on the clock frequency.

So the best option would be modular pcbs, which could be possible but it would be more expensive for the upgrades themselves and it's exactly the most efficient strategy since the devices wouldn't consist of much else than a micro controller and a power supply.

[sounds]

Even they have legitimate reasons to keep the spec secret, there is absolutely no reason to keep the tracking number secret.

You are wrong
no ASIC = no shipment = no tracking number = reason to keep the tracking number (fictitious) secret

I think he was saying the same thing you are saying actually.

The tracking number doesn't exist yet - they're getting the paperwork in order. They're probably also tired of listening to us here, talk about it endlessly.

[greyhawk]

If I were a company sending out the first prototypes of my highly anticipated completely new device the last thing I would want is for all of the world to see where exactly this prototype is floating around, especially considering the communities infiltration with scammers and thieves.

[crazy_rabbit]

I hope they keep their tricks a secret until BFL starts building the PCB's and commits hard money to completing their project.

I doubt at that point any BFL customer will get their cash back.

Josh can correct me if I am wrong about this and tell his customers that they can pull their cash out of this venture at any point (without limitations) prior to shipping.

Even they have legitimate reasons to keep the spec secret, there is absolutely no reason to keep the tracking number secret.

Doesn't having the tracking number give you access to the recipients address? Perhaps it's for privacy.

[nathanrees19]

Doesn't having the tracking number give you access to the recipients address? Perhaps it's for privacy.

Usually not, but most posters would be quite happy with the recipients of the initial 2 units saying "I've got a tracking number and it's currently at X".

[michaelmclees]

Wow.. the self denial is incredible.  Yes, enjoy your 4U, 400w hashing monster, lol. Size a fucking refrigerator for 66 GH/s.

Size matters... at all?  There are only 4 relevant factors: cost, release date, hashing power, and wattage.  That you're even commenting on size, and that you've consistently touted the good looks of the BFL units, tells everyone here that your priorities are way out of whack.

[Korbman]

Wow.. the self denial is incredible.  Yes, enjoy your 4U, 400w hashing monster, lol. Size a fucking refrigerator for 66 GH/s.

Profit per day at 66 GH/s @ 400w = $190
Profit per day at 60 GH/s @ 100w = $174

Hashing monster > vaporware product.

You're talking initial gains. Double or triple the difficulty in your equation and try again..

*Hint* Organofcorti already did all the math for you over a wide range of difficulties over on his blog */Hint*

[MrTeal]

Wow.. the self denial is incredible.  Yes, enjoy your 4U, 400w hashing monster, lol. Size a fucking refrigerator for 66 GH/s.

Profit per day at 66 GH/s @ 400w = $190
Profit per day at 60 GH/s @ 100w = $174

Hashing monster > vaporware product.

You're talking initial gains. Double or triple the difficulty in your equation and try again..

*Hint* Organofcorti already did all the math for you over a wide range of difficulties over on his blog */Hint*

It actually isn't that bad for the first little while. Even at 250TH/s (~35M difficulty), the systems are pretty similar.
For instance, at $17.50/BTC and $0.15/kWh, an Avalon hashing at 66GH/s using 400W would profit $455/30 day month.
At the same conditions, the BFL Single at 60GH/s using 60W would profit $447/30 day month.

Baring a change in stats, it probably won't be until almost all the preorders are out that BFL takes over from Avalon in terms of profitability. Avalon's second batch is still a loser in my eyes if BFL (and Avalon) can ship on time though. Earnings might be close to the same, but the Avalon unit is going to heat up a room and use a lot more space, in addition to costing $200 more. They'll definitely need to rethink their price increase or have some super duper extra features.

I hope they keep their tricks a secret until BFL starts building the PCB's and commits hard money to completing their project.

LOLWUT? You think that building the PCBs is the point in which you put hard money into the project? The cost of all the PCBs for the first batch going to be a lot less than they spend on salaries for a month, it's a drop in the bucket at this point.

[Soros Shorts]

Wow.. the self denial is incredible.  Yes, enjoy your 4U, 400w hashing monster, lol. Size a fucking refrigerator for 66 GH/s.

Profit per day at 66 GH/s @ 400w = $190
Profit per day at 60 GH/s @ 100w = $174

Hashing monster > vaporware product.

You're talking initial gains. Double or triple the difficulty in your equation and try again..

*Hint* Organofcorti already did all the math for you over a wide range of difficulties over on his blog */Hint*

Where I live, a 300W advantage works out to about $2.00 (0.11 BTC) a day less in electricity costs. For most people in the US it would be less.

I would have stopped mining long before my $1,000+ piece of hardware is pulling in only $2/day. It is simply not worth it.