 Show Posts
|
|
Pages: « 1 2 3 [4]
|
|
Does anyone know if there is a way to set a primary and a backup pool in the start command?
|
|
|
|
|
question for you all. One of my guys "upgraded" an Antminer R4 batch 6 to the latest Antminer S9 firmware. That upgrade went through just fine....and now it won't hash, of course.
It won't let me replace the Antminer S9 firmware with the correct Antminer R4 firmware.
Anyone know where in the R4 filesystem there might be a file or setting to change to get it back on track?
|
|
|
|
|
One of the Zclassic devs looked at it, said it appeared to be a large solo miner that started by using Sammy's proxy, then got the dev fee removed, and kept mining to the same address.
My guess it's someone with a big GPU mining operation looking to move over from Ethereum or Monero mining to Zclassic. That's just good news right there.
Or maybe someone who has a large AI GPU farm that's not keeping as busy as he wants to with AI stuff, and wants to get some Zclassic by mining.
|
|
|
|
Colleagues, I observe the mining of this coin 3 days. I am concerned a very strange phenomenon. The miner with the address t1actVSAhVnM2ididm5d3fPwDTZBAey3bbw periodically finds in a row a large number of blocks in a short time. This happens about once every 1.5 hours. Here is a screenshot of the Block explorer where the miner found 11 blocks in a row in 2 minutes. How is this possible?  http://cs1.imwox.com/o/4e/dd/4edd5677e5cd6bd28b12373f1a14a3f0.jpgSame here: http://cs1.imwox.com/o/43/a4/43a4d135b5225238f2adbf88dc810a6b.pngIts repeating and thats a lot mined coins. Look on ballance on that wallet. Devs can you answer us on this issue? It looks like some form of mining attack. Thanks It was brought to attention to the dev's and this has been look into I've heard rumors that some of the bigger Bitcoin mining pools find the answer to the block, then don't publish it right away. They take that answer, and start all their miners mining the next block. By doing this, it can give them 30-90 second head start. Then they float the answer to the first block, and do the same thing with the next block. Is it possible someone is doing the same thing at a private Zclassic mining pool? That would explain getting a number of blocks in row. I don't know how the big mining pools operate, and this is just a theory. But it would explain what we are observing so far. As to what to do about it, miners need to join more distributed pools, and zclassic needs more miners. This comes with time, and possibly with Ethereum and monero miners moving over to Zcash and Zclassic. |
|
|
|
you will ALSO receive a 1:1 amount of ZEN. So, if you have 1000 ZCL before the hardfork, you will have 1000 ZCL after the hardfork plus 1000 ZEN. ZCL will continue to exist.
Wow, it is great! Now I get it, thank's all for clearing  Just keep in mind that no one knows what ZCL will be worth after the harfork, or how much ZEN will be worth when in comes into existence. If everyone just dumps their coins, both ZCL and ZEN may be close to worthless, at least for some period of time. Will have to think about what mining strategy to use on this. After the fork, I will probably point half my machines at Zen and leave half running on Zclassic. Zclassic will have a lower hash rate, so I should be able to pick up more blocks on it if other people move their miners over to Zen. |
|
|
|
|
Woo-hoo! I'm loving 1.7!! It's brought some cards back from the dead that 1.6.2 could not make work. And it's faster! Love it!
|
|
|
|
|
I bought a few Giants directly from Baikal. Excellent customer service, fast shipping. The miners are running great!
I had bought a Baikal from Asicminermarket earlier and that one stopped working, none of the LED's light up. It looks a little different than the ones I got directly from Baikal. Right now I am not sure where to get support on that one.
|
|
|
|
|
My T9's are running like champs. I was able to order a few more this last time. Stayed away from the S9's.
It will be interesting to see how they handle the heat this summer in Atlanta.
|
|
|
|
|
Based on my experience in the IT industry setting up data networks for people, you should charge a significant premium above PC repair prices.
If in the USA, charge $175/hr or $1500 per day. But try not to charge an hourly rate, instead quote a project based rate, and in the project, include:
Discovery
Scoping
Design
Ordering
Installation and Configuration
Testing
Operations Training
Basic support
So whatever your estimate of how long the project is going to take, triple it. If you think it will take 2 days, make it 6. Charge a $1500 daily rate times 6 days, and quote $9,000. If the client wants to negotiate the price down, do it by changing the scope of the project.
Scope, timeframe, and price are all linked together.
Make sure you set a milestone or date for when the support from the initial installation changes to paid support. Then defined a Service Level Agreement for paid support, and set the expectation that response time will be within 4 hours, not instantaneous.
|
|
|
|
I got an R4 of the latest batch on Jan 13, 2017. It ran for 3 days and one board quit hashing. I am noticing there is a flaw in the cooling algorithm. Sometimes after a reboot it runs with both boards at 80 degrees and the fan fluctuates periodically. Other times the fan runs slow and one board runs at 100. There is definitely something wrong with the fan controller in my unit. My basement room is humidity controlled and never gets above 68F. I am so bummed out because I was so excited and spent $1200 to get it. I also read the website many times and some pages tell you to ship back the board and other pages tell you after Sept 2016 to return the entire unit. I am waiting to see if I can ship the board since I already removed it. It is pretty obvious by reading the kernel log which board is not hashing. My guess is they know the defective boards exit and they want you to ship the entire unit in hopes that will make people reluctant and will just run the 50% hash rate and save them some repairs. I plan to play like a good player but if they don't fix my board I am going to be quite vocal about it. After reading this forum they are not QC these boards very well so buyer beware.
I shipped a S9 autofreq board back and they told me I needed to send the entire S9 back. I did that, and set up a second ticket, and in the ticket asked them to put the board in the box. You probably would want to ship the entire R4 back to them, else they might do the same thing to you as they did to me. |
|
|
|
I have switched to Linux ( ubuntu 16.04.1) but have been experiencing my screen not being recognized between switching rigs ( plug in hdmi in one rig and plugging back into same gpu I started with on another rig ) also on startup I have seen my screen not be recognized. Here's what I did: install ubuntu Reboot Sudo apt update Sudo apt dist-upgrade Reboot Sudo install new amd 16.50 Reboot Run miner, see gains  Move to next rig repeat Came back to the first rig and screen isn't recognized. Missing anything? I know i'm not answering exactly what you expect, but how does that matter ? Can't you use ssh to admin your rigs ? Or the issue isn't about administration but rather about the miner not working because the screen is not recognized ? => My setup is 100% headless, no fake hdmi or anything and it works like a charm. Agreed - headless is the way to go. There's an issue with the way xorg works in the new versions of Ubuntu. I have found that installing 3 packages makes it work well. This is what I do: install xorg xserver-xorg-legacy xserver-xorg-video-dummy Edit the config file for xorg-legacy: /etc/X11/Xwrapper.config delete the last line: allowed_users=console insert at end of file: allowed_users=anybody needs_root_rights=yes then set up the xserver-xorg-video-dummy There's a bit more to getting the headless fully working in a screen, but this gives you a start. I don't want to keep flogging my build guide, you can probably find a link in a previous post if you want. It took me a while to figure it out, but once you get full headless going it's wonderful. |
|
|
|
Just upgraded my linux boxes to 1.6.0. Looks like a solid and reliable build so far. Getting an increase across the board. At -i 7 cards are doing:
RX470 8G - 288
Rx480 8G -301
Fury X - 452
Fury Pro - 470
R9 390 - 365
How long have you been using your -i 7 ? mostly out of curiosity, i'm wondering whether that's just as stable as -i 6 on stock cards. But more importantly, have you also compared your Watt/Sol ratio between -i 6 and -i7 ? I've been using -i 7 for about a week now. Good question on Watt/Sol ratio. I had not thought of that! I have wattmeters and will be at the warehouse tomorrow so I will run a test. Hey blackops! Just chatted with you on the other forum. I've looked at power vs intensity, and at least at the "auto" intensity and higher, there is virtually no difference in power consumption, at least on my Nanos. I see a small rate gain with higher intensity, but I favor stability over a couple sols/s. With 120+ GPUs, it doesn't take very many GPUs hanging to wipe out any hash rate gain, for me anyway. Yes, that's a lot of GPU's, I see why you are set up for stability. -i 7 is pretty stable so far, but you're right on the downside of a card hanging, especially if it is at night. It's fun to chase a few extra solutions per sec, but I'm in it for making money, not for bragging rights. I just counted, I'm only at 60 GPU's....it might be time to go shopping! But, I want to wait just a little longer until the Vega cards come out - either they will be much better, or the Nano's will come down in price. |
|
|
|
Just upgraded my linux boxes to 1.6.0. Looks like a solid and reliable build so far. Getting an increase across the board. At -i 7 cards are doing:
RX470 8G - 288
Rx480 8G -301
Fury X - 452
Fury Pro - 470
R9 390 - 365
How long have you been using your -i 7 ? mostly out of curiosity, i'm wondering whether that's just as stable as -i 6 on stock cards. But more importantly, have you also compared your Watt/Sol ratio between -i 6 and -i7 ? I've been using -i 7 for about a week now. Good question on Watt/Sol ratio. I had not thought of that! I have wattmeters and will be at the warehouse tomorrow so I will run a test. |
|
|
|
how much is dev fees at these miner???
optiminer just updated the changelog on his github repo saying it's 2.5% Version 1.6.0 released!  [1.6.0] Asm support for GCN 1 devices. [1.6.0] Reduced CPU utilization. [1.6.0] Fix segfault on reconnect. [1.6.0] Version and Os exported in monitoring. Speed increase about 3% for newer cards, the GCN 1 cards should see a much higher increase with the asm kernel now. 5% increase on my stock rx 470 (xfx true oc) on ubuntu 14.04 with amdgpu-pro 16.50 (netting me roughly ~262.5sols/s now, compared to ~250sols/s with 1.5.0) Just as a side-note, i setup 1.5.0 when it got released and never had to reboot/restart any rig (hardware or software). Just upgraded my linux boxes to 1.6.0. Looks like a solid and reliable build so far. Getting an increase across the board. At -i 7 cards are doing: RX470 8G - 288 Rx480 8G -301 Fury X - 452 Fury Pro - 470 R9 390 - 365 |
|
|
|
First you need to extract the GPU BIOS and keep a back up file, I use atiwinflash but its a windows application, dunno in linux.
Then with a BIOS editor, you change straps and save save as name.rom, again the ones I use are windows base and not linux.
Once you are done you need to flash the new modified BIOS into the GPU and reboot.
Make sure you save the original bios tho
ok, thanks! Sounds like a bit of work - when I have a few minutes I'll dig into it. |
|
|
|
Damn Phil, I never would've considered doing that vs a warranty repair, but that makes a lot of sense. I am with you in re: to Bitmain, they have no respect for their customers. If you're not buying 100+ units at a time, it's just not worth the risk of failed equipment. I'm starting to feel like legitimate Group Buys hosted at DC's by purchasing shares (and paying out mining reward according to shares regardless of who's unit actually breaks down) might start making sense again, like an insurance policy against being the unlucky one with a failed unit taking out a huge chunk of your revenue stream.
Will Bitmain not accept the dead hashboard back to China for repair by itself? Last time I tried warrantying an S7 they specifically mentioned NOT to send the whole miner, only to send the dead boards and grossly under-declare value for them. Doesn't really help you with the fan, but if I was you I would just source a furnace fan and duct it in instead of the type that ships with the R4. If it's covered under warranty it would be a shame not to try to get your value out of it.
On the autofrequency miners they require shipping entire units back. I shipped back some S9 and R4 boards, and they held the repair until I shipped the complete cases. Then they put the systems back together for repair. It gets expensive sending multiple units back. I'm going to start checking for bad capacitors, at least, before sending boards back in the future. |
|
|
|
|
