By the way: anyone having a dumping CGminer or FW_1.0 ? My system is stopping and restarting one to three times daily, this I didn't have with the previous rc9.
I wonder if I'm the only one?
This is what my 'cat /var/log/monitordcdc.log' is telling (it dumped and restarted on 08:14:18 after it ran in one go from 15:38:20 the day before):
[2014-07-13 15:38:20] Restarting cgminer
[2014-07-14 08:12:47] Die 5-3 came DOWN
[2014-07-14 08:13:09] waas re-run
[2014-07-14 08:13:55] Die 5-3 came UP
[2014-07-14 08:14:17] waas re-run
Topic: Hacking The KNC Firmware: Overclocking - page 5. (Read 144362 times)
OK, thanks for the info... I will need to study more, and back to the drawing board.....
I DID send you (unsolicited) detailed instructions via "Message" yesterday.
Please forgive me for bieng so bold but,
I think you are the closest to success.
@ Others,
I don't make a habit of it (messages or mentioning same in threads)
When it's ready it is trivial to educate/distribute widely.
I don't want to participate in cooking cubes of the unprepared miner that WILL click anything for 0.0001 more GH.
Full disclosure
I invited Elenelen to distribute when ready also, no caveats, no secrets, no leetness involved.
Anyone can do it now, few are motivated to go thru the hoops yet.
YMMV
OOPS !! I didn't know I had messages waiting for me (11 of them, never noticed by me)..... such a newbie I am
Thank you Tolip ! I will work on it.
I followed suggestions from another venue about recompiling modified source code.
The bar is higher than editing a script and running it currently.
The RESULTS for me are less than ideal to date so I don't elaborate here.
WIthout cooling and power improvements, not much there to get.
If I were helpful like you are,
explained monster heatsink in this thread,
with pictures,
I could start checking for messages,
yesterday.
It's in this thread is also a cruel correct answer.
It's cryptic tidbit time! Fun with powers of 2.
Code:
root@Neptune:~# cat /var/run/.waas_curfreq
-1 -1 -1 -1 512 512 528 512 512 512 544 512 512 512 544 512 512 512 512 512 512 512 512 512
False is a possibility too.
(BTW I claim it's real)
Don't assume nuthin!
Thanks to KnCMiner,
It's open source,
legal yet fun!
YMMV
This is interesting if it actually works, you have always been able to set whatever you want but waas itself is closed source in the Nov Jupiters, while you can pass it anything you like it will not actually honor the settings.. I can't imagine they open sourced the code for waas itself for the Neptune but if they did then it's an easy fix..
On my firmware I need to set things manually by flipping bits because it's out of the range of waas, I actually intercept the speed (if anyone has looked at the code) and if waas can handle it I let it, anything over the Octover and November waas ranges automatically gets set by flipping bits..
In any case if they now allow waas to do it directly it will be much easier..
well i gotta say this...I may not agree at what knc company does to their customers or how they run things but their boards n chips for oct. n nov jup are holding up like troopers at a moderately high oc setting/overvolting for months...i'm actually surprised they are still running n not blowing up hehe..i'm actually thinking about turning them up even more to see if i can get them to blow just for sheer entertainment...but risk of fire out weighs that i think hehe
I got a hot pepper that is forced to solo at ever increasing clocks with similar fate
well i gotta say this...I may not agree at what knc company does to their customers or how they run things but their boards n chips for oct. n nov jup are holding up like troopers at a moderately high oc setting/overvolting for months...i'm actually surprised they are still running n not blowing up hehe..i'm actually thinking about turning them up even more to see if i can get them to blow just for sheer entertainment...but risk of fire out weighs that i think hehe
OK, thanks for the info... I will need to study more, and back to the drawing board.....
I DID send you (unsolicited) detailed instructions via "Message" yesterday.
Please forgive me for bieng so bold but,
I think you are the closest to success.
@ Others,
I don't make a habit of it (messages or mentioning same in threads)
When it's ready it is trivial to educate/distribute widely.
I don't want to participate in cooking cubes of the unprepared miner that WILL click anything for 0.0001 more GH.
Full disclosure
I invited Elenelen to distribute when ready also, no caveats, no secrets, no leetness involved.
Anyone can do it now, few are motivated to go thru the hoops yet.
YMMV
I took a look at the waas source on github and there is a check for max frequency set at 500. You'd either need to edit and build this utility (and I wouldn't be surprised if that's what tolip did) or you can probably use the same trick that worked for the Jupiters and reset the clock before starting cgminer.
Exactly what I did.
Currently, it MUST be stated.
The solution is worse than the question.
It does not work on the fly.
It's slow.
Just setting to 500 is better or as good.
If you are @ 500 now,
and are not getting 3.6 minus % HW and % rejects already,
no point!
I am not there yet with higher clock.
EDIT I tried the Hoopiter approach.
For shure no success here.
I looked elsewhere.
The SPI has changed, things have moved.
Good option if someone wants to reverse engineer the changes.
My first stab was a bust.
YMMV
OK, thanks for the info... I will need to study more, and back to the drawing board.....
I took a look at the waas source on github and there is a check for max frequency set at 500. You'd either need to edit and build this utility (and I wouldn't be surprised if that's what tolip did) or you can probably use the same trick that worked for the Jupiters and reset the clock before starting cgminer.
Exactly what I did.
Currently, it MUST be stated.
The solution is worse than the question.
It does not work on the fly.
It's slow.
Just setting to 500 is better or as good.
If you are @ 500 now,
and are not getting 3.6 minus % HW and % rejects already,
no point!
I am not there yet with higher clock.
EDIT I tried the Hoopiter approach.
For shure no success here.
I looked elsewhere.
The SPI has changed, things have moved.
Good option if someone wants to reverse engineer the changes.
My first stab was a bust.
YMMV
I must have done something wrong? As I cannot set it to 525.
As by your posts, I modified 'vi /config/advanced.conf' then did 'waas'.
Doing this I have success in modifying the voltage, and I can modify the frequency as well... but when ever I try 525 instead of 475 or 500 the 'waas' will not accept it (500 is maximum).
How did you get 525 ?
I followed suggestions from another venue about recompiling modified source code.
The bar is higher than editing a script and running it currently.
The RESULTS for me are less than ideal to date so I don't elaborate here.
WIthout cooling and power improvements, not much there to get.
If I were helpful like you are,
explained monster heatsink in this thread,
with pictures,
I could start checking for messages,
yesterday.
It's in this thread is also a cruel correct answer.
It's cryptic tidbit time! Fun with powers of 2.
Code:
root@Neptune:~# cat /var/run/.waas_curfreq
-1 -1 -1 -1 512 512 528 512 512 512 544 512 512 512 544 512 512 512 512 512 512 512 512 512
False is a possibility too.
(BTW I claim it's real)
Don't assume nuthin!
Thanks to KnCMiner,
It's open source,
legal yet fun!
YMMV
The teaser
Code:
root@Neptune:~# cat /var/run/.waas_curfreq
-1 -1 -1 -1 500 500 550 500 500 500 525 525 500 500 550 500 475 500 525 500 475 525 500 525
I must have done something wrong? As I cannot set it to 525.
As by your posts, I modified 'vi /config/advanced.conf' then did 'waas'.
Doing this I have success in modifying the voltage, and I can modify the frequency as well... but when ever I try 525 instead of 475 or 500 the 'waas' will not accept it (500 is maximum).
How did you get 525 ?
I took a look at the waas source on github and there is a check for max frequency set at 500. You'd either need to edit and build this utility (and I wouldn't be surprised if that's what tolip did) or you can probably use the same trick that worked for the Jupiters and reset the clock before starting cgminer.
The teaser
Code:
root@Neptune:~# cat /var/run/.waas_curfreq
-1 -1 -1 -1 500 500 550 500 500 500 525 525 500 500 550 500 475 500 525 500 475 525 500 525
I must have done something wrong? As I cannot set it to 525.
As by your posts, I modified 'vi /config/advanced.conf' then did 'waas'.
Doing this I have success in modifying the voltage, and I can modify the frequency as well... but when ever I try 525 instead of 475 or 500 the 'waas' will not accept it (500 is maximum).
How did you get 525 ?
The teaser
Code:
root@Neptune:~# cat /var/run/.waas_curfreq
-1 -1 -1 -1 500 500 550 500 500 500 525 525 500 500 550 500 475 500 525 500 475 525 500 525
That's a great teaser, had to do a double-take. *sigh* heat is my major issue.
I just woke up, and saw that cgminer 4.4.1 dumped again (and restarted automatically)... looks that there is some bug in fw-1.0 or cgminer_4.4.1 (I didn't have crashes with rc9 and 4.4.0, while hardware and Volt-tuning are all the same).
Next week, I hope to have some more time and will check that log-file (thanks for the info).
Did you have Status or Advanced web page running?
ssh connection?
miner.php using the API?
I have stopped using the gui completely when possible.
Not even running 'screen -r'
I'll see how long going native lasts.
I get hashrate with
'api-cgminer devs'
Get VRM details and clock volt temp stuff same as Advanced
'waas -g all-asic-info'
This also identifies choices for filtering.
To filter out Frequency
'waas -g all-asic-info | grep Freq'
or a different way to see clock only
'cat /var/volatile/run/.waas_curfreq'
The teaser
Code:
root@Neptune:~# cat /var/run/.waas_curfreq
-1 -1 -1 -1 500 500 550 500 500 500 525 525 500 500 550 500 475 500 525 500 475 525 500 525
Find sleeping VRM
'waas -g all-asic-info | grep Iout'
Change things manually instead of Advanced page
'vi /config/advanced.conf'
apply button replacement, use after manual edit of /config/advanced.conf
'waas'
WARNING this may result just like clicking on Apply in the advanced page with some VRM off.
It's random same as Advanced.
YMMV
EDIT
I found a way to wake up sleeping VRM
WITHOUT POWER CYCLE!
it's faster!
stop cgminer with
'/etc/init.d/cgminer.sh stop'
if Advanced page is open wait for numbers to drop (20 seconds?)
To just wake up sleepers no changes
'waas'
If you need to apply changes from Advanced do it with cgminer stopped
If you manually edited /config/advanced.conf, run
'waas'
Wait 20-30 seconds for waas to finish.
restart cgminer with
'/etc/init.d/cgminer.sh restart'
look to see if VRM are happy!
Thank you for teaching me all those Linux commands.... Coming from windows I'm still a newbie...
Just tried "waas"... wow very interesting info....
and nice to see you are already above 500 Mhz with some VRM's!
52a is max for the vrm before OCP kicks in n kills the vrm...so with that limitation you don't have much oc'in room...i'd say you'll be lucky to get 550mhz if that outta these units unless you super cool the vrms lmao...but that wouldn't be worth it
I determined experimentally 50A or trouble spreads.
Look closely at my freq above
Rock steady @ 3.54T with warm VRMs, "lmao".
YMMV
Well, I run one of my VRM's on 51A-52A and it doesn't switch of.... but that's maximum for sure (I tried a bit more Volt, and it stops)... of course you need very low Temps !! I'm running @ 3.58T with all on 500 Mhz. I like to try to go higher, but then for sure the Volt needs to go lower, which will cause more HW-errors... I wonder if Mhz/Volt-tuning will give us substantially more hashes than 3.6T
I just woke up, and saw that cgminer 4.4.1 dumped again (and restarted automatically)... looks that there is some bug in fw-1.0 or cgminer_4.4.1 (I didn't have crashes with rc9 and 4.4.0, while hardware and Volt-tuning are all the same).
Next week, I hope to have some more time and will check that log-file (thanks for the info).
Did you have Status or Advanced web page running?
ssh connection?
miner.php using the API?
I have stopped using the gui completely when possible.
Not even running 'screen -r'
I'll see how long going native lasts.
No, I don't use status/adv web-page, but I do run PuTTY with screen -dr.
Currently CGminer is running for 15 hours straight, but I expect some dump later (which isn't that problematic, as it auto-restarts).
yes, someone in picture your mining rigs already put a neppy in novac stuff and the vrms were cool 50c but he said if he tried to push the vrms past 52a they would shut off...hmmm maybe its unique to each vrm and you just have to play around with setting i guess...idk i don't have one to play with...ahh so maybe you can oc more then...dunno but good luck
I did see that Neptune board in 7100; looks suspiciously similar to the giant lab beaker I intend to use.
I will work on uber cooling the rig; hope to see some custom overclock-able firmware by the time I'm done.
52a is max for the vrm before OCP kicks in n kills the vrm...so with that limitation you don't have much oc'in room...i'd say you'll be lucky to get 550mhz if that outta these units unless you super cool the vrms lmao...but that wouldn't be worth it
What is OCP? (please excuse my ignorance)
How super cool are we talking? Pretty sure the Novec would still keep them at 65 if you were running 100A through them/
EDIT:
Over Current Protection right?
yes, someone in picture your mining rigs already put a neppy in novac stuff and the vrms were cool 50c but he said if he tried to push the vrms past 52a they would shut off...hmmm maybe its unique to each vrm and you just have to play around with setting i guess...idk i don't have one to play with...ahh so maybe you can oc more then...dunno but good luck
I'm telling VRM to send lower voltage to ASIC with -0.0439.
The VRM is tired, he likes it.
ASIC likes it too.
If they likes it I likes it.
Look closely @ the default voltage
-0.0366.
IT IS A NEGATIVE NUMBER
-0.0439
is a MORE NEGATIVE NUMBER
The way it is set up, there is a default voltage that is close.
Then add or subtract from that to get what you want.
The VRM is tired, he likes it.
ASIC likes it too.
If they likes it I likes it.
Look closely @ the default voltage
-0.0366.
IT IS A NEGATIVE NUMBER
-0.0439
is a MORE NEGATIVE NUMBER
The way it is set up, there is a default voltage that is close.
Then add or subtract from that to get what you want.
I figured it was 0.77v or whatever from VRM one side and then -0.0366v the other so -0.0439 was actually pulling more voltage across the chip. But you are clearly correct in that it simply modified the default voltage and the other side of the chip stays at 0v.
Thanks
-0.0439 means you're pulling a higher voltage across the chip or not?
I'm telling VRM to send lower voltage to ASIC with -0.0439.
The VRM is tired, he likes it.
ASIC likes it too.
If they likes it I likes it.
Look closely @ the default voltage
-0.0366.
IT IS A NEGATIVE NUMBER
-0.0439
is a MORE NEGATIVE NUMBER
The way it is set up, there is a default voltage that is close.
Then add or subtract from that to get what you want.
Try higher/lower and see what works for you.
YMMV
52a is max for the vrm before OCP kicks in n kills the vrm...so with that limitation you don't have much oc'in room...i'd say you'll be lucky to get 550mhz if that outta these units unless you super cool the vrms lmao...but that wouldn't be worth it
I determined experimentally 50A or trouble spreads.
Look closely at my freq above
Rock steady @ 3.54T with warm VRMs, "lmao".
YMMV
Jump to: