Braiins OS & Braiins OS+ custom ASIC firmware: optimize performance & efficiency

[Apprentice]

At 700M + 8.7V got to 14.85 Th/s and at 725M + 8.7V got to 15.2 Th/s
Temperature was around 80C and fans was working with much lower speed than default.
No idea if its possible to calculate power draw from voltage or I should get a separate power meter to check that.

hmm..well at 670 and 9V am getting 14.7xT
When increasing the frequency some of the chains are not supporting it and hashing lower. Guess this is where the BM autotune comes in handy.

I would check first the autotune frequencies on the original firmware and start from there, and assess which chain is versatile

[papampi_2]

hmm..well at 670 and 9V am getting 14.7xT
When increasing the frequency some of the chains are not supporting it and hashing lower. Guess this is where the BM autotune comes in handy.

I would check first the autotune frequencies on the original firmware and start from there, and assess which chain is versatile

Thats exactly where we need more info.
Whether higher voltage with lower freq is better or higher freq with lower voltage.

[sergw]

Can I ask whats the program you use to get all those details in that simple page?

this is simple php page that i created to collect miners stats via API and show on single page.

[jfaubin]

this is simple php page that i created to collect miners stats via API and show on single page.

Hi, it is possible to share?

[sergw]

Hi, it is possible to share?

sure, i will upload it on github and share here, hope in a few days i find time for it. i'll post link here for it.

[nick5429]

• Second alpha release Wolfram with overt AsicBoost support for all S9's

Presuming this is still derived from the original bitmain firmware's implementation -- would you please release an update for this rebased on the 11/3 update from bitmain that resolves the 'hashrate dropping over time' issue?  (Or advise that it's not necessary for your implementation)

https://twitter.com/Antminer_main/status/1058714676865855488

Thanks

[papampi_2]

These are the hashrates and powers I got in 15 minutes tests with antminer s9i 14Th/s:

Code:

Freq	Volt	Th/s	Watt	Watt/Th
650	8.8	13.80	1270	92.0290
675	8.7	14.28	1280	89.6359
700	8.7	14.70	1335	90.8163
725	8.7	15.30	1375	89.8693
725	8.6	15.00	1350	90.0000
750	8.7	15.65	1420	90.7348
750	8.6	15.30	1390	90.8497

Presuming this is still derived from the original bitmain firmware's implementation -- would you please release an update for this rebased on the 11/3 update from bitmain that resolves the 'hashrate dropping over time' issue?  (Or advise that it's not necessary for your implementation)

https://twitter.com/Antminer_main/status/1058714676865855488

Thanks

I have not tested braiins more than 15 minutes and have not seen drops in that time frame
Have you experienced hash drops over time?
If so after how long?

[jfaubin]

sure, i will upload it on github and share here, hope in a few days i find time for it. i'll post link here for it.

Thanks guy!

These are the hashrates and powers I got in 15 minutes tests with antminer s9i 14Th/s:

[...]

Hi Papampi,

Do you have temperature value with your s9I at all frequency?
I have my S9i at 700MHz, 8.8Volts and 14.81TH, but the temp is between 96 to 102 Celcius

https://imgur.com/gallery/cqAN9Ul

[nick5429]

I'm also seeing reported temperatures *much* higher (~20C hotter) than what the bitmain stock firmware reported.  I'm just not sure if the two are reporting the same measurement -- if not, they're not really comparable.  If these are the same measurements though, that's not great.

My S9's are in an air conditioned room at ~15C ambient air temp.

Stock bitmain 13.5TH: 65-70C reported temps
Braiins at similar hashrate: 85-90C reported temps

I sent in a ticket to braiins to ask for clarification...

[Apprentice]

I'm also seeing reported temperatures *much* higher (~20C hotter) than what the bitmain stock firmware reported.  I'm just not sure if the two are reporting the same measurement -- if not, they're not really comparable.  If these are the same measurements though, that's not great.

My S9's are in an air conditioned room at ~15C ambient air temp.

Stock bitmain 13.5TH: 65-70C reported temps
Braiins at similar hashrate: 85-90C reported temps

I sent in a ticket to braiins to ask for clarification...

Same here but I see the fabs on braiins spin slower, can tell from the lower pitch. Guess that's why they're hotter.

[jfaubin]

I'm also seeing reported temperatures *much* higher (~20C hotter) than what the bitmain stock firmware reported.  I'm just not sure if the two are reporting the same measurement -- if not, they're not really comparable.  If these are the same measurements though, that's not great.

My S9's are in an air conditioned room at ~15C ambient air temp.

Stock bitmain 13.5TH: 65-70C reported temps
Braiins at similar hashrate: 85-90C reported temps

I sent in a ticket to braiins to ask for clarification...

Good idea, keep me in touch!

J-F Aubin

[sergw]

is it possible to upload bitmain firmware via Brains OS web - interface? i would like to roll back firmware.

[jfaubin]

is it possible to upload bitmain firmware via Brains OS web - interface? i would like to roll back firmware.

Hi Sergw, why you would rollback?

[sergw]

Hi Sergw, why you would rollback?

on one of my s9 from feb-2017 batch it is not showing chip and board temps

[jfaubin]

on one of my s9 from feb-2017 batch it is not showing chip and board temps

ok, i understand!!!
Do you have a time to upload your web api page?

J-F

I'm also seeing reported temperatures *much* higher (~20C hotter) than what the bitmain stock firmware reported.  I'm just not sure if the two are reporting the same measurement -- if not, they're not really comparable.  If these are the same measurements though, that's not great.

My S9's are in an air conditioned room at ~15C ambient air temp.

Stock bitmain 13.5TH: 65-70C reported temps
Braiins at similar hashrate: 85-90C reported temps

I sent in a ticket to braiins to ask for clarification...

Hi Nick, i have a stupid question. Where i open ticket to braiins?

[sergw]

ok, i understand!!!
Do you have a time to upload your web api page?

J-F

i uploaded it on github, use the following topic to check:
https://bitcointalk.org/index.php?topic=5065875

[Steevithak]

Are there some screen shots of this BrainOS firmware's UI somewhere? Curious what it looks like compared to the stock firmware

[nick5429]

Dug through the source, recompiled, made several cgminer modifications and tried several tweaks.

Learnings, tested on a 13.5TH autofreq miner from Dec 2017, overclocked to ~15.5TH with asicboost enabled on Braiins and playing with the code:

1) The temperature displayed (at least for my model of S9) is a complete lie. The miner (and this code comes from bitmain) tries to read both PCB and ASIC temp.  The ASIC temps are deemed unreliable by the algorithm and/or can't be read on my model, so the code basically takes the actual PCB temp reading (generally 50-65C for me) adds an arbitrary 30-35C to "fake" an estimate at the ASIC temps.

Braiins web UI displays the faked temp including the offset (~90C), bitmain stock UI (autofreq, pre-asicboost) displays the original PCB reading (~60C).

2) Auto fan control works fine, it's just maybe biased a little slower than bitmain's -- mine hovered around 70% PWM in braiins.  I made modifications to cgminer to push the fan curve higher for mine to ensure good cooling for overclocking, I'm now generally at 80-90% PWM now. The difference between 70% PWM fan and 100% PWM fan is at least 20-40W power increase to run the fans.

3) Power usage on pre-asicboost bitmain firmware at 13.5TH is roughly equivalent to Braiins with asicboost at 15TH (9V and 712mhz set on each chain) for me

4) The 'recommended voltage' goes down as frequency goes up based on Bitmain's internally coded tables, which are intended to help the device stay within a given power usage envelope.  [until you hit the point of frying the thing], more voltage is always going to help the chain go faster -- it's just going to suck more power (and possibly more power than your PSU or cooling can handle)

I'm also working on an auto-frequency-adjuster that can lower frequency per-asic-chip on each board (like that one with the cool red/green UI, but automated).

It works well for modifying frequency on a per-asic basis, but the heuristic I was basing 'when should I turn down the frequency' on appears invalid, at least for some of the chips (I've been using 'reported per-asic hashrate not within 15% of ideal per-asic hashrate').  Anyone know what that red/green UI uses to determine when to display red?

[Breeze]

Nice work mate.
So in other words we have no clue what temps the chips are running? wow, gotta take it easy with OC.
Have mine in the garage so i have a colder enviroment....

[mariooiram]

Hi,

thanks for providing braiins os! We have successfully tested it on our S9's. Now we have tried to
flash it into the NAND memory of some bricked devices.

I put the questions above:

Q1: Is there a way to flash braiins os into NAND memory, when booted braiins os from SD?

Q2: is the step flas_eraseall required to flash?

What we have tried is to boot the bricked device from SD with braiins os. That works perfectly.
But when trying to flash braiins OS into NAND, that's not prepared by the install scripts, we tried several ways.

A) Remote via ssh and 'upgrade2bos.py' . That does not work, ends with "braiins os is allready installed".

B) copy everything required for running stage1.sh (of folder 'factory_transition') into /tmp/firmware of the
  bricked device, which is booted from SD, and run installation locally.
  There are some pitfalls:

   - We found that some binaries are missing on braiins os, that it uses when running stage1.sh, utilzing the
  binaries still in place from the S9:like flash_erase_all, ... we copied them from another miner.

   - Also we found that the mtd layout of the running braiins OS from SD does not hold the installer stage2:

Code:

(mtdparts=pl35x-nand:512k(boot),2560k(uboot),2m(fpga1),2m(fpga2),512k(uboot_env),512k(miner_cfg),22m(recovery),95m(firmware1),95m(firmware2))

Code:

root@m55:~# cat /proc/mtd
dev:    size   erasesize  name
mtd0: 00080000 00020000 "boot"
mtd1: 00280000 00020000 "uboot"
mtd2: 00200000 00020000 "fpga1"
mtd3: 00200000 00020000 "fpga2"
mtd4: 00080000 00020000 "uboot_env"
mtd5: 00080000 00020000 "miner_cfg"
mtd6: 01600000 00020000 "recovery"
mtd7: 05f00000 00020000 "firmware1"
mtd8: 05f00000 00020000 "firmware2"

 - So we patched the SD so it reflects the mtdparams required:

Code:

mtdparts=pl35x-nand:32m(BOOT.bin-env-dts-kernel),144m(angstram-rootfs),60m(upgrade-rootfs),20m(upgrade-tmp)

Then run stage1.sh locally. This seems to work, except - the binary flash_erase_all returns a warning that it's deprecated:

Code:

'flash_eraseall has been replaced by `flash_erase <mtddev> 0 0`; please use it'

Though the script runs through, flashes braiins-os into the nand and completes. But after reboot and jumpering to boot from NAND, the system remains inaccessible.

Do you have any advice how to flash braiins-os into a bricked device?

Thanks and regards, mario!


This has been resolved by the fantastic support, providing an updated bos2bos.py. Thank you!