Swedish ASIC miner company kncminer.com

[tzortz]

Are you very sure?

[1714170105]

1714170105

[1714170105]

1714170105

[1714170105]

1714170105

[1714170105]

1714170105

[samsonn25]

Screw the display!

We need more hash power!!

How? You really think they will sell separate boards/boxes? I doubt it.

they will 'make' such they simply won't sell such to customers...i bet they are board specific to the mounts on their current rack mounting PH farsm

the cloud hashing offer works like this ..you cloud hash you don't get a 2nd miner and when we send you a miner we stop (net gain KNC)

so all the taking cloud hashing does is give you hash 'till you get your miner" then that hash goes to knc and the future miner (which is probably
already being made..in case you did not take cloud hashing optioin ..but now that you have goes plop into their PH farm

eventually (except for your 1st piece of equipment) in an ideal world all hash and 2nd miner go back to KNC for use on their farm

works slick huh

as to the above point ..they will NOT sell modules to us once they have shipped neptunes all batches .indeed i suspect they are producint 2 miners
for every one sent to a customer ...the extra miner out the door in this time period will go directly to their PH farm....also any miners after batch 3
will go directly to PH farm....it is win/win/win for knc

get the cloud hashing 'back' they setup for you to hash..get your machine out while likely making 2x as many machines...also the machines that
were refunded..they are also being built to go right to knc and any 'modules' in the future are plug in play int their current rack setup on the giant
PH farm

brillant i must admit

Searing

Batch 3 is "not refundable"

[samsonn25]

Who else is expecting this announcement from KNC next week...

"We are excited to announce that we have 20,000 "new" Jupiter kits for immediate sale at a special discount price of $1000 per THs."

Whatever they cant sell they can undervolt and still run, basically there is no capital cost since its been paid for already or made it back many times over.

[opentoe]

And advanced page:

Do I see right?

VRMs already running at around 99-100C?

Wtf, this is too much.

It will soon spit an error. Ericsson vrms are very sensitive.

Yes indeed, I saw some of them at 106 C.  I noticed as well that the aluminum plate isn't touching all the VRM's with the same pressure. Especially the one's at the end of the U-shape (at the out-flow side): some might have a small space of air (the DC/DC numbers 0 and 1 have significant higher temps).   I'm thinking, to add extra Fans on both sides blowing below the aluminum plate.

Horrible design for cooling. Seems they couldn't figure a way to really cool the VRM's properly so they just slapped some thick thermal pads on there thinking that would help.  Users are going to start to fry these things by just running them in normal ambient temperatures. For at least $10k I really would expect a "plug it in and forget about it" type hardware.

[The Avenger]

Horrible design for cooling. Seems they couldn't figure a way to really cool the VRM's properly so they just slapped some thick thermal pads on there thinking that would help.  Users are going to start to fry these things by just running them in normal ambient temperatures. For at least $10k I really would expect a "plug it in and forget about it" type hardware.

All we need now is KNC's untested, rushed-out-the-door rig and a nice hot summers day 

http://www.youtube.com/watch?v=5GCSWEgZT94

[The Avenger]

I think he's saying it's inside or under the chip like in processors

Just so it's clear - the black components you very clearly see here:



EQUALS

The numbers you see next to the words "Temp":




These components are about 2 to 3 inches away from the chip and the temperature reading is 58C  

Can you even imagine how hot that chip is in it's core?

That far away from the chip, the readings are almost useless.

But as I said, it was done by design. KNC hiding their shitty engineering.

[opentoe]

I feel screwed when I know that KnC's "hash while you wait" means forfeiting "Neptune two-for-one"!

KnC announced "hash while you wait" on April 8, saying first batch customers will start their cloud mining from early June and the last customer will hash in their data center around the last week of June. Now this is close to the end of June, and finally the Plan B arrived.

On April 28, KnC announced that "KnCMiner, the most trusted brand in Bitcoin mining, announced today that it will give every current customer who has an outstanding order for a machine from its Neptune line’s first two batches a free Neptune from its upcoming “third batch” of Neptunes."  I felt happy when KnC announced this, because it shows KnC at least tries to bring more profits to the customers (although, in fact, two Neptune still cannot reach ROI). To my understanding, this is a bonus/gift/giveaway from KnC, it has nothing to do with the compensation due to delay of delivery.

In the email I receved from KnC two days ago, I am notified to choose either "hash while you wait" or "switch to Titan". There was no terms saying that by select any of the options, the "neptune two-for-one" plan will no longer be valid. I read their T&C (convert.kncminer.com/terms/), the two options they offered has nothing to do with the two-for-one plan, therefore I signed up "hash while you wait", happily waiting for the cloud mining next week, first Neptune by the end of June, and second one in August.

Then I read the forum discussion, and I know that KnC actually use "hash while you wait" in replace of "neptune two-for-one"! I feel totally cheated and pissed off! How come KnC breaks their public promises, and deprive my rights by just giving some vague and ambiguous terms and conditions?

I use the online calculator for some maths: https://bitcoinwisdom.com/bitcoin/calculator. (I set the difficulty increase 20%, power consumption 2.1kW, the rest parameters are with default values)

I spent 12999 USD on 3TH Neptune. With current difficulty, even if I start mining today, I can mine up to 6BTC before electricity fee exceeds the income. Suppose bitcoin price is 600USD, I can get back 3600 USD. There is still a gap of 9399USD!

If I receive a second neptune by Aug 31 (earliest estimated delivery date), then before electricity bill wins, I can mine up to 8.4 BTC together with the first neptune. This is about 5040USD in total. And the gap is still 7959USD!

"hash while you wait" has a max period of 45 days. Suppose I don't receive my neptune and therefore I use up all the cloud mining time, then I can get around 3.8BTC, equivalent to 2280USD. Of course, if KnC ships my first Neptune by end of June, then I only get 0.8-0.9 BTC while waiting...

So, even if I have both "hash while you wait" and "neptune 2-for-1", I can get maximum $(5040+2280)=$7320. The gap is $5679.

If I only have "hash while you wait" and one neptune, then revenue is $(3600+2280)=$5880, gap is $7119!

Now I can buy a 1TH miner for about 1500-1700 USD(e.g. http://www.lketc.com/goods/show-172.aspx  the dragon miner, one of the well-known mining rigs in China), and such mining rig is in stock and can be delivered in 2-3 days. With 12999 USD, I can buy up to 7-8 TH mining power! This is far more than 6TB that KnC can provide.

KnC said Neptune is to be delivered in Q1/Q2, and they make customers believe it refers to March/early April. Now it is the very end of Q2, and I feel disappointed.

KnC said the "hash while you wait" will be started in early June, and it is delayed, and I feel disappointed.

Now KnC will revoke their promises of give away a second Neptune, I am totally pissed!

We ALL said soon as we saw Q1/Q2 that means the end of Q2. I must have said that 20 times when they announced those dates but everyone ran out and dropped $10k (some $13K) on a pre-order that was 6 months away. I feel bad for people that received shitty deals, but don't have any empathy for the complainers about receiving hardware now and complaining it is so late. That's just the way it is with bitcoin hardware. It is more of a gamble these days picking the right horse to win. Who knows.

[jelin1984]

Final no roi for anyone




Knc must simple
Sell at December undervolt Jupiter chips
At 10.000 dollars
Why they spend so many dollars for make an awful chip
I can not understand

[samsonn25]

I think he's saying it's inside or under the chip like in processors

Just so it's clear - the black components you very clearly see here:



EQUALS

The numbers you see next to the words "Temp":




These components are about 2 to 3 inches away from the chip and the temperature reading is 58C  

Can you even imagine how hot that chip is in it's core?

That far away from the chip, the readings are almost useless.

But as I said, it was done by design. KNC hiding their shitty engineering.

More smoke and mirrors. That way they show you a lower temp instead of 100 degreed Celsius on cbip. Which is pretty damn hot

[The Avenger]

More smoke and mirrors. That way they show you a lower temp instead of 100 degreed Celsius on cbip. Which is pretty damn hot

Exactly.

And we all know that excessive heat and electronic components do not mix very well.

[tzortz]

Core temps cannot be correctly measured that far away.

You are right. That is why those chips shown circled in red in the original post are not used for core temperature measurements.


You need to learn when AVNGR is just being a troll and when he is serious. Hint:  he's always a troll.

Can you please explain, whether or not is a temp sensor inside the core chipset?

[opentoe]

Get an iR temperature dot on that VRM and see how hot it really is. I'll guarantee you the bottom PCB's of those boards are already starting to turn brown from burn.

My Dad's Jupiter had a DIE die on him and KNC replaced it with a used board. We really didn't expect a new one anyway, and the bottom of the "new" PCB board was all brown from burning in. Long story short the replacement board worked and I made sure to tell him to try and keep it cool as possible. The VRMs were running so damn hot they physically burnt the board. Crazy that KNC would actually send that one back to replace a broken one. Cost $65 to send one little PCB back to Sweden, that's why we didn't ask for another one.

And where the heck is an open box??? Someone must have recorded one by now?

[samsonn25]

There should be.  But only KNC can see it through their programming. And it depends if they want you to know or not.

[samsonn25]

Get an iR temperature dot on that VRM and see how hot it really is. I'll guarantee you the bottom PCB's of those boards are already starting to turn brown from burn.

My Dad's Jupiter had a DIE die on him and KNC replaced it with a used board. We really didn't expect a new one anyway, and the bottom of the "new" PCB board was all brown from burning in. Long story short the replacement board worked and I made sure to tell him to try and keep it cool as possible. The VRMs were running so damn hot they physically burnt the board. Crazy that KNC would actually send that one back to replace a broken one. Cost $65 to send one little PCB back to Sweden, that's why we didn't ask for another one.

And where the heck is an open box??? Someone must have recorded one by now?

Was your temp over 95?

[Collider]

Just ssh into the control unit and reattach the cgminer screen, then look for temperatures

On the other hand, most silicon does fine at 124°C, so no worries about temps.

[dropt]

You are right. That is why those chips shown circled in red in the original post are not used for core temperature measurements.

What do you propose they are measuring way out there on the edge of the board?

[fubly]

Todo for KNC:

absolutely new design of the asic and controller board.

It could not be, that others squeezze from an 28nm chip 750 Gh/s and knc only 660 Gh/s on 20nm Chip. I guess batch 4 will based also on 20nm chip but new bord-design.

[samsonn25]

 But for longevity and life span under 60 degrees is highly recommended.

[The Avenger]

Can you please explain, whether or not is a temp sensor inside the core chipset?

There is no temperature sensor in the chips. The black components I've pointed out are used to give a rough indication of how hot the chips are. If you see my last post, it shows where this data is displayed in the interface.

This temperature component was closer to the chips on Jupiter, but much further away on Neptune boards. I've already explained why I think this was done by knc, but people can make up their own mind if they think differently.

I can't wait until someone with a thermal imaging camera gets one of these rigs and we get to see some REAL core temperatures 

[samsonn25]

You are right. That is why those chips shown circled in red in the original post are not used for core temperature measurements.

What do you propose they are measuring way out there on the edge of the board?

To move the sensor closer to where the chip is at.  Lol