Swedish ASIC miner company kncminer.com

[Templer]

OOooh, I like the colored ring on the fans... haha

i think its that reflective hologram type stuff

lol 

[1715481037]

1715481037

[1715481037]

1715481037

[1715481037]

1715481037

[1715481037]

1715481037

[chrono030]

Does anyone know why there is a 4 pin connector on the front PCB?  I assume that it is to power the on board Linux OrSOC device?

[Phoenix1969]

Does anyone know why there is a 4 pin connector on the front PCB?  I assume that it is to power the on board Linux OrSOC device?

yes, the linux device is powered by the molex connector...

[dropt]

I assume that it is to power the on board Linux OrSOC device?

It looks like a Beaglebone.   Further, the cut-out to the left  looks as though it might be for a rPi.

[Phoenix1969]

hahaha, no....no Rpi... it has its own linux onboard, with GUI & ssl ...a pi would be useless..
but is does seem, the chips are in, and assembly & test has begun!!...!!..!!

[jimrome]

hahaha, no....no Rpi... it has its own linux onboard, with GUI & ssl ...a pi would be useless..
but is does seem, the chips are in, and assembly & test has begun!!...!!..!!

I'm not sure how/why you believe a Pi would be useless..

I'm also not sure that chips are in, did they indicate otherwise?

EDIT: you do know that a Pi would offer onboard linux providing GUI, SSL, etc right?

[rolling]

hahaha, no....no Rpi... it has its own linux onboard, with GUI & ssl ...a pi would be useless..
but is does seem, the chips are in, and assembly & test has begun!!...!!..!!

Source?

[Miner538]

hahaha, no....no Rpi... it has its own linux onboard, with GUI & ssl ...a pi would be useless..
but is does seem, the chips are in, and assembly & test has begun!!...!!..!!

That would mean my bitbet bet is gonna pay off!!

[Phoenix1969]

hahaha, no....no Rpi... it has its own linux onboard, with GUI & ssl ...a pi would be useless..
but is does seem, the chips are in, and assembly & test has begun!!...!!..!!

I'm not sure how/why you believe a Pi would be useless..

I'm also not sure that chips are in, did they indicate otherwise?

What could the pi do that an onboard linux couldn't? That's why useless....
The new.pics have the heatsinks mounted, which means the chips are under them. a 1/4 mounted alterra fpga would most likely damage the board if you tried to mount those heatsinks to them, so I'm saying, the real chips must be in, and in those pics.

[DeathAndTaxes]

So I guess the corsair is quite a good choice:
* dedicated single +12V rail with user-configurable virtual "single rail" and "multi-rail" software modes
* features a massive 71.6 Amp (859.2 Watt) single +12V rail
* can also be configured as a multi-rail device with individual PCI-E over-current protection (OCP) trip points
=> http://www.corsair.com/us/blog/ax860i_technical_details/

"Probably".  The big unknown on any single rail PSU is what safety overcurrent protection is set to.  Looking at the screenshot in the review the user adjustable OCP is up to 40A per PSU connector (not to be confused with the downstream PC connectors).  That likely means the default is 40A and obviously they would have designed their plugs, pins, and wires to handle that much current.  If you used just a pair of PSU connectors each with 1 PSU connector running to two PCIe connectors in series, then each PSU per connector "virtual rail" would have half the load of a Jupiter.  If we assume Jupiter is 790W that is 65A total or 32.5A per connector.  Looks good to me.  Not really going to know for sure until the first person plugs on in but I would be comfortable buying it (if I didn't have 10 spare Seasonic 1050W PSUs ).

Worst case scenario (for any PSU) I am sure someone like Cablez can sell a EPS12V to PCIE 8 pin connector.  The EPS12V connector (8 pins usually split as 4pin + 4pin) which connects to the motherboard is designed for >280W.  To any DIY who later finds this post by a search and acts without thinking the pinout for EPS12V and PCIe-8pin are different.  If you connect them 1:1 you will destroy your PSU and equipment.  It would require a custom cable to properly connect the 12V to 12V and ground to ground.

[jimrome]

hahaha, no....no Rpi... it has its own linux onboard, with GUI & ssl ...a pi would be useless..
but is does seem, the chips are in, and assembly & test has begun!!...!!..!!

I'm not sure how/why you believe a Pi would be useless..

I'm also not sure that chips are in, did they indicate otherwise?

What could the pi do that an onboard linux couldn't? That's why useless....
The new.pics have the heatsinks mounted, which means the chips are under them. a 1/4 mounted alterra fpga would most likely damage the board if you tried to mount those heatsinks to them, so I'm saying, the real chips must be in, and in those pics.

Maybe the "onboard linux" is an RPi? Also, the heatsinks could easily be mounted without the chips in place for product photography purposes.

You're jumping to conclusions - although I'd be very happy if you're right. 

[Syke]

The new.pics have the heatsinks mounted, which means the chips are under them. a 1/4 mounted alterra fpga would most likely damage the board if you tried to mount those heatsinks to them, so I'm saying, the real chips must be in, and in those pics.

And what would happen if the heatsinks were mounted on board without any chips?

[DeathAndTaxes]

hahaha, no....no Rpi... it has its own linux onboard, with GUI & ssl ...a pi would be useless..
but is does seem, the chips are in, and assembly & test has begun!!...!!..!!

Pi can run a "webserver" with SSL with no problem. 

[DeathAndTaxes]

Maybe the "onboard linux" is an RPi?

Early KNC indicate the host would be a embedded sodimm board so not a RPi.  Then again it doesn't really matter.  The host is relatively low end stuff.  It needs a tiny amount of memory, a low end processor, ethernet, and some sort of serial connectivity to the ASIC boards.  An embedded linux board will work fine, so will a RPi.  I assume they simply have more experience with this board so they are using it.

[Phoenix1969]

The new.pics have the heatsinks mounted, which means the chips are under them. a 1/4 mounted alterra fpga would most likely damage the board if you tried to mount those heatsinks to them, so I'm saying, the real chips must be in, and in those pics.

And what would happen if the heatsinks were mounted on board without any chips?

that could have been done weeks ago... it wasn't.  Nobody wants to assemble something that will never function, and we've been waiting for this...  I'd bet on it. It's an obvious teaser! I'm stoked!

[crumbs]

The new.pics have the heatsinks mounted, which means the chips are under them. a 1/4 mounted alterra fpga would most likely damage the board if you tried to mount those heatsinks to them, so I'm saying, the real chips must be in, and in those pics.

And what would happen if the heatsinks were mounted on board without any chips?

that could have been done weeks ago... it wasn't.  Nobody wants to assemble something that will never function, and we've been waiting for this...  I'd bet on it. It's an obvious teaser! I'm stoked!

You obviously have never built anything in your life -- the concept is called "mockup."

For instance, the pics make it obvious that the air partitions suggested in the cad rendering will be difficult to mount.

[noodle73]

I'm so glad it's shiny. Shiny = "the future".

[kingcoin]

I assume that it is to power the on board Linux OrSOC device?

It looks like a Beaglebone.   Further, the cut-out to the left  looks as though it might be for a rPi.

It's an OrSoC embedded Linux DIMM module:

[DeathAndTaxes]

The new.pics have the heatsinks mounted, which means the chips are under them. a 1/4 mounted alterra fpga would most likely damage the board if you tried to mount those heatsinks to them, so I'm saying, the real chips must be in, and in those pics.

And what would happen if the heatsinks were mounted on board without any chips?

that could have been done weeks ago... it wasn't.  Nobody wants to assemble something that will never function, and we've been waiting for this...  I'd bet on it. It's an obvious teaser! I'm stoked!

That is kinda a leap.  KNC is still selling Nov units.  Showing a populated board would boost sales as it would give new customers confidence that if they are on track for Sept/Oct then they are on track for Nov.  At this point it is a race between various companies.  There is only so much money which can go into hardware before ROI approaches 0% so a missed sale today may not be made tomorrow.  Given that if they had the chips and the chips on a board I would have assumed they would snap a couple pics.  If they had an operating board I am sure they would span a pic of it mining.  Both would explode sales for Nov.

Also the Alterra FPGA requires active cooling.  Why would they mount it to the board if it wasn't to test the board?  If they are testing the board they need to cool it and the mounting hole only allow a full size heat sink.  So I am sure they intend to use the Alterra with the production heat sink as the goal of using a FPGA is to test the entire system (host, power supply, connectors, cooling, etc).  It isn't that difficult of a problem.  The simple but expensive option would be to just mount 4 FPGA on one prototype board which would make the most sense as it would allow the most comprehensive testing.  If they are only using a single FPGA a nonconductive shim isn't that difficult.

[ASIC-K]

i tend to agree that it is probably just a mock up as well.