Sitarow (OP)
Legendary
Offline
Activity: 1792
Merit: 1047
|
|
October 18, 2013, 09:58:00 PM |
|
-------------------- | 0.94 | 0.95 | 0.96 | Average CGminer | 270Gh/s | 274Gh/s | 280Gh/s | Average Pool | 275Gh/s | 263Gh/s | 278Gh/s (link) | Consumption | 485 Watts | 305 Watts | 313 Watts | Temputure | 54 & 62°C | 42 & 47°C | 42 & 47°C | Rejected | 1.5% | 0.22% | 0.38% | HW | 2.81% | 0.79% | 1.08% | WU | 3975/m | 3717/m | 3933/m |
Anyways I'll stick to firmware 0.96. From the beginning I wanted to mine on Eligius and with firmware 0.90 to 0.95 I had the flushwork issue. It seems that 0.96 solved that issue; it's been 24 hours that I'm mining on Eligius. Nice post. I will be testing v0.9.6 soon. However My 3rd Jupiter restarted cgminer 6hrs into my test. I took steps to swap out the intake fans on Jupiter 3 as it performances similar to Jupiter 1. I hope this will solve the cgminer reset.
|
|
|
|
-ck
Legendary
Offline
Activity: 4144
Merit: 1638
Ruu \o/
|
|
October 18, 2013, 10:01:08 PM |
|
I will be testing v0.9.6 soon. However My 3rd Jupiter restarted cgminer 6hrs into my test. I took steps to swap out the intake fans on Jupiter 3 as it performances similar to Jupiter 1. I hope this will solve the cgminer reset.
Do you have a Jupiter 2 then as well? Why did I suddenly think of 'Lost in Space'?
|
Developer/maintainer for cgminer, ckpool/ckproxy, and the -ck kernel 2% Fee Solo mining at solo.ckpool.org -ck
|
|
|
DeathAndTaxes
Donator
Legendary
Offline
Activity: 1218
Merit: 1079
Gerald Davis
|
|
October 18, 2013, 10:01:31 PM |
|
1hr result with firmware v0.9.5 Not much different at this time.
Does v0.9.5 improve or worsen power consumption at the wall?
|
|
|
|
Sitarow (OP)
Legendary
Offline
Activity: 1792
Merit: 1047
|
|
October 18, 2013, 10:21:00 PM |
|
I will be testing v0.9.6 soon. However My 3rd Jupiter restarted cgminer 6hrs into my test. I took steps to swap out the intake fans on Jupiter 3 as it performances similar to Jupiter 1. I hope this will solve the cgminer reset.
Do you have a Jupiter 2 then as well? Why did I suddenly think of 'Lost in Space'? nice.
|
|
|
|
Sitarow (OP)
Legendary
Offline
Activity: 1792
Merit: 1047
|
|
October 18, 2013, 10:23:07 PM Last edit: October 18, 2013, 10:36:36 PM by Sitarow |
|
1hr result with firmware v0.9.5 Not much different at this time.
Does v0.9.5 improve or worsen power consumption at the wall? I have yet to load v0.9.6, So I can't say at this time. However in the next 36hours I will be able to definitively answer that question. Here are the results after 4 hours on firmware v0.9.5 - Ambient Room Temp 20 Replaced Both KnC (Jupiter 1) and KnC3 (Jupiter 3) intake fans have been replaced with high CFM ones. Those systems seem to have the poorest performance of the lot. Results after 3 hoursResults after 4 hours
|
|
|
|
arorts
|
|
October 18, 2013, 10:31:43 PM |
|
-------------------- | 0.94 | 0.95 | 0.96 | Average CGminer | 270Gh/s | 274Gh/s | 280Gh/s | Average Pool | 275Gh/s | 263Gh/s | 278Gh/s (link) | Consumption | 485 Watts | 305 Watts | 313 Watts | Temputure | 54 & 62°C | 42 & 47°C | 42 & 47°C | Rejected | 1.5% | 0.22% | 0.38% | HW | 2.81% | 0.79% | 1.08% | WU | 3975/m | 3717/m | 3933/m |
Anyways I'll stick to firmware 0.96. From the beginning I wanted to mine on Eligius and with firmware 0.90 to 0.95 I had the flushwork issue. It seems that 0.96 solved that issue; it's been 24 hours that I'm mining on Eligius. Thanks for the nice, clean, non-verbal-crap summary!
|
|
|
|
Miz4r
Legendary
Offline
Activity: 1246
Merit: 1000
|
|
October 18, 2013, 11:12:37 PM |
|
Could someone please summarize their findings with the last firmware versions? Something like these data points:
- Firmware version - Power consumption - Hashrate - Temp - Errors, etc/ I updated from 0.95 to 0.96 and I'm sad to say that I regret it. Firmware 0.95512 Gh/s (538 Gh/s avg on Eligius) 41-55 °C 5% HW errors Firmware 0.96476 - 518 Gh/s 41-55 °C 4-12% HW errors Strange how 0.96 seems very unstable for me compared to 0.95. At first after I updated the hashrate was immediately down and had 10% HW errors. So I did a hard reset and after a reboot things seemed fine again with 4% HW errors and 518 Gh/s avg on cgminer, a tiny improvement over 0.95. Then after I restarted cgminer the hashrate started to suck again with a lot of errors. I tried rebooting but it seems to only get worse with 12% HW errors now and only 476 Gh/s avg. I guess I will go back to 0.95 and hopefully get the old stability back even though 5% errors still seems a bit too high.
|
Bitcoin = Gold on steroids
|
|
|
r3animation
|
|
October 18, 2013, 11:51:14 PM |
|
I do think 0.96 recovers from "flushwork" a tiny faster then the previous firmwares. No concrete proof though.
|
|
|
|
DPoS
|
|
October 19, 2013, 12:29:18 AM |
|
it seems they only test their firmwares on one machine before they release them., so much variance
I am getting great results from my one miner that was running .93 for a week ~520 and now is ~550
not all HW errors / cores failing are the same. Some miners will be a clean slate after .96 while others that have bigger issues won't squeeze the same performance out
tl;dr if you have a miner with above par components, use .96 (if you have 8 VRMs of course don't bother going past .94)
|
|
|
|
arorts
|
|
October 19, 2013, 12:38:35 AM |
|
it seems they only test their firmwares on one machine before they release them., so much variance
I am getting great results from my one miner that was running .93 for a week ~520 and now is ~550
not all HW errors / cores failing are the same. Some miners will be a clean slate after .96 while others that have bigger issues won't squeeze the same performance out
tl;dr if you have a miner with above par components, use .96 (if you have 8 VRMs of course don't bother going past .94)
Seems to contradict the findings of Afox. Perhaps some random variance? BTW, does anyone know how to do p2pool with KNCminers? Any details on how that'd be achieved would be appreciated (especially on changes needed on the KNC miner itself, not so much so on what needs to be done in the host PC running the p2pool software).
|
|
|
|
|
DPoS
|
|
October 19, 2013, 12:57:55 AM |
|
yeah that's what my post was about.. some core issues are 'easy fix' and .96 wipes them clean, while other core issues are 'your'e screwed!' and can't do much. Enable cores is part of ,96 install. i guess they gave it out separate so you can run it with other firmware
|
|
|
|
ImI
Legendary
Offline
Activity: 1946
Merit: 1019
|
|
October 19, 2013, 01:06:43 AM |
|
|
|
|
|
r3animation
|
|
October 19, 2013, 01:26:15 AM |
|
yeah that's what my post was about.. some core issues are 'easy fix' and .96 wipes them clean, while other core issues are 'your'e screwed!' and can't do much. Enable cores is part of ,96 install. i guess they gave it out separate so you can run it with other firmware Ah I see. When I upgraded to 0.96 I was wondering why when I downgraded back to 0.94 I had a few more cores enabled (from 41/48 to 47/48). But when I used the EnabledCores flash, it kicked the last core online. But in the end, no hashrate or WU/m increase for me.
|
|
|
|
arorts
|
|
October 19, 2013, 01:51:25 AM |
|
What's EnableCores for? What happens if I don't run that patch? I'm not sure I quite understand...aren't all chips/cores to be re-enabled on any reboot??? Thanks
|
|
|
|
gnar1ta$
Donator
Hero Member
Offline
Activity: 798
Merit: 500
|
|
October 19, 2013, 01:57:08 AM |
|
Seems to contradict the findings of Afox. Perhaps some random variance?
BTW, does anyone know how to do p2pool with KNCminers? Any details on how that'd be achieved would be appreciated (especially on changes needed on the KNC miner itself, not so much so on what needs to be done in the host PC running the p2pool software).
KNC's don't work well with p2pool. The ridiculously slow flush work causes a lot of issues and around 150 Ghash loss on a Jupiter.
|
Losing hundreds of Bitcoins with the best scammers in the business - BFL, Avalon, KNC, HashFast.
|
|
|
arorts
|
|
October 19, 2013, 02:09:36 AM |
|
How do I see the cgminer mining details? All I know is how to access the web interface.... which doesn't provide actual hashing rate
|
|
|
|
DPoS
|
|
October 19, 2013, 02:31:57 AM |
|
What's EnableCores for? What happens if I don't run that patch? I'm not sure I quite understand...aren't all chips/cores to be re-enabled on any reboot??? Thanks that's the murky part we are trying to figure out I'd like to know all the ways that a 'core' goes down. Why a whole die of 48 cores just shuts off. And also why some dies have 10-30 cores off all the time and then others that get 'cleaned up' easy and back to perfect
|
|
|
|
sbfree
|
|
October 19, 2013, 03:03:51 AM |
|
so after 30hrs or so, here is where I am at with my Saturn on .95 firmware....should I update it to .96
miner_knc 233.65 GH/s 5534080 (99.82%) 7936 / 31616 / 2176
temps are 47 slot 1, and 40.5c slot 5
|
|
|
|
arorts
|
|
October 19, 2013, 04:39:50 AM |
|
Can someone please explain how to see the cgminer outcome?? I use screen -r but it won't take me the process name like with bit fury. How do you see the outcome of that process?
|
|
|
|
|