Topic: Asic.to Firmware S17+ 95th/s • T17+ 80th/s T17 40w/t • S17/T17 on over 200k Asic - page 4. (Read 8400 times)

Hey guys, I see earlier in the thread there was mention of the asic.to firmware working with the S17e's? Is this correct? Don't see it on the website available for download, unless you just use the S17 version of the firmware.

I used slushpool.

Power consumption is an estimate and every unit draws different amount at same speed. Reported chip temperature SHOULD be valid as it's drawn from temp sensors same as in original Bitmain firmware. For underclocking it's great, for some overclocking results, I am not sure what is going on. We will have to wait for taserz to come and chime in.

Meanwhile, if you haven't reverted back yet, take some screenshots and copy/paste Kernel Log in .txt file so we can check what is going on.

I know you don't, I tend to trust your words but it's common sense that you should provide a proof to support such a serious claim, I have interacted with you a lot of times and as I said I do tend to believe you, but if people ask for a proof you shouldn't take it personal, it's normal, now back to the real subject:

I highly think that there is something techinally wrong here, there are a ton of reasons why would your mining gear report less hashrate on the pool than shows on the miner status page, in order for anyone including the firmware devs to troubleshoot the issue - a few informations are needed.

1- Ping stats to the pool.
2- Screenshot from the miner status page that shows everything from time-elapsed, hashrate, hardware errors, stale, rejected, and etc.
3- Screenshot from the pool status showing all hashrate related details.
4- The kernel log.

There is a good chance that your gears simply didn't like the firmware and they were mining with a ton of hardware errors, or during the test your connection to the pool got lousy and a good amount of shares got lost in the way, or the miner was tunning for a whole hour or so (not exactly hashing to the pool) and you took that time into considerations, I mean there is a lot of reasons that can cause a huge difference of 14%, one of them of course is the dishonesty of the dev-fees which i highly doubt in this given situation.

I've no reason to lie about it. Which pool were you testing with?

Perhaps the difference varies with different settings, but there IS a difference and it's significant, which gives me trust issues. If the hashrate reported is inaccurate, how can I trust that the power consumption is accurate, or reported chip temperatures etc.

I'm not saying I'd never use this firmware, but for the 2 x units I tested on (which are perfectly healthy, albeit 9 months old) - I think stock firmware is best for now.

I may use this in future for underclocking, but I won't trust the figures in the GUI.

I already informed taserz about this problem long ago, but it wasn't that noticable for me. Fee is 3% and danieleither is losing almost 14% of his hashrate (if he is saying truth as I do not see any pictures).

My difference is closer to 5-7% instead of 3% but that is only if I OC. As soon as I downclock, it's closer to 4-5% (which is still higher than 3%) but it's nowhere near the 14% that danieleither is saying.

I sent tasers a message on telegram, he will respond to your post soon (hopefully) he probably has an answer on how to go back to the original firmware, the difference in hash rate is huge to say the least, it is strange that you are the first to mention it, let us see what tas has to say on this matter.

I did indeed - it's definitely still Asic.to firmware.

I tried this on 2 different units, one overclocked and one underclocked. These are T17+ 58 TH/s, around 11 months old.

In both cases, the performance at the pool is significantly less than the performance reported from the miner GUI. To be fair, I used the 24 hours measurements from pool with the miner GUI showing > 24hr elapsed hashing.

On the overclocked machine: GUI showing consistent 67 TH/s - pool reporting around 58 TH/s (24 hour period)
On the underclocked machine: GUI showing consistent 34 TH/s - pool reporting around 28 TH/s (24 hour period)

Pool is F2Pool. I wondered perhaps it's a pool thing, however I have 10 x S9 (14TH/S) overclocked with Asicseer firmware and they are spot-on (GUI reports 16.8 TH/s, Pool reports same).

So my conclusion, the miners work better with Bitmain firmware. I have no use for this at the moment. The only use I can think is perhaps if the miner is faulty and this custom firmware can be used to make it more stable, something like that...

Did you try to press CTRL+F5 after flashing the new firmware? in most cases, the flash is successful using only the web update method but the browser's cache needs to be refreshed.

Also would you explain the reason/s for leaving Asic.to? what problems did you face? what did you dislike about it?

Not really sure that can be done, there is however a few of sd card images around web which Bitmain sent officially to someone for purposes of troubleshooting. Try to find it or just join Vnish group on Telegram, Taserz is active there every day, unlike here.

Is that simply a case of putting the original firmware on an SD card (as available from the Bitmain download page) or is a specific SD card version required?

Afaik, you can just use sdcard with original firmware. Not really sure if that is the only option. I cannot remember if I used normal "upgrade" via web or sdcard when I was flashing back my S17 PRO to original after it broke (it broke on original, vnish's firmware helped me use it for couple of months with 2 hashboards but then I decided to RMA it as it was broken before I even touched anything).

What is the process to revert to Bitmain firmware? I have tried to do this using the standard 'web' method but it doesn't work. It appears to upload the firmware but when the device reboots, it's still the Asic.to firmware. This doesn't appear to be documented in the manual or on the website...

Honestly, this is the only rational way of treating these fragile gears, except for the S17 pro probably all other gears are already clocked very near what they can handle, and with the terrible solder paste on all of them every single degree of temp matters, so what we know?

1- Running the 17 series on default settings has about 30% failure/RMA rate
2- Overclocking them will with a doubt set you above 30% failure rate, so maybe 50%?
3- Underclocking them will obviously get you a much lower rate than 30%

So if one doesn't want to go with the third option, you would at least not go with the second one, as phill mentioned before, in order for these gears to remain profitable, they need to remain functional in the first place.

Last I heard kano lives in the land down under. Not the USA.

It's actually the underclocking and using less power option that appeals to me. I have 18 x new S19 Pro on order from Bitmain, but only enough power available for 15. I was planning to sell on some of my older gear, but having read this about custom firmware, I was thinking perhaps to underclock some of my older T17+ units which have had a hard life to free up some power rather than selling them on. That being said, I do also have 12 x new T17+ units (in a different installation).

I've now installed this firmware on 2 of my old (non-warrantied) T17+ units to do some testing. Will post some results shortly

Yes, but you have trust the firmware maker on that, taserz can lock you in forever if he wants to, he could probably even disable the SDcard boot with some signature/security stuff and then you will be stuck with the firmware forever unless you buy a new control board, but taserz won't do such a thing, he is a nice lad and highly trusted.


If we all agree that COULD and WOULD mean two different things then "could they" is probably yes, it's technically possible for them to know if they have implemented the right method to record firmware updates but I doubt they have any of that, now the "would they" part is a  no, I have heard from a few people whom I trust very much that they returned their miners to bitmain after having used a custom firmware and bitmain served them without an issue ( obviously after having flashed back the firmware), but there is no guarantee that it will work for everyone so that will be at your own risk, and honestly you shouldn't send the gear back if you flash a custom firmware.

If I were to run a custom firmware and toast my gear, I would most certainly not send it back to bitmain because that is plain dishonesty and stealing, it's like putting olive oil in your brand new car's engine and once the engine goes bad you send it back to the dealer falsely claiming it's not your fault, of course, Asic.to isn't like olive oil in your car engine, in fact you could very well underclock it and bitmain would be happy with that since it reduces the chances of you sending the gear back, but since they don't allow you to do so, it's nothing but cheating if you do it.

Forgive me if these are stupid questions but:

1) After installing this firmware, if you're not happy with it, is it possible to flash the miner back to original Bitmain firmware?
2) If you did the above and had to return the unit back to Bitmain for repair, could they tell that you've been running aftermarket firmware on the device?

You joined 8 days before him. Odd...

Imagine this conspiracy theory: Kano is actually Taserz and he is doing this as publicity stunt  

JK, but yeah, you guys joined almost exactly 2 years after bitcointalk was launched.

I mean f2pool, viabtc, slush, poolin, binance must I go on

Also, look when I joined this forum I lurked a while before signing up I just like the tech aspect of mining.

He can't go after Bitmain, because he allows their gears which violate the cgminer license on his pool, and he directly or indirectly makes a lot (probably most) of his profit from those mining gears.


We all agree on Kano's terrible behavior and personality but I have to disagree to this, Kano has been in the game from the get-go, if not for his behavior and perhaps weak hands in Hodling BTC - he would have been so rich to the point where he wouldn't bother running a mining pool, with that being said we have to give credit where credit is due, Kano is probably in the top 5 of best pool developers. if only his attitude was half as good as his skills.