Swedish ASIC miner company kncminer.com

[adrd]

https://www.kncminer.com/news/news-60
now is official

typo in link in news http://www.kncminer.com/userfiiles/file/kncminer-0.98.bin
should be http://www.kncminer.com/userfiles/file/kncminer-0.98.bin

[chrono030]

I can't seem to download the file, it failed. - Edit - solution posted before me

[xyzzy099]

I can't seem to download the file, it failed. - Edit - solution posted before me

Use the link on the support page.

[soapmodem]

Early notes on 0.98...

- cgminer updated to 3.6.6
- getting fewer hardware errors
- no change in flushwork frequency
- performance roughly 2% higher
- temps slightly higher

[carlotech]

any reports on .98 firmwares?

[rolling]

any reports on .98 firmwares?

I am testing on my poor performing units and so far looks great!

The disabled cores now seem to have a smaller negative effect on overall hashing power.

[chrono030]

any reports on .98 firmwares?

So far, my hashrate in CG miner has gone from 540 gh/s to 550

 cgminer version 3.6.6 - Started: [2013-10-29 14:29:59]
--------------------------------------------------------------------------------
 (5s):627.6G (avg):566.4Gh/s | A:27648  R:512  HW:594  WU:7846.7/m
 ST: 2  SS: 0  NB: 1  LW: 29933  GF: 0  RF: 0

Take this all with a grain of salt, it has been hashing for about 20 seconds - havent applied bertmod quite yet.

[edgar]

is bertmod & enablecores the same thing?   

i have 2 separate .bin files but neither 'seem' to change anything.

[sickpig]

any reports on .98 firmwares?

So far, my hashrate in CG miner has gone from 540 gh/s to 550

 cgminer version 3.6.6 - Started: [2013-10-29 14:29:59]
--------------------------------------------------------------------------------
 (5s):627.6G (avg):566.4Gh/s | A:27648  R:512  HW:594  WU:7846.7/m
 ST: 2  SS: 0  NB: 1  LW: 29933  GF: 0  RF: 0

Take this all with a grain of salt, it has been hashing for about 20 seconds - havent applied bertmod quite yet.

what was your hash rate with your prev fw?

[jelin1984]

yes the same

[rolling]

is bertmod & enablecores the same thing?

NO!

[soapmodem]

Careful with this one, guys.. monitor your temps. My machines are running hot. Each chip about 2-4 degrees higher on 0.9.8.

Rolling back to 0.97 for now.

[sickpig]

is bertmod & enablecores the same thing?

not at all

[sickpig]

Careful with this one, guys.. monitor your temps. My machines are running hot. Each chip about 2-4 degrees higher on 0.9.8.

Rolling back to 0.97 for now.

absolute values? before and after?

[chrono030]

is bertmod & enablecores the same thing?

No - Bertmod is used in order to display more details in summary form including disabled cores and hardware error %'s.  KNC seem to have approved it for use, as it is not permanent.

[chrono030]

.97 Firmware
ASIC slot #1 39.5 ℃
ASIC slot #2 27.0 ℃
ASIC slot #3 43.0 ℃
ASIC slot #4 37.0 ℃

Avg. Hash Rate 544 Gh/s
Difficulty Accepted 21,176,320
Difficulty Rejected 64,512 (0.3 %)
Hardware Errors 1,952,367 (7.8 %)
WU 7,874
Difficulty Stale 8,192
Network Blocks 350
Pool Rejected 0.3 %
Pool Stale 0
Best Share 9,295,152
Found Blocks 0

.98 firmware
ASIC slot #1   41.0 ℃
ASIC slot #2   28.0 ℃
ASIC slot #3   45.5 ℃
ASIC slot #4   38.5 ℃

Last Checked   Tue Oct 29 14:45:20 UTC 2013
Avg. Hash Rate   563 Gh/s
Difficulty Accepted   104,960
Difficulty Rejected   512 (0.4 %)
Hardware Errors   2,378 (1.9 %)
WU   7,847
Difficulty Stale   0
Network Blocks   2
Pool Rejected   0.5 %
Pool Stale   0
Best Share   479,722
Found Blocks   0

I'll agree with the comments above, this firmware does run hotter than .97.  With the increased hashrate I suppose that is to be expected.  

[oyeTorry]

Careful with this one, guys.. monitor your temps. My machines are running hot. Each chip about 2-4 degrees higher on 0.9.8.

As long as it does not affect HW errors %, 2-4 degrees higher doesnt matter, right?

[sickpig]

.97 Firmware
ASIC slot #1 39.5 ℃
ASIC slot #2 27.0 ℃
ASIC slot #3 43.0 ℃
ASIC slot #4 37.0 ℃

.98 firmware
ASIC slot #1   41.0 ℃
ASIC slot #2   28.0 ℃
ASIC slot #3   45.5 ℃
ASIC slot #4   38.5 ℃

I'll agree with the comments above, this firmware does run hotter than .97.  With the increased hashrate I suppose that is to be expected.  

it's harmless imho, this is my jup with 0.95:

 KNC 0: 51.0C | 135.6/132.1/126.1Gh/s | A:11149 R:160+0(.46%) HW:160159/4.8%
 KNC 1: 50.5C | 134.8/130.0/126.4Gh/s | A:11284 R:126+0(.46%) HW: 82402/2.5%
 KNC 2: 47.5C | 135.3/132.0/128.0Gh/s | A:11467 R:109+0(.24%) HW: 91937/2.8%
 KNC 3: 40.0C | 142.7/139.5/137.0Gh/s | A:12301 R:132+0(.28%) HW: 33157/.96


let say the I would have used "warmer" rather than "hotter"

[greenbtc]

Careful with this one, guys.. monitor your temps. My machines are running hot. Each chip about 2-4 degrees higher on 0.9.8.

As long as it does not affect HW errors %, 2-4 degrees higher doesnt matter, right?

Some people have been doing testing around these chips and they actually have an "optimal" temperature that they hash at. It's not always just "the coolest" = "highest hashrate."

Some people were even creating shrouds around the units to get them to a higher temp.

As long as the HW error % doesn't go up like the guy said, the higher temps are not a reason to "roll back" to .97

[davewr2013]

As long as you are under 100C I would not worry too much about an increase.

Maybe you should be concerned if it hits 80-90C -- maybe not.

In any case below 70C -- quit worrying take the extra hashes.
https://www.kncminer.com/pages/support

I will upgrade after I finish a short outing -- will give additional info tonight.