Swedish ASIC miner company kncminer.com - page 1396.

Puppet

legendary

Activity: 980

Merit: 1040

Quote from: donch on October 11, 2013, 04:53:39 AM

I think the interesting point that hasn't been raised yet (as far as I've seen in this epic thread) is how KNC have binned their chips, if at all. BFL have 2 chips per Jalapeno and each can be (and usually seems to be) of a different quality in terms of hash rate and they must manually balance which chips get put on which board to achieve the advertised hashrate. Bigger boards have way more chips and it then becomes a classic partitioning problem.

Very good point, AFAIK, the chips are currently not binned at all. Not even tested for functioning (neither wafer probing or final chip testing) before being soldered on the PCB. Thats a huge mistake IMO, one that was pointed out months ago. They could be testing/binning the completed modules, but Ive not heard a word on that.

Searing

copper member

Activity: 2898

Merit: 1465

Clueless!

Quote from: demonmaestro on October 11, 2013, 04:03:23 AM

or maybe they should of gave a miner to the designer of bfgminer/cgminer...

screw that ....put them both on a plane....with "agreeable blondes" ....fly them to Sweden work them like slave pay them well....they can keep the blondes

(hey just saying for coders you could run the world on software.....maybe that's how google does it/ ... with this method)

but truthfully a plane ride accomodations and 'good pay' would more then make it self up with the miner software all buffed up and shiny imho

Searing

donch

member

Activity: 84

Merit: 10

Updated ironic image.

I think the interesting point that hasn't been raised yet (as far as I've seen in this epic thread) is how KNC have binned their chips, if at all. BFL have 2 chips per Jalapeno and each can be (and usually seems to be) of a different quality in terms of hash rate and they must manually balance which chips get put on which board to achieve the advertised hashrate. Bigger boards have way more chips and it then becomes a classic partitioning problem.

Given that a Jupiter has only 4 chips, each of which must be of a different quality (bin) how are they choosing which chips go in which box? Pure lottery or an undisclosed system? I understand that each chip has many cores and the firmware is attempting to disable bad cores in an effort to manage the binning at runtime, but the variance in the aggregate chip hashrates shown by people on this thread show that this strategy isn't yet working as intended.

Simple question: KNC are you speed testing each chip/board before placing it in a box?

Second question: What is the minimum guaranteed hashrate of a Jupiter?

trepex

sr. member

Activity: 266

Merit: 250

I don't want to complain too much as I am a DAY1 customer that did receive one Saturn at home and another one
online hashing at the KnC datacenter. I also received my (first ?) bonus/compensation from KnC.
But I also thought I post this comparision (as reported by Bitminter):

http://postimg.org/image/82x69elyj/

My Saturn unit at home runs 0.93, case closed (but optimized the flaps/wings inside the case = bent them to not disturb the airflow)
and uses 270W (at 230V).

r3animation

hero member

Activity: 756

Merit: 500

Quote from: cypherdoc on October 11, 2013, 04:22:13 AM

Quote from: r3animation on October 11, 2013, 04:14:14 AM

Quote from: cypherdoc on October 11, 2013, 04:12:23 AM

Quote from: donch on October 11, 2013, 03:54:23 AM

Cypherdoc - FUDcaster General, coming to an unrelated thread near you very soon.

Actually if you look carefully, I never came into this thread to criticize or spread fud prior to now. That really doesn't make sense if I was trying to disrupt their business. In fact I was one of the first to congratulate them a couple of weeks ago after the demo video.

Buy yeah, you know what? I am a big protector of bitcoin in general and when I see any company screw up the economy after having collected so much money and behaved so arrogantly Imo, I'll let them have it.

Did you let BFL, "have it" then?

Good point. No.

But the KNC warrior bunch was a one of a kind virulent bunch of trolls that tried to destroy Hashfast and my reputation based on nothing except to try and protect their investment.

They also kept this thread pushed up to the top for visibility.

Now things have gone terribly wrong.

Well what about Avalon then? They are currently doing a few bad things.

Mota

legendary

Activity: 804

Merit: 1002

Quote from: cypherdoc on October 11, 2013, 04:22:13 AM

Quote from: r3animation on October 11, 2013, 04:14:14 AM

Quote from: cypherdoc on October 11, 2013, 04:12:23 AM

Quote from: donch on October 11, 2013, 03:54:23 AM

Cypherdoc - FUDcaster General, coming to an unrelated thread near you very soon.

Actually if you look carefully, I never came into this thread to criticize or spread fud prior to now. That really doesn't make sense if I was trying to disrupt their business. In fact I was one of the first to congratulate them a couple of weeks ago after the demo video.

Buy yeah, you know what? I am a big protector of bitcoin in general and when I see any company screw up the economy after having collected so much money and behaved so arrogantly Imo, I'll let them have it.

Did you let BFL, "have it" then?

Good point. No.

But the KNC warrior bunch was a one of a kind virulent bunch of trolls that tried to destroy Hashfast and my reputation based on nothing except to try and protect their investment.

They also kept this thread pushed up to the top for visibility.

Now things have gone terribly wrong.

Hashfast still does not accept paypal or cc. Easy as that. When they do, you can come back here and troll on. Until then my consumer rights are more protected with KNC. Also - A miner Protection Program for a whooping 5400k premium? Bitch, please....

cypherdoc

legendary

Activity: 1764

Merit: 1002

Quote from: r3animation on October 11, 2013, 04:14:14 AM

Quote from: cypherdoc on October 11, 2013, 04:12:23 AM

Quote from: donch on October 11, 2013, 03:54:23 AM

Cypherdoc - FUDcaster General, coming to an unrelated thread near you very soon.

Actually if you look carefully, I never came into this thread to criticize or spread fud prior to now. That really doesn't make sense if I was trying to disrupt their business. In fact I was one of the first to congratulate them a couple of weeks ago after the demo video.

Buy yeah, you know what? I am a big protector of bitcoin in general and when I see any company screw up the economy after having collected so much money and behaved so arrogantly Imo, I'll let them have it.

Did you let BFL, "have it" then?

Good point. No.

But the KNC warrior bunch was a one of a kind virulent bunch of trolls that tried to destroy Hashfast and my reputation based on nothing except to try and protect their investment.

They also kept this thread pushed up to the top for visibility.

Now things have gone terribly wrong.

The Avenger

hero member

Activity: 798

Merit: 1000

Quote from: CYPER on October 10, 2013, 08:35:54 PM

I'm thinking of buying 2x 140mm fans and just put them with cables ties on the outside of the case at the rear (exhaust side), so they can help a little bit with removing the hot air.

Thermalright TY 143 look nice and coupled with a fan controller you can run them quiet and slow or noisy and fast.

L152 mm x H140 mm x W26.5 mm
Fan Weight : 170g
Fan Speed : 600 - 2500 rpm (PWM)
Sound Level : 21 - 45 dB(A)
Airflow : 31.4 - 130.0 CFM

Or one can get a more advanced one with a display and thermal probes = monitor the temps inside the case directly

Hey Cyper, wouldn't it be better to add the souped-up, more powerful fans as the inlet fans and put the existing f14's at the rear to suck out the air? I think it would be more important for temperatures to improve the inflow than the outflow.

donch

member

Activity: 84

Merit: 10

Updated ironic image.

CypherDoc - feel free to start a new thread with your views. You are just spamming this one.

r3animation

hero member

Activity: 756

Merit: 500

Quote from: cypherdoc on October 11, 2013, 04:12:23 AM

Quote from: donch on October 11, 2013, 03:54:23 AM

Cypherdoc - FUDcaster General, coming to an unrelated thread near you very soon.

Actually if you look carefully, I never came into this thread to criticize or spread fud prior to now. That really doesn't make sense if I was trying to disrupt their business. In fact I was one of the first to congratulate them a couple of weeks ago after the demo video.

Buy yeah, you know what? I am a big protector of bitcoin in general and when I see any company screw up the economy after having collected so much money and behaved so arrogantly Imo, I'll let them have it.

Did you let BFL, "have it" then?

cypherdoc

legendary

Activity: 1764

Merit: 1002

Quote from: donch on October 11, 2013, 03:54:23 AM

Cypherdoc - FUDcaster General, coming to an unrelated thread near you very soon.

Actually if you look carefully, I never came into this thread to criticize or spread fud prior to now. That really doesn't make sense if I was trying to disrupt their business. In fact I was one of the first to congratulate them a couple of weeks ago after the demo video.

Buy yeah, you know what? I am a big protector of bitcoin in general and when I see any company screw up the economy after having collected so much money and behaved so arrogantly Imo, I'll let them have it.

Puppet

legendary

Activity: 980

Merit: 1040

Quote from: cypherdoc on October 11, 2013, 03:52:29 AM

I've already said I'll eat my words if the f*ck up is a bad as this. I don't think being late is the issue here. It's mostly product design.

But I really don't think it will be this bad. We'll see though.

Arguably, HF already fucked up about as bad as KnC.
KnC promised end of September and only a few days later at least a fair number of users received their hardware working better than originally promised.

HF's current one week delay (assuming it ends there, which I highly doubt) would make achieving what KnC did, shipping less than a week late, over spec to at least a tangible fraction of their customers, very unlikely. And the real pitfalls for HF are yet to come.

Remains to be seen if the problems for the rest of KnC can be solved, we dont know that yet. Its at least feasible a firmware update will fix it for most other customers. Personally, I doubt it, but I doubt just as much that HF's execution will be better.

ur0pl

newbie

Activity: 56

Merit: 0

Quote from: demonmaestro on October 11, 2013, 04:03:23 AM

or maybe they should of gave a miner to the designer of bfgminer/cgminer...

Luke-Jr will have a look at it next week. The source code will be released. He is the author of bfgminer. But yeah, they should have given him a free machine and let him do testing. He could probably do better firmware work than they are doing too.

demonmaestro

hero member

Activity: 574

Merit: 500

Mining for the hell of it.

or maybe they should of gave a miner to the designer of bfgminer/cgminer...

ur0pl

newbie

Activity: 56

Merit: 0

Quote from: timmmers on October 11, 2013, 03:42:36 AM

Quote from: ur0pl on October 11, 2013, 03:28:46 AM

I think they say it is a firmware issue mostly because the hardware is basically the same in all the units.

I hope that's right because it looks to me like they have a shitload of bad chips and only found out when they arrived and didn't perform, and are now having to test them as they go which is slowing them right down. They definitely didn't have time to test them before they announced that happy hashrate in the video and sent a bunch out. I remember people saying that rigs arrived with no activity on their pool accounts too, so maybe the first time they ran in anger was when they arrived at the customers?

Maybe they can make a firmware where cores become disabled, but then they are re-enabled quicker or something, or maybe a firmware where disabled cores are enabled, but at 80% of what they were trying to do , or something.

donch

member

Activity: 84

Merit: 10

Updated ironic image.

Cypherdoc - FUDcaster General, coming to an unrelated thread near you very soon.

cypherdoc

legendary

Activity: 1764

Merit: 1002

Quote from: Puppet on October 11, 2013, 03:47:14 AM

Quote from: cypherdoc on October 11, 2013, 03:38:27 AM

Oh, so I'm the first one around here to be doing that? These are rough parts around here in case you hadn't noticed. I sure did back when I endorsed HF.

Seems people can dish it out but can't take it back.

Cant speak for others, but Im mainly curious how you will react when HF runs in to similar problems and delays. Their schedule is every bit as unbelievable as KnC's was. HF chips havent even made it through the fab yet and already they are running a week behind schedule making their october promise all but impossible. The most tangible thing we've seen from HF so far is renderings and a FPGA simulation, not even running on their own PCB. Ie, about what we've seen from Cointerra which isnt due for another 2-3 months.

Time will tell if you turn out to be a hypocrite or not.

I've already said I'll eat my words if the f*ck up is a bad as this. I don't think being late is the issue here. It's mostly product design.

But I really don't think it will be this bad. We'll see though.

BitThink

legendary

Activity: 882

Merit: 1000

Quote from: lindeanin on October 11, 2013, 03:43:53 AM

Quote from: ur0pl on October 11, 2013, 03:09:04 AM

If KNCMiner doesn't ship both of my Jupiter units by the end of next week or refund any money, then I might kill myself for financial reasons.

People overvalue one or two weeks. In fact, your miner will make only slight return in two weeks.

Try to chill out

With the current difficulty increase rate, you can only mine 70% BTC if late for 10 days and 49% BTC if late for 20 days.

cypherdoc

legendary

Activity: 1764

Merit: 1002

Quote from: -Redacted- on October 11, 2013, 03:25:25 AM

Quote from: cypherdoc on October 11, 2013, 02:42:55 AM

Yeah where is that idiot Bitcoinorama when you need him?

Imo, this is a multi - pronged problem. Not only was the packaging inadequate but more importantly we have a design problem. going from 8 to 4 VRM was a mistake. Double the phases means half the current, means smoother power, means more stability for over clocking, means less heat, means less blowing up. I doubt it's just firmware.

@Cypherdoc

Why are you, a HashFast representative, in a competitors thread piling on? I'm not saying there isn't criticism due, but this kind of stuff really makes HashFast look bad. You might want to think about that some...

And furthermore, I remember you to be one of the more disrespectful ones criticizing.

And now I know why Mr.$50K order.

Puppet

legendary

Activity: 980

Merit: 1040

Quote from: cypherdoc on October 11, 2013, 03:38:27 AM

Oh, so I'm the first one around here to be doing that? These are rough parts around here in case you hadn't noticed. I sure did back when I endorsed HF.

Seems people can dish it out but can't take it back.

Cant speak for others, but Im mainly curious how you will react when HF runs in to similar problems and delays. Their schedule is every bit as unbelievable as KnC's was. HF chips havent even made it through the fab yet and already they are running a week behind schedule making their october promise all but impossible. The most tangible thing we've seen from HF so far is renderings and a FPGA simulation, not even running on their own PCB. Ie, about what we've seen from Cointerra which isnt due for another 2-3 months.

Time will tell if you turn out to be a hypocrite or not.

Topic: Swedish ASIC miner company kncminer.com - page 1396. (Read 3050074 times)