i think the voltage setting "--bitburner-voltage" is not correctly implemented.
if i set bitburner-voltage 1300 cgminer shows about 1020mV. if i set it to 1400 (maximum value) cgminer shows 1109mV.
I think when the voltage is correctly set to ~1300mV-1400mV it should be possible to reach 64ghash.
With avalon-setting "45:282" and bitburner-voltage "1400" i reach ~54ghash.
( i used both cgminer 3.5.0 from someone42 and default )
That is correct. The default voltagefor the Bitfury chips is 900mV, but the Bitburner Fury boards support under- and overvolting, so they support a voltage range of 700mv-1100mV. The current firmware emulates an Avalon so it subtracts 300mV from the specified bitburner voltage - if it didn't the default voltage of 1200mV would fry your board! So "--bitburner-voltage 1300" selects a voltage of 1000mV, which is approximately what we see. If you could really set the core voltage to 1300mV you'd almost certainly destroy the chips. roy |
|
|
|
|
EMC down again?
ETA: Oh, just us1 and us2 down. Am I really supposed to add all the servers to my config? I figured it would be counterproductive (and antisocial) to do that as I'd be tying up excessive EMC resources for no benefit. But maybe it really is the right way to configure?
|
|
|
|
Also, for the bitburners you can't have a 10 connector cable on 8 boards.
As long as you use use the connectors at both end of the CANBUS cable, unused connectors shouldn't matter |
|
|
|
There are 2 posts... the first is on bottom. I guess you havent updated the firmware yet either then.
Are the new boards still shipping with out-of-date firmware, then? |
|
|
|
Yeah, that is insane.
I had all three EMC servers plus two backup pools in my config and at one moment (a couple of days ago) cgminer just stopped because there were no pools alive to work with. All of them were under a heavy DDoS.
Yeah, happened to me, too. Having no idea what was going on, I just put it down to my ISP being flaky. I mean, what are the chances of three major pools being down at the same time - must be close to zero, right? roy |
|
|
|
Things are mostly back to normal now. DDoS is peaking at over 12 Gbps and 3 million PPS, pretty insane.
Make sure you have backup pools setup for US1, US2, US3, EU, etc... and things should fail over properly now. Still some work to do on the back end here, but it should be mostly stable.
I want to say dial 911 and report a serious crime in progress. But sadly you'd get arrested for wasting police time. Why is it that the authorities don't give a damn about DDoS....? |
|
|
|
The --bitburner-fury-voltage option will only take effect if a board identifies as BBF (as opposed to BTB) in cgminer. The current firmware will always identify as BTB, although getting it to identify as BBF is as simple as changing the USB product string from "BitBurner" to "BitBurner Fury".
I'm going to hold off on changing the USB product string until pull #499 makes it into an official release. This is because changing the USB product string will stop the boards from being recognised by older cgminer versions.
Yes, this is a bit of a mess. But hopefully in a few weeks time, everyone will be using new firmware and new cgminer which will render all this "add 300 mV to the desired core voltage" business obsolete.
Sorry for spreading misinformation. I saw the code, and I saw burnin's statement that the new firmware would only work with the cgminer fork, and I put two and two together and made five... roy BTW.... Do think of those of us who will shortly be mining with both boards... I do understand wanting to get it into cgminer first, but can we try and get this resolved within days rather than weeks? roy |
|
|
|
The --bitburner-fury-voltage option will only take effect if a board identifies as BBF (as opposed to BTB) in cgminer. The current firmware will always identify as BTB, although getting it to identify as BBF is as simple as changing the USB product string from "BitBurner" to "BitBurner Fury".
I'm going to hold off on changing the USB product string until pull #499 makes it into an official release. This is because changing the USB product string will stop the boards from being recognised by older cgminer versions.
Yes, this is a bit of a mess. But hopefully in a few weeks time, everyone will be using new firmware and new cgminer which will render all this "add 300 mV to the desired core voltage" business obsolete.
Sorry for spreading misinformation. I saw the code, and I saw burnin's statement that the new firmware would only work with the cgminer fork, and I put two and two together and made five... roy |
|
|
|
It's seems the hashrate protection is active. I ask them to add another board rather than € refund but they said boards already sold out for now and my order will ship in a few hours.
That's a shame. I was hoping that 64+ GH/s might still be achievable. Has burnin/cryptx confirmed that the boards will not do 64 GH/s ? roy |
|
|
|
AFAICS there are changes in the way voltage selection is done in this new driver. I think you now specify the true voltage with a new argument, i.e. for standard (not overvolted) operation: --bitburner-fury-voltage 900
But I could be wrong. Hopefully @someone42 or @burnin will confirm soon. Until then I suggest caution. roy EDIT: Please ignore for now - AIUI no change in how you use this version |
|
|
|
But there is no burnin here to discuss this. He is gone like Yifu, working on doing it better next time, perhaps.
I'm in exactly the same situation as you, and I'm likewise disappointed. Although I would say the accusations here (that burnin is mining with customer equipment - with no evidence cited, almost certainly because the poster is making it up) is pretty much par for the course here. And after putting up with these forums for months, I can quite understand why he just can't take it any more. Would still really appreciate a bit more communication.... both on the Avalon and Bitfury boards. Burnin, I've stood buy you, I've defended you. I still think you mean well, but it's hard to continue to believe in you when things don't go according to plan and you fail to give any update.... roy |
|
|
|
You can run multiple cgminer's, one for each stack of boards.
Yes, but if they are all handled by the same driver, how will we specify which boards we want to use? We'll be at the mercy of which order the devices are detected. I probably misunderstand you, but if you have a stack of BitBurners and a stack of FuryBurners, each connected with their own USB cable you would run one cgminer and force it to connect to the BitBurners with cgminer command line arguments applicable to the BitBurners. and run a second cgminer and force it to connect to the FuryBurners with cgminer command line arguments applicable to the FuryBurners. I hope so. I worry that I won't be able to. kano's comment in the pull request seems to suggest that the fury boards still come up as BTB in cgminer, which is just wrong. |
|
|
|
Is there a solution for setting different parameters for XX and Fury boards if we're running both on the same cgminer instance? Particularly with voltage, I *really* don't want to risk getting it wrong and setting the voltage of the board to something other than what I was expecting.
You can run multiple cgminer's, one for each stack of boards. Yes, but if they are all handled by the same driver, how will we specify which boards we want to use? We'll be at the mercy of which order the devices are detected. There must be a better way. Can the driver not tell whether it is an XX or Fury board? I tend to agree with kano that this is the wrong approach. (Note, this is not saying you can't keep the Avalon protocol, if you want to, but the driver needs to know these are Fury boards). I also think the 300mV offset is a recipe for confusion, and is sure to result in fried boards due to people getting confused. roy |
|
|
|
Wow, that sounds great. I guess this means you can put both Bitburner XXs and Furys to the same CAN-bus and control all boards from one cgminer :-?
In principle, yes, since the CAN-bus protocol is also the same. But in practice, you will get HW errors because of how cgminer's Avalon driver is implemented*. Also, frequency settings are broadcast across the CAN-bus and it's unlikely that you would want Avalon and BitFury to run at exactly the same frequency. Maybe in the future, when a well-designed cgminer driver and corresponding firmware is written, you will be able to mix Avalon/BitFury/Cointerra/whatever BitBurner boards. cgminer doesn't control the voltage of any miner to date. All existing miners (and almost certainly this product as well) operate on a fixed voltage. Changing the voltage requires changing a resistor on the board.
Voltage control for BitBurner boards from within cgminer was added in August (see https://github.com/ckolivas/cgminer/commit/b0f4d55be77adfacc3fd3e1bf0540a0252610455). Thanks someone42, that's just the update we were looking for. This all sounds very positive. So do you have somebody on the team working on the cgminer tweaks, so that we can get hashing as soon as the boards arrive?
See https://github.com/ckolivas/cgminer/pull/499. *cgminer's Avalon driver rotates through an array of work units. Mixing Avalon and BitFury will cause the array rotation to happen too fast for the Avalon boards, causing cgminer to forget about work units and resulting in "no matching work" errors. Thanks for the update. Is there a solution for setting different parameters for XX and Fury boards if we're running both on the same cgminer instance? Particularly with voltage, I *really* don't want to risk getting it wrong and setting the voltage of the board to something other than what I was expecting. |
|
|
|
I think that the problem here is that there has never been official cgminer support for bitfury chips, as the developers don't have access to the hardware.
Not quite true. 3.5.0 supports red fury/blue fury. roy |
|
|
|
As I understood this tracking nr. was generated automaticly and doesn't represent the real shipping date ... I'm wrong?
Btw. what software do we need on rasp? ... Any how to ...
Sadly, you're right, the shipping date was not correct in the emails that we received  . They got our hopes up a bit there, but apparently still expect to begin shipping this week - fingers crossed this is the case. We will need a modified version of cgminer I believe, but there's no news about progress on this yet. I really don't want to run a different, modified, version of cgminer just for my Bitburner Fury boards. I want burnin and cryptx to work with ckolivas and kano to integrate support into the official cgminer, as happened with the Bitburner XX boards. I don't understand why this board is going down the 'modified version' route.... roy |
|
|
|
|
I think cryptx has already said that those dates are autogenerated by Fedex when the shipping labels are printed, and that we shouldn't rely on them. But that the units are still on track to ship this week.
Disappointed at the lack of info though, if our units are really days away from shipping. Surely they must have at least rough specs and preliminary software at this stage? I'd much prefer it if they were working with ckolivas and kano to get the driver into cgminer, rather than just promising us some magic version of cgminer will be available on the day...
roy
|
|
|
|
|
Con, did you forget to create a tag? Running head fine though...
Thanks,
roy
|
|
|
|
|
Well, the Ebay one went for $202 or about BTC1.6. I'm really curious as to why the buyer paid that much for it - but I guess I'll never know.
Sure, it's possible that the buyer has some other use for the miner, or the chips within it - but I do wonder whether these things just get bid up by clueless newbies who think it will make them rich and have no understanding of the economics of mining.
I'm not holding my breath, but if anyone wants to sell me a working Single for BTC0.5 (or £40 cash) plus shipping, do let me know. Even better if you're in London and we can do the trade in person.
roy
|
|
|
|
Interesting indeed. Although that doesn't come with a power supply. I wonder what BFL did with all of their trade-in units. You can power it off an ATX PSU with the right cables.... it's what many people did (although I ran mine with the supplied brick). roy |
|
|
|
|
Show Posts
