EBang 14nm 6.3T (140W/T) BTC Miner for sale soon

[HagssFIN]

Let them just give to some of us a test unit (for testing period or with postpaid), this guy (these guys may be) will review it and claim to all community that they are legit.

It is available @ Aliexpress,

Any links, please?

https://www.aliexpress.com/item/Ebit-E9-Miner-6-5TBitcoin-Miner-Newest-14nm-Asic-Miner-only-1000w-Btc-Miner-Better-Than/32783023069.html?spm=2114.01010208.3.320.EB03yX&ws_ab_test=searchweb0_0,searchweb201602_4_10065_10068_10000009_10084_10000025_10083_10000029_10080_10082_10081_10000028_10110_10111_10112_10060_10113_10062_10114_10056_503_10055_10054_10059_10099_501_10078_10079_10000022_10000012_10103_10073_10102_10000015_10096_10000018_10000019_10052_10053_10107_10050_10106_10051,searchweb201603_2,afswitch_4,single_sort_2_default&btsid=ee895104-cb4b-429d-b0a9-6f3a50908c2e

At that price it doesn't make any sense to buy one because the Antminer R4, batch 5 (7.5Th/s) costs only 916 USD.

[philipma1957]

I see 4 pcie jacks  and a meter at 1020 watts.

I smell melted plastic

I would never use any gear that looks for 255 watts a jack   yeah I know it is at the wall reading but I have melted enough wires to know it is a fuckup on the builders part to spec 4 jacks  and not six.

1020 watts at the wall  the psu is 93% so 950 watts is 237 watts at the jack run that 24/7/365  I won't



4 pcie jacks

[QuintLeo]

I see 4 pcie jacks  and a meter at 1020 watts.

I smell melted plastic

 No worse than the Spondoolies SP20E - most of which ran QUITE reliably, if not at their quoted hashrate, pulling a machine-limited 288 watts per connector (which IS the spec for the CONNECTOR).

 Titans and Neptunes had issues 'cause they were trying to pull 300+ watts through those same connectors, and had marginal setups on the circuit board to handle the current level.

 They definitely need to work on communications and on pricing to be competative though, except perhaps for folks that want miners "RIGHT NOW" and don't want to wait on the next batch of whatever out of BitMain or pay Avalon pricing for the 721.

[tntdgcr]

I see 4 pcie jacks  and a meter at 1020 watts.

I smell melted plastic

 No worse than the Spondoolies SP20E - most of which ran QUITE reliably, if not at their quoted hashrate, pulling a machine-limited 288 watts per connector (which IS the spec for the CONNECTOR).

 Titans and Neptunes had issues 'cause they were trying to pull 300+ watts through those same connectors, and had marginal setups on the circuit board to handle the current level.

 They definitely need to work on communications and on pricing to be competative though, except perhaps for folks that want miners "RIGHT NOW" and don't want to wait on the next batch of whatever out of BitMain or pay Avalon pricing for the 721.

the SP20/20E pull that 288 like nothing. Out of 1000s , maybe 2-3 have had PCIe failure. good guage, good psu, good running environment, best single box model for profit per box.

[philipma1957]

Many sp20s melted cables.

Using 2 jacks per board is simply not wise.

The biggest reason the sp20s did not melt more then they did was they became really huge power hogs at top speed most of use ran them at 700-800 watts and 1250-1350 gh.

When people pushed to get 1550-1650gh the power went up and many pcie cables failed.


A six pin cable rates 150. Which is conservative 200 is doable but 237. Watts  at  24/7/365 is asking for melted plugs and or jacks

[adaseb]

From that photo it looks like there are 6 PCIe connections

[philipma1957]

From that photo it looks like there are 6 PCIe connections

Maybe you are correct. But on the website expanded as big as I could it looks like two.

I am a gear head I buy tons of different gear.  All the miners that draw 225 watts or more per jack have melted or heated cables too much...

So I will let others be first on this gear.

[QuintLeo]

Per the initial review in another thread it's 3 connectors per board, so the point is moot - that's well under 200 watts/connector.


BTW, it's not the fault of the CONNECTOR that some folks had cables melt 'cause they didn't use cables with adequate wire gauge.

[tntdgcr]

Per the initial review in another thread it's 3 connectors per board, so the point is moot - that's well under 200 watts/connector.


BTW, it's not the fault of the CONNECTOR that some folks had cables melt 'cause they didn't use cables with adequate wire gauge.

Correct, 16 AWG or bust.

[italianMiner72]

it seem eBang have updated theyr ASIC miner...
from 4XPCIe port to 6XPCIe.
some docs here...

[fanatic26]

I posted pictures of the 6 PCI-E version months ago in the other ebang thread. I am actually going to do some amp clamp tests to get REAL power numbers today. I will post what I find so everyone can have an accurate idea of the efficiency.

[italianMiner72]

I posted pictures of the 6 PCI-E version months ago in the other ebang thread. I am actually going to do some amp clamp tests to get REAL power numbers today. I will post what I find so everyone can have an accurate idea of the efficiency.

i'm waiting for this date ...
i was chatting on Skype with EBIT E9 customer support ...
they Told me 780$+100$ for delivery in EU eBit E9 + psu + shipment.
i also ask why, with a 14nm chip, ASIC E9 is less efficient than the S9 with a 16nm chip ..
if i ask too is a problem of firmware optimization ..
but i have got no answer!

[NotFuzzyWarm]

I posted pictures of the 6 PCI-E version months ago in the other ebang thread. I am actually going to do some amp clamp tests to get REAL power numbers today. I will post what I find so everyone can have an accurate idea of the efficiency.

i'm waiting for this date ...
i was chatting on Skype with EBIT E9 customer support ...
they Told me 780$+100$ for delivery in EU eBit E9 + psu + shipment.
i also ask why, with a 14nm chip, ASIC E9 is less efficient than the S9 with a 16nm chip ..
if i ask too is a problem of firmware optimization ..
but i have got no answer!

14/16nm for all intents and purposes are the same regarding efficiency. To date only Samsumg and Global Foundries (under license with Samsung) is producing 14nm chips and TSMC doing 16nm which is slightly easier to produce.. Intel and IBM don't count because they do not share their production fabs with outsiders.

As for comparing to the s9 -- same thing as comparing a s9 to the Avalons or BitFury rigs which both use 16nm chips or even to the T9 (same chip as s9): When used in the s9 and R4 miners the design was driven to be be "highest efficiency on the planet". It worked but at the expense of the miners often being unstable/unreliable because the chips are running at bare minimum core voltage (Vcore).

Same chip used in the s9/R4 is used in the T9 and so far no one has trouble with the T9. Why? Because Bitmain is using a higher Vcore which increases power usage but also make the chips more stable. Same for Avalons and BitFury and probaly EBang: Higher Vcore for more stability. Good choice to make.

[fanatic26]

Also this is Ebangs first mining chip and it has not been on the market very long. On something as complicated as in house design of an ASIC the chances of them being the best/most efficient on the first try is most likely pretty low.

[tntdgcr]

Also this is Ebangs first mining chip and it has not been on the market very long. On something as complicated as in house design of an ASIC the chances of them being the best/most efficient on the first try is most likely pretty low.

while the HTTPS broken admin in Chinese was a PITA , after setup Ebang pose 0 problem. Consistent run times for ages now.

[fanatic26]

Sorry about the delay on the power numbers.


The miner is running 6.48th. I am seeing 4.72 amp on 208 (PSU is 93% efficient) so they seem to pull around 1000w at the wall.

[alred911]

Is anyone compare the noise level between 1 S9 and  2 Ebang E9?

[fanatic26]

with just a single fan the e9 is really loud, comparable to an S9 by itself