[2018-02-12]Bitmain To Release Ethereum ASIC Miner F3 With 72Gb DDR3 Dram In Q2

[Vektast]

168Mh/s @ 500W? There is no point in buying that asic...

"hash will be 200-220 MH / s.   Price -    $ 2500- $ 3000."
-------------------

16 February 2018
The world’s largest producer of mining equipment Bitmain is working on the creation of Antminer F3 - ASIC-miner for the second largest capitalization of crypto currency Ethereum. This is reported by IBTimes referring to local technology publications. This is reported by Ib Times.

As the source notes, Antminer F3 will be equipped with three motherboards in each of which will be installed on six devices for the Ethereum mining. Each board will have 32 DDR3 RAM modules with a capacity of 1 GB. According to a number of sources, the hash of the device will be 200-220 MH / s, but with the specified specifications, this figure can be expected to be higher

It is assumed that the production of new equipment for the production of ETH will begin at the end of this month. In the world market the novelty will appear approximately in II-III quarter of this year. The price of Antminer F3 will be set at $ 2500- $ 3000.

Well, actually that's numbers can be valids, because if you do the math with 32 1Gbit ddr3 1866 module*3 (4GB per board) then 32*3*14.9/192*30=~223.5Mh/s (=~7x RX 580).

[1715153884]

1715153884

[1715153884]

1715153884

[kodokbuduk]

168Mh/s @ 500W? There is no point in buying that asic...

"hash will be 200-220 MH / s.   Price -    $ 2500- $ 3000."
-------------------

16 February 2018
The world’s largest producer of mining equipment Bitmain is working on the creation of Antminer F3 - ASIC-miner for the second largest capitalization of crypto currency Ethereum. This is reported by IBTimes referring to local technology publications. This is reported by Ib Times.

As the source notes, Antminer F3 will be equipped with three motherboards in each of which will be installed on six devices for the Ethereum mining. Each board will have 32 DDR3 RAM modules with a capacity of 1 GB. According to a number of sources, the hash of the device will be 200-220 MH / s, but with the specified specifications, this figure can be expected to be higher

It is assumed that the production of new equipment for the production of ETH will begin at the end of this month. In the world market the novelty will appear approximately in II-III quarter of this year. The price of Antminer F3 will be set at $ 2500- $ 3000.

Well, actually that's numbers can be valids, because if you do the math with 32 1Gbit ddr3 1866 module*3 (4GB per board) then 32*3*14.9/192*30=~223.5Mh/s (=~7x RX 580).

How many watt?

[cuteman]

Well, actually that's numbers can be valids, because if you do the math with 32 1Gbit ddr3 1866 module*3 (4GB per board) then 32*3*14.9/192*30=~223.5Mh/s (=~7x RX 580).

Then, this carries no real threat to the estblished GPU mining market.   It is quite similar to Panda B3 miner with 230 MH/s, 1250W  (8 x RX470)

[cuteman]

168Mh/s @ 500W? There is no point in buying that asic...

"hash will be 200-220 MH / s.   Price -    $ 2500- $ 3000."
-------------------

16 February 2018
The world’s largest producer of mining equipment Bitmain is working on the creation of Antminer F3 - ASIC-miner for the second largest capitalization of crypto currency Ethereum. This is reported by IBTimes referring to local technology publications. This is reported by Ib Times.

As the source notes, Antminer F3 will be equipped with three motherboards in each of which will be installed on six devices for the Ethereum mining. Each board will have 32 DDR3 RAM modules with a capacity of 1 GB. According to a number of sources, the hash of the device will be 200-220 MH / s, but with the specified specifications, this figure can be expected to be higher

It is assumed that the production of new equipment for the production of ETH will begin at the end of this month. In the world market the novelty will appear approximately in II-III quarter of this year. The price of Antminer F3 will be set at $ 2500- $ 3000.

Well, actually that's numbers can be valids, because if you do the math with 32 1Gbit ddr3 1866 module*3 (4GB per board) then 32*3*14.9/192*30=~223.5Mh/s (=~7x RX 580).

How many watt?

Power consumption is not mentioned on the article which i cited.    I doubt it is 500W as previously mentioned, as similar miners (Panda) has more than 1kW

[ThugB33]

I don't even know why u ppl panic, cuz you just switch ur rigs to another algo and that's it... that the power of GPU ... multi algo Asic is signle algo ... you are not in a D3 wonderland

[CJNCRYPTO]

DMM Japan most likely have it already ...
and they are test running it in their farms.

F3 specs..
650MH/s  @  750Watts

 

That is like a full 10-12 GPU RIG. anyone know what is gonna be the F3 price?

[Buyingallcoinsz]

I though that developing Ethereum ASIC would be unprofitable because of their plans to switch to PoS, but it seems like Bitmain has calculated otherwise. Maybe it will be usable to mine other coins with the same algorithm. But since it won't be able to mine all GPU algorithms, GPU mining will still be around for a long time, so if this ASIC will take over ETH mining, current ETH miners will just switch to other coins.


I thought so too :O

[moonstruck]

DMM Japan most likely have it already ...
and they are test running it in their farms.

F3 specs..
650MH/s  @  750Watts

 

That is like a full 10-12 GPU RIG. anyone know what is gonna be the F3 price?

As stated above, when taking the 3x32GB boards as true (total 96GB) it would give a theoretical maximum of around 230MH/s.
It's main advantage could be lower cost than equivalent GPU hashing power at around 60% of the power usage.

[VoskCoin]

Fwiw what I’ve been told is that they will be releasing new boxed style gpu rig withs stripped down cards aka 8 generic rx580 etc (think pandaminer style but better)

[coinzoid]

I though that developing Ethereum ASIC would be unprofitable because of their plans to switch to PoS, but it seems like Bitmain has calculated otherwise. Maybe it will be usable to mine other coins with the same algorithm. But since it won't be able to mine all GPU algorithms, GPU mining will still be around for a long time, so if this ASIC will take over ETH mining, current ETH miners will just switch to other coins.


I thought so too :O

If they are able to produce ASIC miners for Dagger, they will be able to produce for other algos and it will cause trouble for gpu miners in the future.

[moonstruck]

I though that developing Ethereum ASIC would be unprofitable because of their plans to switch to PoS, but it seems like Bitmain has calculated otherwise. Maybe it will be usable to mine other coins with the same algorithm. But since it won't be able to mine all GPU algorithms, GPU mining will still be around for a long time, so if this ASIC will take over ETH mining, current ETH miners will just switch to other coins.


I thought so too :O

If they are able to produce ASIC miners for Dagger, they will be able to produce for other algos and it will cause trouble for gpu miners in the future.

They are able to make ASICs for every algo, the question is only when it becomes economically viable to do so.
When you have a coin that is only profitable to mine with up to 10.000 gpu's for example it will not be viable to make an ASIC for only that.

[kiver]

an asic miner for eth hash algo that will shift the whole mining world upside down i know there is alot of other algorithm that can be mined with a gpu but that means that any algo that clams to be an asic proof in the future will be an asic specific for it

[RYXES]

If it is indeed 650Mh/s @ <1000W then I will be making a business trip to Asia this spring.

Scary/Exciting times.

[Wefud]

If it is indeed 650Mh/s @ <1000W then I will be making a business trip to Asia this spring.

Scary/Exciting times.

If that is true, I think there will be a change of the mining algorithm.

[nerdralph]

Well, actually that's numbers can be valids, because if you do the math with 32 1Gbit ddr3 1866 module*3 (4GB per board) then 32*3*14.9/192*30=~223.5Mh/s (=~7x RX 580).

I came up with ~250Mh if they have optimal straps on ddr3-1866.  When it comes to power consumption I expect it to be well under 1000W.  I have no intention of buying one, and I'm not at all concerned it will render my GPUs worthless.  I might even be able to buy a nice new Rx580 for < $200 then.

[nerdralph]

If it is indeed 650Mh/s @ <1000W then I will be making a business trip to Asia this spring.

Scary/Exciting times.

It will be less than half that.  The math is straightforward: ethash requires 8 gigabytes/s of memory bandwidth per megahash.  Add up the number of DDR-3 chips in the miner (18 x 32 = 576), and even if they are overclocking 16-bit wide DDR3-1866, the total speed will be less than 300Mh.

[Branko]

650mh/s with 72 DDR3 memory? Its Simply Nonsense, and I tell u why:
So what we know is the RX 580 has 192Gb/s bandwidth and able to perform ~30mh/s
Well if we choose the best DDR3 memory with 1866Mhz (1866000000 Hz) then we can calculate 1 ddr3 memo chips's bandwidth. It's has a 64bit lane so 1866000000*64*8/1024/1024/1024=14.9 GB/s this is 1 DDr3 memo chip's bandwidth.
14.9*72=1074 Gb/s this is the total memory bandwidth with 72 ddr3 chip. So 1074/192*30=~168 Mh/s this could be the maximum eth mining performance with 72 ddr3 chips. Everything else is bullshit. 

Maybe difference between GDDR5 and DDR3 affects your formula, too? Ethash algorithm maybe actually favor DDR3?

[nerdralph]

650mh/s with 72 DDR3 memory? Its Simply Nonsense, and I tell u why:
So what we know is the RX 580 has 192Gb/s bandwidth and able to perform ~30mh/s
Well if we choose the best DDR3 memory with 1866Mhz (1866000000 Hz) then we can calculate 1 ddr3 memo chips's bandwidth. It's has a 64bit lane so 1866000000*64*8/1024/1024/1024=14.9 GB/s this is 1 DDr3 memo chip's bandwidth.
14.9*72=1074 Gb/s this is the total memory bandwidth with 72 ddr3 chip. So 1074/192*30=~168 Mh/s this could be the maximum eth mining performance with 72 ddr3 chips. Everything else is bullshit. 

Well, you've got the right idea but the wrong numbers.  Rx 580 @2Ghz has 256GB/s of bandwidth not 192 (R9 380 @1.5Ghz has 192).  ddr3 is available in x8 and x16, and since x16 has twice the bandwidth, that's likely what Bitmain will use.  Going with those numbers I come up with 250-300Mh/s.

[wheelz1200]

I dont think people quite understand the amount of gpus are mining eth.  If, and that is a big if this comes out you cant take that amount of gpus and just switch mining to another coin.  Yes in theory you can but that will crush diff in a massive amount of those coins.

[Chromexnet]

I dont think people quite understand the amount of gpus are mining eth.  If, and that is a big if this comes out you cant take that amount of gpus and just switch mining to another coin.  Yes in theory you can but that will crush diff in a massive amount of those coins.

This works both ways. Massive number of GPUs mining Eth means that adding a few thousands ASICs wont affect the difficulty at all. Even adding a hundred thousands ASICs wont render GPU mining non-profitable. I wonder how many ASICS Bitmain can produce. There's a shortage of DDR3/4 chips so they will be limited by the number of chips they can source.