Was there ever a .05 firmware update?
Is there anybody out there? Nod if you can hear me...
Nod, there was. |
|
|
|
I received one of these devices courtesy of LordTheron, thanks! I've committed some changes to cgminer master git to detect the number of active chips on the fly. It will now change to hexfury while running even if it reports the wrong number of chips to start with (this one reports 5). Display output: 22: HXF 2 : 47.7C | 10.49G / 10.19Gh/s WU: 146.6/m
API output: [STATS22] =>
(
[STATS] => 22
[ID] => HXF2
[Elapsed] => 206276
[Calls] => 0
[Wait] => 0.000000
[Max] => 0.000000
[Min] => 99999999.000000
[Version] => 1.2
[Revision] => 1
[Chips] => 5
[NonceRate] => 1.353960
[NoMatchingWork] => 0
[Temperature] => 45.000000
[Max DeciTemp] => 451
[Clock] => 54
[Core0 hwerror] => 4
[Core0 jobs] => 137
[Core0 submits] => 98
[Core1 hwerror] => 7
[Core1 jobs] => 137
[Core1 submits] => 92
[Core2 hwerror] => 2
[Core2 jobs] => 134
[Core2 submits] => 84
[Core3 hwerror] => 5
[Core3 jobs] => 133
[Core3 submits] => 93
[Core4 hwerror] => 0
[Core4 jobs] => 133
[Core4 submits] => 96
[USB Pipe] => 0
[USB Delay] => r0 0.000000 w0 0.000000
[USB tmo] => 1989 0
)
EDIT: Actually extended results only ever seem to indicate 5 chips as well. Average after 24 hours: 10.47GH |
|
|
|
|
As far as I'm aware there is nothing to make the U2 faster, only that it has better cooling. Cgminer can't discriminate a U1 from a U2 so it's still being told the same default frequency of 200. The heatsink is larger on the U2 meaning it can easily run at 2GHs but only if you actually tell it --anu-freq 250
|
|
|
|
5% is still okay in my book, but that's about the limit. LiteGuardian does 2% PPS, and it's not doing any hybrid.
The reserves a pool needs to handle the swings of PPS are too much for it to be a consideration. Combine that with the block withholding attack risk that can drive the pool bankrupt it's not worth it. I like mmpool with or without PPS, I'm just giving my opinion on PPS in general. Oh, well yea, I agreed with you 100% with PPS in general. Personally, I think PPS rates should be trailing; based on the pool's average rate at which they find blocks for one Difficulty Term. I would first set my break-even, minimum PPS %, then choose an arbitrary number + % as the first control term. After the Network Difficulty adjusts, get the avg. Time between blocks on network and between blocks found in pool. I would see how the pools avg. Hashrate changed +/- then adjust the PPS n% higher, or lower. Of course you would need an upper bound limit pool miners would agree too, but adjusting the PPS between let's say 1.25% and 5.05% could average out nicely for Pool managers and Devoted Miners. My math is 'Pseudo Math' I would need time to write an actual formula, but that's my idea. PPS with a variable fee based on luck. Interesting. This is a lousy idea and just as hoppable as the old proportional type payment in disguise. If you have a week of good luck all your miners will jump across while the pps rates are high. As soon as luck is bad they'll all run away. You can bet your bottom dollar as soon as the hordes are aware you're offering this they'll abuse it. It still leaves you wide open for a block withhold attack too. |
|
|
|
That is good news. Thanks for the fast work on that.  The new version of cgminer also allows you to specify the bits on the command line as a complement to your overclocking attempts. --bxf-bits <arg> Set max BXF/HXF bits for overclocking (default: 54)
Wasn't something similar already there for the bifuries? How different is this? Isn't default 50 for them? No, the bifuries only had target temperature control and were always set to a maximum of 54. This allows you to set the maximum now in concert with the target temperature control. There are a number of different bitfury chip based usb devices supported in cgminer and they all have different clocking command line options. |
|
|
|
|
For those of us who get a lot of forum notifications, and the mods that get reports need to have a simple way to start/stop all forum notifications temporarily for example for when we are away and do not wish to come home to hundreds of emails, and to reject all attempts to send private messages to let people know we're currently away and not accepting messages.
|
|
|
|
Sure but they've not tried submitting it for inclusion. You seem to be missing how the process works despite me describing it a few times. |
|
|
|
Any chance there'll be support for Technobit's HEX16A2 boards?
If they submit their code in a way consistent with other driver submissions, or provide one of us devs hardware to help massage the code to get into mainline cgminer then sure. There is a patch for these boards: https://bitcointalk.org/index.php?topic=355031.0Is it possible to integrate it into the main code? No, we do not go out of our way to scrounge up patches from elsewhere and make them work in the master cgminer code. I already described the process above. |
|
|
|
I would like to respectfully request you post your hardware driver's custom modifications to the cgminer source code as per the agreed upon license when you chose to develop on cgminer.
You must make an open letter to the manufacture. I believe Lee is just a reseller. Thanks, who is the manufacturer? |
|
|
|
|
I would like to respectfully request you post your hardware driver's custom modifications to the cgminer source code as per the agreed upon license when you chose to develop on cgminer.
|
|
|
|
So many people in the web are just assuming and guessing I was reading thru the readme and made the jump in the cold water by getting the newest version where other people say dont get the new one stay with 3.1.1 because everybody is using that one. Please let me know if u want me to add or remove more from the config file. Thank you very much to both of you and I keep u posted about the usb hub results
That ancient version has no support for antminers anyway. It thinks they're block erupters and probably runs them at ~500MH. |
|
|
|
|
The rate at which you're bumping your own thread is annoying when no one else is posting on this thread. Please just edit your existing posts.
|
|
|
|
Any chance there'll be support for Technobit's HEX16A2 boards?
If they submit their code in a way consistent with other driver submissions, or provide one of us devs hardware to help massage the code to get into mainline cgminer then sure. |
|
|
|
|
Source code? And does it have BTC support?
|
|
|
|
Hi jbcheng Fortunately, the author of cgminer is a Mod of this forum so he can fully clarifly the licensing. Two A few questions: - will your driver and/or mining software be open source?
- In the case where GPL code is used, will you respect the license and post your modifications? You know, unlike your competitors (I'm looking at you bitmain and lketc
 ). - will you make your circuit schematic docs open source - or at least available as PDF?
If the answer to the above questions is 'no' how exactly do you expect customers to maintain/repair your products? As per the GPL license you must supply source code and all modifications. This is mandatory.Here are some relevant snippets from https://www.gnu.org/copyleft/gpl.html: As always I will give the benefit of the doubt and assume there is some misunderstanding on your part about the license you have agreed to when choosing cgminer and deciding to develop your driver based on its code, or that you have misunderstood what is being asked of you. So I will remind you as simply as possible as a summary of that document above (also as per the file COPYING included with the cgminer source and binaries). The license you have agreed to by using cgminer means: You are free to modify and distribute binaries based on cgminer provided you distribute the source with your modifications, or a link to a publicly available download site that hosts the source, or provide the option to send a copy of the source on media for no more than the cost of the media, along with a copy of the license included in all binaries and source distributed or a link to the COPYING file. If you do not do this within 30 days, then your license to distribute the binaries will be revoked and you will be open to prosecution. |
|
|
|
mmpool, you could implement a referall system. So if you refer someone you get 5% of their earnings or 1 of every 1000 shares or something
Who would use a pool with 6.5% fee? No one... I suppose you could get 5% of the 1.5% fee the pool charges, but seriously, how much money we talking? I've been promoting this pool for over 2 years... are you saying I should be earning something for that? If so, I'm cool with that, heh. You really think no one would use a pool with a 6.5% fee? I can't describe how funny that is. Check out the PPS fee on BTC Guild and come back. BTCguild abolished PPS due to the great risk it posed. A large fee is appropriate. |
|
|
|
I'm willing to do some cgminer testing with the latest stable kernel (mainline 3.14), albeit I'll have to make some fake devices to test the cgminer API on the latest kernel (I don't own any ASICs). On that note, I'm making my own GNU/Linux distribution - I've compiled the kernel (3.14) and bash (4.3) in its own fs/partition and I'm creating the linux directory hierarchy by hand. Basically my policy is everything mainline - everything latest stable. (Reason for creating this is if I want to run 3.14 none of the distros want to know about me, and I want to see the driver code for my hardware and muck around with it.) And yes, Ted Ts'o is a bit of a smarty pants on the linux kernel mailing list, which I've found out personally. (Not sure how he treated you, conman) The choice and features of each linux kernel pretty much has precisely zero impact on cgminer so there's nothing to test. |
|
|
|
|
Unless you have some idea of how long the password is and what some of it is likely to be, then if it's a long password, potentially you will never find the password, even if that hardware tried to brute force for as long as the universe has existed.
|
|
|
|
Perhaps you want this: https://bitcointalk.org/index.php?topic=557866.0And this: https://bitcointalk.org/index.php?topic=557991.0And this: 2013 Feb, after discussion, mining.subscribe was extended to take 2 arguments, so now it's mining.subscribe(String useragent, sessionid) where sessionid is optional, but if provided must match a previous session's mining.notify subscription id. If the server allows resuming a session, it will give you back the same subscription id and extranonce information.
cgminer handles it. No idea bout the rest. |
|
|
|
what does this mean?:
AMU 57 SendWork usb write err:(-9) LIBUSB_ERROR_PIPE
I'm getting hashrate to build up to 2/3rds the way then half of it shuts off and it stats to reload again until the cycle repeats. Cannot get everything on and mining in harmony.
It means there has been a fatal usb communication error and cgminer drops the device to allow it to restart again. You probably don't have enough power or dodgy connections etc. Its not that simple though - I've have them all up and running before but for some reason I see that after a little that a string of them would turn on (an entire hub wont seem to connect, I have 9 hubs, of those 3 are daisy chained to a 4th. Doesn't matter how complex your setup is, the error means the same thing. Often when something fails it can take out/reset a whole hub since the communication channel is shared. |
|
|
|
|
Show Posts
