HF seems to submit a bankrupt request:
http://ia902509.us.archive.org/13/items/gov.uscourts.canb.522575/gov.uscourts.canb.522575.1.0.pdf
Topic: HashFast Sierra's Owners Thread - page 3. (Read 23757 times)
but in reality i shouldn't have to overclock as it should be hashing at 1.2 TH/s minimum?
While i was tuning the boards to find their max. stable hashing rate, cgminer occasionally gets ZOMBIEs and doesn't recover. But once I lay off the OCing, they can hash stable and if one should fail, cgminer tents to be able to recover it automatically after 60 secs of 0 share submitted.
Interesting to know, i don't think i waited 60 seconds to find out if it would auto kick in.
also, are you running cgminer in linux? if so, which one would you recommend or doesn't it really matter?
also, is the sierra being damaged in anyway by setting the clock rate to 590? as i thought i read somewhere 612 was a sweet spot, i didn't fancy pushing it that far.
but in reality i shouldn't have to overclock as it should be hashing at 1.2 TH/s minimum?
in reality, BTC H/w manufacturers tent to estimate the product's hashing speeds and power consumptions. Real world figures can differ dramatically.
My Sierra was stable on 604Mhz and hashing ~1.3THs, but once I took it apart. I could get it to about 1.4THs. Each board behaves differently. See http://imgur.com/a/94wo7#28 for my last screen shot of cgminer and pics of my Sierra. You will notice different clock speed for each board in the Sierra.
Here's a snap shot of cgminer at the time of posting:
Code:
cgminer version 4.3.3 - Started: [2014-05-15 17:29:56]
----------------------------------------------------------------------------------------------------
(5s):787.5G (1m):1.207T (5m):1.309T (15m):1.332T (avg):1.469Th/s
A:17328394 R:20517 HW:17106 WU:20522.3/m | ST: 5 SS: 0 NB: 97 LW: 20093194 GF: 0 RF: 0
Connected to us.ozco.in diff 461 with stratum as user quiveringgibbage.larry
Block: 4200d5e2... Diff:8.85G Started: [09:02:45] Best share: 58.2M
----------------------------------------------------------------------------------------------------
[U]SB management [P]ool management [S]ettings [D]isplay options [Q]uit
0: HFB AL : 550MHz 75C 10% 0.79V | 102.6G / 429.1Gh/s WU: 5994.9/m A:5058818 R:5550 HW:14752
1: HFB KEN : 656MHz 85C 48% 0.78V | 299.2G / 484.7Gh/s WU: 6771.5/m A:5711371 R:6063 HW: 1529
2: HFB ROBERTA : 666MHz 85C 72% 0.78V | 453.3G / 555.2Gh/s WU: 7755.9/m A:6558205 R:8904 HW: 825
----------------------------------------------------------------------------------------------------
[2014-05-16 09:02:58] Accepted 41b0b091 Diff 998/461 HFB 0 pool 0
----------------------------------------------------------------------------------------------------
(5s):787.5G (1m):1.207T (5m):1.309T (15m):1.332T (avg):1.469Th/s
A:17328394 R:20517 HW:17106 WU:20522.3/m | ST: 5 SS: 0 NB: 97 LW: 20093194 GF: 0 RF: 0
Connected to us.ozco.in diff 461 with stratum as user quiveringgibbage.larry
Block: 4200d5e2... Diff:8.85G Started: [09:02:45] Best share: 58.2M
----------------------------------------------------------------------------------------------------
[U]SB management [P]ool management [S]ettings [D]isplay options [Q]uit
0: HFB AL : 550MHz 75C 10% 0.79V | 102.6G / 429.1Gh/s WU: 5994.9/m A:5058818 R:5550 HW:14752
1: HFB KEN : 656MHz 85C 48% 0.78V | 299.2G / 484.7Gh/s WU: 6771.5/m A:5711371 R:6063 HW: 1529
2: HFB ROBERTA : 666MHz 85C 72% 0.78V | 453.3G / 555.2Gh/s WU: 7755.9/m A:6558205 R:8904 HW: 825
----------------------------------------------------------------------------------------------------
[2014-05-16 09:02:58] Accepted 41b0b091 Diff 998/461 HFB 0 pool 0
While i was tuning the boards to find their max. stable hashing rate, cgminer occasionally gets ZOMBIEs and doesn't recover. But once I lay off the OCing, they can hash stable and if one should fail, cgminer tents to be able to recover it automatically after 60 secs of 0 share submitted.
Hash wild and Hash Fast,
QG
In a perfect world yes. It's not ck's fault though.
Are you seeing squelching messages in the cgminer output?
Are you seeing squelching messages in the cgminer output?
i didn't say it was ck's fault, im just asking if there's a solution to this problem if anyone else has come across it as well.
I have been told to use a program called cgwatcher to monitor cgminer & if it zombies it will auto restart cgminer. haven't tested that theory yet, but before i do i thought i would ask if there was any other solution.
When you say squelching messages, what do you mean? i think the only message i have seen is something to do with no response from head then it say's zombie, sorry i didn't really pay much attention, was just concerned the sierra had knocked off & i rapidly restarted everything to get it back up and running
It seems that when the miner is pushed too far, specific cores are "squelched". At least that is the message that cgminer reports. I've not seen one in some while now.
Mine is running comfortably at 604 and has been for a few weeks averaging 1.261Th during that time. I've converted to a .conf file, but the original .bat file simply had this preceding pool information:
"cgminer.exe --hfa-hash-clock 604 --hfa-fan 85"
Your temps seem to be under control, so it seems like you just got a bit of a dud.
ok thank you for your help, do you think running it in linux will make a difference? and il try the .conf file setup as well.
I don't know why i expected any different from hashfast, waited this long and got a dud lol
il also test each board to see if i can locate it down to one board that seems to be causing the issue.
again, thanks for the info, at least now i know its something to do with the sierra
In a perfect world yes. It's not ck's fault though.
Are you seeing squelching messages in the cgminer output?
Are you seeing squelching messages in the cgminer output?
i didn't say it was ck's fault, im just asking if there's a solution to this problem if anyone else has come across it as well.
I have been told to use a program called cgwatcher to monitor cgminer & if it zombies it will auto restart cgminer. haven't tested that theory yet, but before i do i thought i would ask if there was any other solution.
When you say squelching messages, what do you mean? i think the only message i have seen is something to do with no response from head then it say's zombie, sorry i didn't really pay much attention, was just concerned the sierra had knocked off & i rapidly restarted everything to get it back up and running
It seems that when the miner is pushed too far, specific cores are "squelched". At least that is the message that cgminer reports. I've not seen one in some while now.
Mine is running comfortably at 604 and has been for a few weeks averaging 1.261Th during that time. I've converted to a .conf file, but the original .bat file simply had this preceding pool information:
"cgminer.exe --hfa-hash-clock 604 --hfa-fan 85"
Your temps seem to be under control, so it seems like you just got a bit of a dud.
Edit: windows + cgminer requires zadig.
ok, this is what cgminer says
HFS 0: Failed to reset after write failure, disabling
HFS 0 failure, disabling!
HFS 0: Failed to get data associated with header
USB init, open device failed, err -12, you need to install a WinUSB driver for - HFA device 2:2-i1
See README.txt file included for help
hashfast detect (2:2) failed to initialise (incorrect device?)
has anyone come across this before?
HFS 0: Failed to reset after write failure, disabling
HFS 0 failure, disabling!
HFS 0: Failed to get data associated with header
USB init, open device failed, err -12, you need to install a WinUSB driver for - HFA device 2:2-i1
See README.txt file included for help
hashfast detect (2:2) failed to initialise (incorrect device?)
has anyone come across this before?
In a perfect world yes. It's not ck's fault though.
Are you seeing squelching messages in the cgminer output?
i didn't say it was ck's fault, im just asking if there's a solution to this problem if anyone else has come across it as well.Are you seeing squelching messages in the cgminer output?
I have been told to use a program called cgwatcher to monitor cgminer & if it zombies it will auto restart cgminer. haven't tested that theory yet, but before i do i thought i would ask if there was any other solution.
When you say squelching messages, what do you mean? i think the only message i have seen is something to do with no response from head then it say's zombie, sorry i didn't really pay much attention, was just concerned the sierra had knocked off & i rapidly restarted everything to get it back up and running
In a perfect world yes. It's not ck's fault though.
Are you seeing squelching messages in the cgminer output?
Are you seeing squelching messages in the cgminer output?
if i remove --hfa-hash-clock 590 will it clock at its normal clock rate (550) only reason why i over clocked it is cause it barely passes 1TH/s without overclocking
Yes I understand why people overclock... but not all hardware is created equal and it's called overclocking for a reason.
if i remove --hfa-hash-clock 590 will it clock at its normal clock rate (550) only reason why i over clocked it is cause it barely passes 1TH/s without overclocking
Yes I understand why people overclock... but not all hardware is created equal and it's called overclocking for a reason. so I have received my sierra but at least once or twice a day it zombies 
and doesn't kick back in until I restart cgminer
does anyone know why this happens?
Are you overclocking it (too much)? and doesn't kick back in until I restart cgminer does anyone know why this happens?
clock 590, current hashrate of 1.22TH/s, temp 77'c 0.79v
if i remove --hfa-hash-clock 590 will it clock at its normal clock rate (550) only reason why i over clocked it is cause it barely passes 1TH/s without overclocking
so I have received my sierra but at least once or twice a day it zombies and doesn't kick back in until I restart cgminer
does anyone know why this happens?
Are you overclocking it (too much)? and doesn't kick back in until I restart cgminer does anyone know why this happens?
clock 590, current hashrate of 1.22TH/s, temp 77'c 0.79v
so I have received my sierra but at least once or twice a day it zombies and doesn't kick back in until I restart cgminer
does anyone know why this happens?
Are you overclocking it (too much)? and doesn't kick back in until I restart cgminer does anyone know why this happens?
clock 590, current hashrate of 1.22TH/s, temp 77'c 0.79v
and i have to restart cgminer twice, first time it just hangs then second time it kicks in straight away
so I have received my sierra but at least once or twice a day it zombies and doesn't kick back in until I restart cgminer
does anyone know why this happens?
Are you overclocking it (too much)? and doesn't kick back in until I restart cgminer does anyone know why this happens?
so I have received my sierra but at least once or twice a day it zombies and doesn't kick back in until I restart cgminer
does anyone know why this happens?
and doesn't kick back in until I restart cgminer does anyone know why this happens?
cgminer + winblowz + usb asic = zadig. Are you using zadig?
thanks! got it working. that did the trick.
another not so good news on HF:
http://www.coindesk.com/hashfast-cuts-50-of-staff-denies-bankruptcy-rumors/
http://www.coindesk.com/hashfast-cuts-50-of-staff-denies-bankruptcy-rumors/
cgminer has issues with reported speeds, so the results are inaccurate there. Each ASIC with all the cores enabled can do 768 Bitcoin hashes per clock, so if you are 700MHz your absolute maximum hashrate is 538GH/s.
i c i c.. as I did try the 700MHz again overnight to try to get KEN up to 800+GHs and yeah it only hashed around 500GHs. Will need more testing and will display results.I did get up to a total of 1.5ths with the pool when all 3 boards were hashing at the same time:
Code:
[19:37:05] <+QG> .w larry
[19:37:06] <@ozbot> This round (23d 19h 24m) QG.larry has 601854762 shares (2900 stales; 0.0%), avg 1539708 MH/s
[19:37:06] <@ozbot> This round (23d 19h 24m) QG.larry has 601854762 shares (2900 stales; 0.0%), avg 1539708 MH/s
Cheers,
QG
so i've started tuning each board's Clock speed and the results are somewhat surprising:
My original setting and hash rate with the Sierra was clocked at 604MHz hashing a total of 1.3THs (~433GHs was expected from each board)
Now I am hashing each of the 3 boards individually with it's own USB cable:
Yes, that's a single Sierra board (aka BabyJet) hashing at over 800GHs (wall measured to be ~610w for this setting). So my total hashing speed went from 1.3ths to ~1.7THs. I am still in the early stages of tuning cgminer for the best hashing rates and I need to repeat these test results. But looks very promising to hash those board individually.
Anyone else planning on hashing their boards individually? I'd love to compare results.
Pics can be found here: http://imgur.com/a/94wo7
Cheers,
QG
PS: Units are for sale (located in Sydney, Australia).
cgminer has issues with reported speeds, so the results are inaccurate there. Each ASIC with all the cores enabled can do 768 Bitcoin hashes per clock, so if you are 700MHz your absolute maximum hashrate is 538GH/s. My original setting and hash rate with the Sierra was clocked at 604MHz hashing a total of 1.3THs (~433GHs was expected from each board)
Now I am hashing each of the 3 boards individually with it's own USB cable:
Code:
0: HFB KEN : 700MHz 85C 81% 0.79V | 725.4G / 860.0Gh/s WU: 12021.7/m A: 675079 R: 437 HW: 311
1: HFB ROBERTA : 700MHz 85C 61% 0.78V | 674.6G / 513.1Gh/s WU: 7170.9/m A: 270783 R: 366 HW: 73
3: HFB AL : 583MHz 82C 5% 0.80V | 537.3G / 587.6Gh/s WU: 8209.1/m A: 564308 R: 688 HW: 626
1: HFB ROBERTA : 700MHz 85C 61% 0.78V | 674.6G / 513.1Gh/s WU: 7170.9/m A: 270783 R: 366 HW: 73
3: HFB AL : 583MHz 82C 5% 0.80V | 537.3G / 587.6Gh/s WU: 8209.1/m A: 564308 R: 688 HW: 626
Yes, that's a single Sierra board (aka BabyJet) hashing at over 800GHs (wall measured to be ~610w for this setting). So my total hashing speed went from 1.3ths to ~1.7THs. I am still in the early stages of tuning cgminer for the best hashing rates and I need to repeat these test results. But looks very promising to hash those board individually.
Anyone else planning on hashing their boards individually? I'd love to compare results.
Pics can be found here: http://imgur.com/a/94wo7
Cheers,
QG
PS: Units are for sale (located in Sydney, Australia).
so i've started tuning each board's Clock speed and the results are somewhat surprising:
My original setting and hash rate with the Sierra was clocked at 604MHz hashing a total of 1.3THs (~433GHs was expected from each board)
Now I am hashing each of the 3 boards individually with it's own USB cable:
Yes, that's a single Sierra board (aka BabyJet) hashing at over 800GHs (wall measured to be ~610w for this setting). So my total hashing speed went from 1.3ths to ~1.7THs. I am still in the early stages of tuning cgminer for the best hashing rates and I need to repeat these test results. But looks very promising to hash those board individually.
Anyone else planning on hashing their boards individually? I'd love to compare results.
Pics can be found here: http://imgur.com/a/94wo7
Cheers,
QG
PS: Units are for sale (located in Sydney, Australia).
My original setting and hash rate with the Sierra was clocked at 604MHz hashing a total of 1.3THs (~433GHs was expected from each board)
Now I am hashing each of the 3 boards individually with it's own USB cable:
Code:
0: HFB KEN : 700MHz 85C 81% 0.79V | 725.4G / 860.0Gh/s WU: 12021.7/m A: 675079 R: 437 HW: 311
1: HFB ROBERTA : 700MHz 85C 61% 0.78V | 674.6G / 513.1Gh/s WU: 7170.9/m A: 270783 R: 366 HW: 73
3: HFB AL : 583MHz 82C 5% 0.80V | 537.3G / 587.6Gh/s WU: 8209.1/m A: 564308 R: 688 HW: 626
1: HFB ROBERTA : 700MHz 85C 61% 0.78V | 674.6G / 513.1Gh/s WU: 7170.9/m A: 270783 R: 366 HW: 73
3: HFB AL : 583MHz 82C 5% 0.80V | 537.3G / 587.6Gh/s WU: 8209.1/m A: 564308 R: 688 HW: 626
Yes, that's a single Sierra board (aka BabyJet) hashing at over 800GHs (wall measured to be ~610w for this setting). So my total hashing speed went from 1.3ths to ~1.7THs. I am still in the early stages of tuning cgminer for the best hashing rates and I need to repeat these test results. But looks very promising to hash those board individually.
Anyone else planning on hashing their boards individually? I'd love to compare results.
Pics can be found here: http://imgur.com/a/94wo7
Cheers,
QG
PS: Units are for sale (located in Sydney, Australia).
Jump to: