Swedish ASIC miner company kncminer.com

[semaster]

eligius pool anyone can connect?

i see errors in my cgminer

use elizium.name

[sbfree]

I was right when I signaled two weeks ago at #kncminer channel that the Saturn seems to work better in higher temperature.
Now many people confirm it.
Going from 40-42 C to 48.5-50 C moves avg from 267 to 274 (at cgminer AND at the pool) and WU from 3850-3900 to 4010.

just plain crazy isn't?

My rig room is at 84 degree F, and my saturn is running the best it's ever ran....

btcg stats:
miner_knc   256.86 GH/s   44,536k (99.83%)

when before w/ extra fans and a/c I was at 220/230GH/s max at btcg....

[sbfree]

I was right when I signaled two weeks ago at #kncminer channel that the Saturn seems to work better in higher temperature.
Now many people confirm it.
Going from 40-42 C to 48.5-50 C moves avg from 267 to 274 (at cgminer AND at the pool) and WU from 3850-3900 to 4010.

just plain crazy isn't?

My rig room is at 84 degree F, and my saturn is running the best it's ever ran....

btcg stats:
miner_knc   256.86 GH/s   44,536k (99.83%)

when before w/ extra fans and a/c I was at 220/230GH/s max at btcg....

Dam, now I am starting to wonder about the earlier firmwares that ran w/ low error rates, but low hashing too, maybe with this "HEATING" solution they might run even better than .97?

[FeedbackLoop]

Boh! You should have posted here like Phoenix did!    I also suspected this in the very first two days and thought to myself I was being silly...   It also takes some colective courage to disable fans.

After disabling the two sink fans for the two worst boards and replacing that flow with two extra case fans placed at a distance, my jupiter went from a 500 GH/s melon to nearly 550 GH/s at pool.  

I'm getting about 10 extra GH/s at KNC's cgminer value and about 50 GH/s extra at pool and using the formula in the previous page (they measure the same) with a corresponding decrease in HW (16% HW to about 7% so far). Getting about 13 cores extra as reported by BertMod (2 to  4 bad ones remaining). Temps 52 in the front boards (the bad ones) and 63 in the back ones which either means the temp sensors are indeed in the DC/DC modules or that I got the bad boards wrong (but a good result...).
 
If these stats change significantly with longer averages I will post here. Otherwise it holds

Firmware 0.95.

[sbfree]

soy, phoenix1969, btc_uzr, (and all others I forget) WHO ARE THE UNOFFICIAL CUSTOMER SUPPORT TEAM for KNC,

I THANK YOU.

I know there many who have contributed their knowledge, expertise, and testing stats, you are included in that thank you.

I know there should be more honorable mentions, so please cont. to mention them in your replies....since this is the only true form of support we have currently. I do thank knc for their efforts as well, as they did surmount that mountain peak.

[FeedbackLoop]

soy, phoenix1969, btc_uzr, (and all others I forget) WHO ARE THE UNOFFICIAL CUSTOMER SUPPORT TEAM for KNC,

I THANK YOU.

I know there many who have contributed their knowledge, expertise, and testing stats, you are included in that thank you.

I know there should be more honorable mentions, so please cont. to mention them in your replies....since this is the only true form of support we have currently. I do thank knc for their efforts as well, as they did surmount that mountain peek.

+1 and sbfree and FiatKiller in that list too

[Phoenix1969]

Thanks for the mention and thank you for thanking me...lolz
you made my morning...Good morning all

just to clarify... I truly believe SOME of the Temperature phenomena is caused by loose heatsink bar screws. Mine are rock solid now, all night, no flux  
This was especially true with the saturn that took a long time to ramp up to speed, and would speed up during the day, and slow down at night. Its the fastest one now, and ramps up very quickly now.

[timmmers]

I was right when I signaled two weeks ago at #kncminer channel that the Saturn seems to work better in higher temperature.
Now many people confirm it.
Going from 40-42 C to 48.5-50 C moves avg from 267 to 274 (at cgminer AND at the pool) and WU from 3850-3900 to 4010.

So, how is my merc running at 140 at 38C ? that equates to 280 for 2 modules.
I think temp increases only help some modules that for some reason aren't in the best condition. Expansion somewhere being the logical reason.

[timmmers]

soy, phoenix1969, btc_uzr, (and all others I forget) WHO ARE THE UNOFFICIAL CUSTOMER SUPPORT TEAM for KNC,

I THANK YOU.

I know there many who have contributed their knowledge, expertise, and testing stats, you are included in that thank you.

I know there should be more honorable mentions, so please cont. to mention them in your replies....since this is the only true form of support we have currently. I do thank knc for their efforts as well, as they did surmount that mountain peak.

You got that right, KNC have been useless. Can you imagine how much trouble people would have had without the threads and discussions on here and their own forum? They can't even answer the phone from what I see, that's so bad after all the money they have taken from customers.

[FeedbackLoop]

I was right when I signaled two weeks ago at #kncminer channel that the Saturn seems to work better in higher temperature.
Now many people confirm it.
Going from 40-42 C to 48.5-50 C moves avg from 267 to 274 (at cgminer AND at the pool) and WU from 3850-3900 to 4010.

So, how is my merc running at 140 at 38C ? that equates to 280 for 2 modules.
I think temp increases only help some modules that for some reason aren't in the best condition. Expansion somewhere being the logical reason.

It seems to help with hardware errors. Either the clock tolerance thing or one of the assorted mechanical/ solder ball stuff. Who knows! Maybe the way to know for sure would be to play with the clock (if possible in any way). Slightly lower clock leading to significantly less errors would indicate the clock  tolerance thing rather than a mechanical/connection thing.

Also make sure you are not just taking those 140 for granted from KNC's cgminer and that you  are either computing those from shares, measuring at pool, or that your hardware error rate is negligible.

[FeedbackLoop]

Mine are rock solid now, all night, no flux  

No flux means all fans off case closed?

[timmmers]

I was right when I signaled two weeks ago at #kncminer channel that the Saturn seems to work better in higher temperature.
Now many people confirm it.
Going from 40-42 C to 48.5-50 C moves avg from 267 to 274 (at cgminer AND at the pool) and WU from 3850-3900 to 4010.

So, how is my merc running at 140 at 38C ? that equates to 280 for 2 modules.
I think temp increases only help some modules that for some reason aren't in the best condition. Expansion somewhere being the logical reason.

It seems to help with hardware errors. Either the clock tolerance thing or one of the assorted mechanical/ solder ball stuff. Who knows! Maybe the way to know for sure would be to play with the clock (if possible in any way). Slightly lower clock leading to significantly less errors would indicate the clock  tolerance thing rather than a mechanical/connection thing.

Also make sure you are not just taking those 140 for granted from KNC's cgminer and that you  are either computing those from shares, measuring at pool, or that your hardware error rate is negligible.

Pool and cg miner agree, within half a GH and have done since I got it the only change being updating to .97 which added 4 Gh , 36C at the moment.
I'd love to know the story behind temps ..but what works is more important I suppose.

[ekiro]

Can KnC ever release another miner that will be profitable?

Looking at the difficulty growth, and the pricy hardware, I see no positive ROI on mining yourself anymore. 

From now on KnC will manufacture their own, mine with them, then sell the hardware off to recover some money.

Don't be the guy buying second hand garbage.

[FeedbackLoop]

Pool and cg miner agree, within half a GH and have done since I got it the only change being updating to .97 which added 4 Gh , 36C at the moment.
I'd love to know the story behind temps ..but what works is more important I suppose.

I imagine your HW is mighty low then, like with all nice performing machines. Also I suspect that no mercuries have been shipped with poor boards since machines get tested after being fully assembled (more bad boards slipping/being masked in jupiters, less in mercuries).

[RHA]

Temperaturewise, it was wise move on KnC side to choose UPS to have the devices delivered with the fans off heatsinks.
Being "off" is a factory state.
Why we made a mistake of remounting them?

[DigginDeep]

So once again keeping the room warm at night has paid off.   When i was letting the room cool off at night  when i woke up and checked my stats on eligius it would be in the Low -mid 400 GH/s  around 430 ,  then i switched to BTCguild to try it out.  witch it was not any better there until i kept the door shut at night.   Before when i woke up the temps would all be in the high 30c   Mostly about 39c   Now two of them run in low 50s  high 40s.  Seems my back two boards are the more troubled ones and need to be more warm to preform better, When the run cooler they have more disabled cores and my error rate shoots over in to the double digits.   Now i don't need to get them as hot as some people on here witch are bringing them in the to 60c range.     But then again i ave not tried this to see if there would be any improvement.   And doubt i will,  I don't mind keeping them running in the low 50s,    But here are what my stats where this morning when waking up this morning.  
https://i.imgur.com/P43JuNI.png
https://i.imgur.com/JLxVDl4.png

Edited....
Now only if i could get it to be stable at the 695gh/s  I could dream.  Wonder if i could ever see it break in to the 700s 
https://i.imgur.com/qmM1VQj.png

[jelin1984]

So final
The Jupiter work better when is harm or when is it cold???

[DigginDeep]

So final
The Jupiter work better when is harm or when is it cold???

Mine without a doubt when its a bit warmer,  Mine dose not need to be super hot.  But  dose not like the cold.     Maybe its just because mine is a Hawaiian Miner lol. 

[sbfree]

So final
The Jupiter work better when is harm or when is it cold???

jelin, I think each saturn/jupiter has it's own individual quirks. So the changes you make will depend on your unit, of course with the intent of more hashing power which ultimately leads to more submitted work/shares to your mining op. WITHOUT destroying your miner. Ask away....or read up....last 50 so pages have some good stuff...

[sbfree]

Wow, my first found block.  On Slush.  A few days ago, I just noticed.  Can't think of any other reason this one Found Block is listed on my profile page.  The block number 265105.

soy, I am approaching 100M submitted (plus others I had on bitminter) shares on BTCG and still no block found....lucky you! and so much for solo mining unless you have terrahashes