Hacking The KNC Firmware: Overclocking

[elasticband]

Can help via team viewer if you wish. Just PM your details.

The student becomes the teacher  LOL

Well done!!

As embarrassing as it is, i was just being a fool. The keyboard I was using is set to a different configuration from what the keys display. I was never pushing :   silly me

thanks for the help though!

[1714056518]

1714056518

[1714056518]

1714056518

[elasticband]

just had success reviving my dead die. currently receiving 50w power.


Lowering it back down to around 30-40w now and it's still running.

If anyboady would like me to have a play with thiers i can help via teamviewer

[joeventura]

Can help via team viewer if you wish. Just PM your details.

The student becomes the teacher  LOL

Well done!!

As embarrassing as it is, i was just being a fool. The keyboard I was using is set to a different configuration from what the keys display. I was never pushing :   silly me

thanks for the help though!

Just happy it got fixed!!!

Happy to help!!

[Cablez]

For those that are unsure of their way around vi as they are new to it here are the basic commands.  http://www.cs.colostate.edu/helpdocs/vi.html

It was a help to me as I had only used nano prior.

[CYPER]

Someone might be interested in my communication with CoolerMaster:

Question:

Maximum safe temperature for PCI-E cables: I have a Cooler Master V850 to power a specific hardware that is using 4
PCI-E cables with a total DC power consumption of around 700W.
So each PCI-E cable is drawing around 175W or a current of 14.5A.
My main concern is that the cables near the PSU reach a temperature of
45-46 degrees. Then again the markings on the cables says 80 degrees,
which I assume is the maximum safe temperature?

So is a temperature of 46 degrees on the outside of the hottest cable safe?

Thank you.

Answer:

Apologies for the late reply.
That is correct, all cables are rated 80 degrees Celsius for testing. You can withstand temparature up to 80 degrees Celsius with any issues.

Kind regards,

Michiel Bos

[mruiter]

I Just let 2 dies of my coolest boards run on 251
They survived the first day. Put them back for the night as i'm not near the miner at that time.

Amps on the VRM where 65 an 64
Temp was 76 degrees

Speed 45Ghash on the dies i enabled.

Tomorrow i will try to lower volt.

But correct me if i'm wrong.... The chips need a certain amount of watts to opperate.

If i lower the volt the amps wil go up.
If i upper the volts the amps wil go down.

Anyone tested and can confirm this ?

[sickpig]

No, at least in my expirience.

Every time I've increased voltage I've had higher Amps and Watts.

One quest though: how did you check hashrate at die level?

[elasticband]

any success doing 8 VRM's?

[mruiter]

No, at least in my expirience.

Every time I've increased voltage I've had higher Amps and Watts.

One quest though: how did you check hashrate at die level?

Bfgminer had the average of the whole board
Just math - the half of the default / 2 is the single die

[elasticband]

got my 8VRM runnign stable around 600 GH

[CYPER]

642GH/s with 8VRMs here.
Been running like that for almost 2 weeks.

[elasticband]

642GH/s with 8VRMs here.
Been running like that for almost 2 weeks.

how? I was a bit premature, running fine for about 10mins then just stared getting HW increase and cores shutting off.
Only Running 1F1 - 800MHZ

[CYPER]

642GH/s with 8VRMs here.
Been running like that for almost 2 weeks.

how? I was a bit premature, running fine for about 10mins then just stared getting HW increase and cores shutting off.
Only Running 1F1 - 800MHZ

Sheer luck - that is my best performing October Jupiter.
Using only 211, so 850Mhz.
HW errors are really low too: 0.56%

[tolip_wen]

I've asked EK-Waterblocks http://www.ekwb.com/ to design a waterblock for the KnC hardware, my aim is for a higher stable overclock;

http://thinkcell.ekwb.com/idea/knc-miner-waterblock/

As you can see from the thread the KnC response has not been positive so far. 

I'd like to know anyone else is interested in this small project,

Target = 1Th 6 module KnC

I like the way you think, but think it all the way through.

Spending lots on cooling at this point in time may not be worthwhile.
RAM blocks are expensive and you need 2 per module.

Large air cooled aluminum heatsinks might be more cost effective.

FWIW: It cost me lots more than the cost of waterblocks for a WC system.
I justified it, you may also, just please make the decision based on all the facts and set expectations accordingly.

Fun  _IS_ a good reason!


YMMV

[elasticband]

On a merc I am currently running 900mhz - 231  with very little problem

[tolip_wen]

I have one die running 950MHz  (251)  for over a month.
Most others are way slower though.

Amps and average temp for that die?

the 950MHz die is pulling 59A 47W 
(those are the highest readings I ever see so slightly lower average)

I don't run the whole ASIC at that speed just one die.
The neighbors are ~50A 40W (average but all similar)
Directly adjacent are @ 39W and opposite corner is @ 41W
41.5W average per die for the ASIC
My personal do not exceed is less than 45W per die AVERAGE per ASIC
I'm still under 190W at the wall per ASIC (platinum PSU's at the sweet spot of 50%)
1.2W at the wall per Gh/s

Temps is a guesstimate at best.
IR temp from backside of running module is 60C for that die.
IR temp for that VRM is 73C measured on the bottom of the PCB!
There is a part in the VRM that is much hotter!
Temps were cross checked with thermocouple and measured at location of many via's conducting heat thru PCB

The watercooling system runs about 30C (I trade some temp for lower noise)
I have parts to improve the delta C but don't want the downtime yet as I don't see much to gain from current settings based on experience to date.

The LM75 on the PCB says ~ 50-55C depending on draftiness in cold house.

as usual
YMMV

[merv77]

Can overclocked Jupiter with 4vrm or not??

yes, it was originally OC'ed for the 4VRM boards. Many thanks to tolip_wen and others here.

and thanks to joeventura for a very nicely written tutorial to follow.

If you don't know how to SSH into your Miner with PuTTY then there is going to have to be some payment for me to write you a tutorial with screen shots.

Here are the full instructions, hopefully you know how to SSH into your Miner with PuTTY

DISCLAIMER: Not my fault if you blow it up! Smiley


Ok first make sure your Miner is updated with the .99 Firmware (tuning)

Then with Putty SSH into your miner

login with your root ID
and your password

type:

screen -r

You will see cgminer running.

Hit Q to stop cgminer

then type this at the command prompt:

vi /etc/init.d/cgminer.sh

Use the cursor keys on your keyboard to scroll down till you see this code:

Look for the line that says:

cmd=$(printf "0x86,0x%02X,0x01,0xF1" $c) 


NOTE: some of the characters at the end (especially the F1) may be different.

The key is the line with the 0x86, when you see that you are on the right line.


change this:

cmd=$(printf "0x86,0x%02X,0x01,0xF1" $c)

to this:


cmd=$(printf "0x86,0x%02X,0x02,0x11" $c)



Hit escape when finished. Then hit the : key, then x to save and exit


relaunch cgminer with the command line:


/etc/init.d/cgminer.sh restart


You should then see cgminer and you should be hashing between 150-154 per module. (Saturn should be 300-310GH)



NOTE:
If you don't know how to use VI then you can screw things up

here is a shortcut,
when you move the cursor over the 1 in "0x01" that you want to change to a 2, do this:
hit r
then hit 2

and it should Replace the 1 with a 2
then do the same with the F
hit r
then hit 1

Hit escape

then hit the COLON which is ":"

then type

wq

then

<enter>

[idee2013]

Just to share my settings/performance:
- October unit (Asic 1 = 4 VRM, other 3 are 8 VRM but only 4 active)
- Firmware 0.99-tune
- Clocksetting on 231, no other firmware changes.
- Harware mod: Lowered the Asicboard-fans (more airflow over the board), added 2 fans at the outlet (so now push + pull) and slightly lowered fan speed to reduce the noise (80%).

Images
Status page --> 676 GH/s / 9131 WU
Advanced page --> 2.5V SPI and 256kHz frequency, slightly modded volts (higher, as required for the increased clock), not completely fine-tuned.
CGminer --> Error rate used to be 3%@560GH/s, since I lowered the volts a bit after the initial increase this increased to just below 4%. Fine with me as it lowered the temps 5-10 degrees an decreased power to 'around' 40 W per DC instead of closer to 50 W.
Pool stat  --> 650ghs average
Power consumption in bertmod used to be 453 W, now it is 651 W (43% increase in consumption for 18/19% performance increase). With voltage finetuning the power consumption should be able to come down a bit, now it is all pretty much on the safe side (safe as in: higher voltage as required to avoid that cores get shut down).

Have done exaktly the same (231) with my october jupiter (3x8vrms and 1x4vrms), but gave 2,5V SPI and 256000hz and gave more volts at the asics.
I am reaching now about 670 Gh/s at Eligius pool 12hours avg. Power consumption now is 743 Watts (50-55amps per asic). Before i had 555Gh/s avg and about 593 Watts.
PSU PC and Cooling  MK 3 1200 (platinum).

[jelin1984]

how i can do that?
\
Then with Putty SSH into your miner

login with your root ID
and your password

i go at putty login as admin
then put the password
but access denied

why??

[idee2013]

how i can do that?
\
Then with Putty SSH into your miner

login with your root ID
and your password

i go at putty login as admin
then put the password
but access denied

why??

because you are not the admin...you have to login with root. KNC has the admin data