btcgarden-AM-v1 down to $0.53/GHs. In stock for international selling!

[jimmothy]

Next gen ST = close to 70x the density though (and also 50-70th of the $/GH)

How did you get 70x?

Think of the footprint and hash rate of a fully populated M-H board setup. Compare that with a SP30. Mucho density.

Have you seen the bf3500? Apparently can hold up to 2880 bitfury chips (11th/s) in a 6u case.

[dogie]

Next gen ST = close to 70x the density though (and also 50-70th of the $/GH)

How did you get 70x?

Think of the footprint and hash rate of a fully populated M-H board setup. Compare that with a SP30. Mucho density.

Have you seen the bf3500? Apparently can hold up to 2880 bitfury chips (11th/s) in a 6u case.

Yeah its insane, will have a review on it next week. Density is only about 3-4x higher than that and 6 weeks later.

[zulover]

Next gen ST = close to 70x the density though (and also 50-70th of the $/GH)

How did you get 70x?

Think of the footprint and hash rate of a fully populated M-H board setup. Compare that with a SP30. Mucho density.

Have you seen the bf3500? Apparently can hold up to 2880 bitfury chips (11th/s) in a 6u case.

where can i get one?

[grn]

Bitmain is not the price leader nor is spondoolies the efficiency leader.

Bitmain s1 at 200gh is $1.3/gh where as hashratio (using AM chips) is as low as $1/gh for twice as efficient miners. Btcgarden is about $1.55/gh with psu.

Also bitfury gen1 is as low as 0.38w/gh chip level which is about equal to spondoolies next gen estimates.

Hmmmmm  i always thought simple math was easy

ant s1 .425 - coupon .05 = .375

.375 x $529.48 = $198.56

$198.56 / 200ghs = $ 0.9928/gh

[kendog77]

Psst...

You know what I see?

Cheaper miners finally. This is going to be a proper price war this time.

Agreed, it's hard to complain about having more competitively priced options.

Products based on ASICMiner Gen3 have been a bit disappointing, though. They are competitive with other offerings currently on the market, but they are also the last to arrive at the 1W/GH dance, so their window of opportunity to profit looks pretty small (2-3 months). Much more efficient miners will be on the market in August.

Most of the hype about ASICMiner Gen3 seems to be coming from their shareholders that are simply trying to pump the share price. Miners don't seem to be too excited about them.

I see nothing in this product that would make we want to buy it over a S2 at $1.65 / GH.

[kendog77]

BTCGarden,

Here are some honest suggestions on how to improve the next version of your product:
1.) Try to get the power usage under 1W/GH (the lower the better).
2.) Include it in a case that is rack mountable, and make sure the power supply (supplied or not) fits in the case.
3.) Include a raspberry pi (or equivalent) so a host computer is not needed.
4.) Use PCI-E connectors to power the boards.
5.) Try to pack at least 1TH/sec in the box.

Basically, the current version of your product is not very data center friendly. It also looks like a poor attempt to copy the S1 design.

[jimmothy]

Miners don't seem to be too excited about them.

It seems the only people not excited are the miners who paid $2+/gh.

They are competitive with other offerings currently on the market, but they are also the last to arrive at the 1W/GH dance, so their window of opportunity to profit looks pretty small (2-3 months)

2-3 months? Are you sure your magic crystal ball is working correctly?

Where are the miners which will make them obsolete?

[xstr8guy]

Miners don't seem to be too excited about them.

It seems the only people not excited are the miners who paid $2+/gh.

They are competitive with other offerings currently on the market, but they are also the last to arrive at the 1W/GH dance, so their window of opportunity to profit looks pretty small (2-3 months)

2-3 months? Are you sure your magic crystal ball is working correctly?

Where are the miners which will make them obsolete?

You don't have to look very hard. The SP10 (and in the near future, SP30) is easily the best option for anyone who is concerned about power efficiency. And I don't even have any Spondoolies miners.

But I, as many others, have reached the max capacity of miners that I can host in my own house. If I'm ever going to upgrade and grow my total THs, waste heat and power efficiency, more so than cost, are the MOST important factors. And the only other option is datacenter hosting. The AM devices announced so far haven't addressed any of these concerns.

Again... too little, too late. Outdated, non-datacenter-friendly enclosures and very inefficient for a 3rd gen design. They'd better hurry and blow these out the door because they have a 6 week shelf life before they are totally obsolete.

[raskul]

Where are the miners which will make them obsolete?

these are the miners which make them obsolete
 
you're welcome

[jimmothy]

Where are the miners which will make them obsolete?

these are the miners which make them obsolete
 
you're welcome

Thats only twice as efficient. I say" only" because it will take much more to make the miners obsolete. Need I remind you that some people are still making a profit off first gen AM cubes.

I doubt that these AM devices will he flying off the shelves when the next gen < 0.5w/gh miners are out but that won't make them obsolete especially for those with cheap electricity.

I'd guess these machines will run for an entire year before needing to be shut down. (Unless electricity costs are abnormally high)

[RoadStress]

Where are the miners which will make them obsolete?

these are the miners which make them obsolete
 
you're welcome

Thats only twice as efficient. I say" only" because it will take much more to make the miners obsolete. Need I remind you that some people are still making a profit off first gen AM cubes.

I doubt that these AM devices will he flying off the shelves when the next gen < 0.5w/gh miners are out but that won't make them obsolete especially for those with cheap electricity.

I'd guess these machines will run for an entire year before needing to be shut down. (Unless electricity costs are abnormally high)

It's not only about cheap electricity. The filling of the power circuit matters too. If a home miner has 5kW available then it's better for him to fill that with as many TH/s as possible by having the most efficient miners. By having 0.8W/GH the miner can have 6TH, instead of 4.5TH/s. It's an almost 50% more in hash rate for the same power circuit.

[jimmothy]

It's not only about cheap electricity. The filling of the power circuit matters too. If a home miner has 5kW available then it's better for him to fill that with as many TH/s as possible by having the most efficient miners. By having 0.8W/GH the miner can have 6TH, instead of 4.5TH/s. It's an almost 50% more in hash rate for the same power circuit.

True, but it is not necessarily about how much TH you can cram in to your garage or else bitfuries would be the clear choice(as of now). $/GH is just more important than w/gh at least at this moment in time. Sure you could spend twice as much money to get 6TH instead of 4.5TH but you might ROI faster with a cheap 4.5th assuming electricity costs are not a limiting factor.

[jimmothy]

They'd better hurry and blow these out the door because they have a 6 week shelf life before they are totally obsolete.

Are you willing to bet on it?

[antirack]

1. The btcgarden miners might be overclocked, not yet stable, badly designed, cheap components, etc... (v1).

2. I am missing information on the power supply used (efficiency, load, etc).

3. The kill-a-watt thing (TM9 power meter) is a copy of a copy and available on the street in China for less than $5. I can only guess why they would use this, but don't expect this to be very accurate.

Have you seen the bf3500? Apparently can hold up to 2880 bitfury chips (11th/s) in a 6u case.

Yeah its insane, will have a review on it next week. Density is only about 3-4x higher than that and 6 weeks later.

4. Isn't the BF3500 is supposed to be 3.5TH at 2800 watts when it is released? Source. And here.

5. BF3500 needs to do 3.5X to 10X better per 6U to keep up with AM's density (38.4TH vs. 3.5TH) or did I get that math wrong (AM's immersion blades are 80 chips @ 10G).

[jimmothy]

4. Yes it comes with 11 blades(up to 24) and 120 chips per blade(2.5gh/chip default) so you can get up to 2880 chips at 3.8gh/chip assuming you are able to properly cool it.

5. AMs density is not that much. IIRC it is around 8kw/8u. Gen3 might be able to step up the density. (I think allied control said their condensers are only working at ~30%)

[Syke]

3. Sales Policy
a) Retail price is around $500 for each single miner

That would be a nice price if it was "shipping today". That's overpriced for "shipping next month".

[Gogreen]

Order and Install the controller board from bitmain? No need repi.  I bet it will sell like hot cake If the prices is right around 0.6 btc

[btcgarden_miner]

BTCGarden,

Here are some honest suggestions on how to improve the next version of your product:
1.) Try to get the power usage under 1W/GH (the lower the better).
2.) Include it in a case that is rack mountable, and make sure the power supply (supplied or not) fits in the case.
3.) Include a raspberry pi (or equivalent) so a host computer is not needed.
4.) Use PCI-E connectors to power the boards.
5.) Try to pack at least 1TH/sec in the box.

Basically, the current version of your product is not very data center friendly. It also looks like a poor attempt to copy the S1 design.

Thank you for your advice.
Now we have new plan to improve our products.

[antirack]

5. AMs density is not that much. IIRC it is around 8kw/8u. Gen3 might be able to step up the density. (I think allied control said their condensers are only working at ~30%)

AM's density is 4.4kW per U, so 8U would be 35.2kW.

From my signature (PDF):
https://docs.google.com/uc?export=download&id=0ByWHHc0u_thNdzB3c2hvVzJkcTQ

DataTank Immersion Slots are simple 19-Inch sheet metal frames holding two 12V power supplies delivering up to 4400W, a power distribution backplane with high current edge connectors and space for up to 8 ASIC boards (blades).

[jimmothy]

5. AMs density is not that much. IIRC it is around 8kw/8u. Gen3 might be able to step up the density. (I think allied control said their condensers are only working at ~30%)

AM's density is 4.4kW per U, so 8U would be 35.2kW.

From my signature (PDF):
https://docs.google.com/uc?export=download&id=0ByWHHc0u_thNdzB3c2hvVzJkcTQ

DataTank Immersion Slots are simple 19-Inch sheet metal frames holding two 12V power supplies delivering up to 4400W, a power distribution backplane with high current edge connectors and space for up to 8 ASIC boards (blades).

That is not the system AM is using unless this has changed: https://drive.google.com/folderview?id=0ByWHHc0u_thNSWhHS18xM2ZXSUU

But yes you are capable of several KW/U using immersion cooling. It is not really even fair to compare it to air cooling regarding efficiency and density.