Author

Topic: cgminer 3.8.5 - what is acceptable number of HW errors? (Read 7910 times)

sr. member
Activity: 364
Merit: 250
Not found a definitive answer to this myself, so working with target of <1% HW errors with my AntMiners.

Using the formula:

HW / (A+R+HW)

For you, this would be:

508 / ( 5063 + 12 + 508 ) = 0.09 or 9% which is too high, look at your cooling, i'm hitting <1% with my AntMiners near a cold window. Got some USB fans on the way.

Good luck!

Cfx.
member
Activity: 75
Merit: 10
I have this with the AntMiners. Does this look ok? Or too many hardware errors?



 (5s):17.69G (avg):14.13Gh/s | A:5063  R:12  HW:508  WU:179.4/m
 ST: 2  SS: 0  NB: 7  LW: 10592  GF: 0  RF: 0
 Connected to stratum.bitcoin.cz diff 12 with stratum as user xx.
 Block: bd07ef04...  Diff:1.42G  Started: [15:45:25]  Best share: 1.83
----------------------------------------------------------------------
 [P]ool management [ S ]ettings [D]isplay options [Q]uit
 AMU 0:                | 2.386G/1.731Gh/s | A:583 R: 0 HW:59 WU:22.1/m
 AMU 1:                | 2.493G/1.709Gh/s | A:552 R: 0 HW:66 WU:21.5/m
 AMU 2:                | 2.473G/1.538Gh/s | A:511 R: 0 HW:48 WU:19.7/m
 AMU 3:                | 2.184G/1.706Gh/s | A:555 R: 0 HW:60 WU:21.6/m
 AMU 4:                | 2.232G/1.683Gh/s | A:603 R: 0 HW:52 WU:21.5/m
 AMU 5:                | 1.742G/1.658Gh/s | A:567 R:12 HW:62 WU:20.8/m
 AMU 6:                | 1.827G/1.620Gh/s | A:513 R: 0 HW:62 WU:20.3/m
 AMU 7:                | 3.259G/1.679Gh/s | A:597 R: 0 HW:52 WU:21.4/m
 AMU 8:                | 1.985G/1.729Gh/s | A:606 R: 0 HW:47 WU:22.3/m
sr. member
Activity: 364
Merit: 250
I have 4 rigs some which have been mining the same pool for over a month with no hardware error's.... when I was 1st setting them up if I got error's I would drop the --gpu-engine buy 5 till the error's would stop.

I wonder if ASIC hardware is different? I'm just wondering what is an acceptable number. Zero would be great. But, how bad is 1%? What is the effect of 1% HW errors? Is it any effect? Or MASSIVE effect? I have no idea.

If someone knows what is bad % of HW then I would be very grateful for your thoughts!

Cfx.
newbie
Activity: 15
Merit: 0
I have 4 rigs some which have been mining the same pool for over a month with no hardware error's.... when I was 1st setting them up if I got error's I would drop the --gpu-engine buy 5 till the error's would stop.
sr. member
Activity: 364
Merit: 250
What do you mean by a little longer? How much time does it take to get 30 HW
What happens if you underclock a little bit

It's about 90 minutes. Those 30 HW errors are the total between all three AntMiner devices, so they have avg 10 each.
sr. member
Activity: 381
Merit: 251
What do you mean by a little longer? How much time does it take to get 30 HW
What happens if you underclock a little bit
sr. member
Activity: 364
Merit: 250
Thank you, I do not seem to be able to eliminate them completely, I get the same even when not overclocking.

After running for a little while longer, here are the figures:

(5s) 3.796G
(avg) 5.483Ghs
A: 5084
R: 0
HW: 30
WU: 75.7/m

A formula I found elsewhere suggests that <1% of HW/(HW+A+R) is acceptable and expected. Is this correct?

Thanks!

Cfx.
newbie
Activity: 8
Merit: 0
HW are not positive, limiting them closer to zero is always better.
sr. member
Activity: 364
Merit: 250
Hello!

I am running cgminer 3.8.5 with 3 x AntMiner U1 at 0881 frequency (slightly overclocked).

What is an acceptable number of HW errors?

I've seen some formulas but i'm not quite sure if they apply, so here are some stats after running for 30 minutes...

(5s) 3.796G
(avg) 5.483Ghs
A: 1664
R: 0
HW: 19
WU: 75.7/m

Is 19 HW errors too many or within an acceptable range? Or do I need to run it for longer to get a larger data set?

Thank you in advance for your help!

Cfx.
Jump to: