Pages:
Author

Topic: Efudd Z-Series Fuddware 2.3 -Z11/Z11e/Z11j/Z9/Mini - page 16. (Read 45527 times)

jr. member
Activity: 82
Merit: 1
Cooldragon
So I hear you want to give me a fucked up image.... Thats pretty sad and low of anyone to sit and play games with someones machine like that. I dont care what your intention was or is. Thats fucking sad and your users need to be aware of sick thoughts or ideas. Whats to say you wont do that to them if you dont like what they say.

I am trying to help people and your out to fuck them so your gonna send me fucked up files trying to make it look like I dont know jack about whats going on so you can get more users.....I have seen it before and will probly see it again. I told you I have been messing with code since the age of 12 and it dont take long to get back into the groove of things. Well, I am back!!

And your the fucking idiot who wants to think he is a know it all. I will explain the voltage to your good and quick with the locations to test even... This is how to change a 7 to a 8 and get a little more speed.

In the image i started with 3/12/19 use ida do a search for 0x348 located at 3 places that is the default voltage of 840mV
Next search for 0X320 that is the low volt of 800mV located at least 6 places
Next do a search for 0x366 that is the current voltage of 870mV
Now search for 0x370 that is the high volt of 880mV

Located at 00049764 is the first byte of the running voltage and is 4 bytes to the code.
Change the 0x366 to 0x370 and you have now increased the voltage by .01 mV
If you set that over 0x370 (880mV) say to 0x371 (881mV) you get illegal voltage message even if you raise the high limit
Now you can change the 0x348 to increase the default from 840mV to whatever you want but if it is still out of the 800-880
you will get an illegal voltage message and the default is set. Even though you raise the default the pic is rejecting it and
automatically setting the 840mV

You can take a stock image change the 0x366 to 0x370 at 49764 and will automatically get an increase over the 142k
You can also undervolt them this way but may lose speed depending on the clocking.

Now anyone with winSCP and a hex editor can try it for themselves and see what they get. They can also look for the speeds like the default at location 4985C change the first or second byte of that 4 byte block and the default is now changed....And so on




Dear sir.

You dont oc the miners according to the source code of bmminer or cgminer.
But you directly edit the binary file?

Really cool.
jr. member
Activity: 559
Merit: 4
So I hear you want to give me a fucked up image.... Thats pretty sad and low of anyone to sit and play games with someones machine like that. I dont care what your intention was or is. Thats fucking sad and your users need to be aware of sick thoughts or ideas. Whats to say you wont do that to them if you dont like what they say.

I am trying to help people and your out to fuck them so your gonna send me fucked up files trying to make it look like I dont know jack about whats going on so you can get more users.....I have seen it before and will probly see it again. I told you I have been messing with code since the age of 12 and it dont take long to get back into the groove of things. Well, I am back!!

And your the fucking idiot who wants to think he is a know it all. I will explain the voltage to your good and quick with the locations to test even... This is how to change a 7 to a 8 and get a little more speed.

In the image i started with 3/12/19 use ida do a search for 0x348 located at 3 places that is the default voltage of 840mV
Next search for 0X320 that is the low volt of 800mV located at least 6 places
Next do a search for 0x366 that is the current voltage of 870mV
Now search for 0x370 that is the high volt of 880mV

Located at 00049764 is the first byte of the running voltage and is 4 bytes to the code.
Change the 0x366 to 0x370 and you have now increased the voltage by .01 mV
If you set that over 0x370 (880mV) say to 0x371 (881mV) you get illegal voltage message even if you raise the high limit
Now you can change the 0x348 to increase the default from 840mV to whatever you want but if it is still out of the 800-880
you will get an illegal voltage message and the default is set. Even though you raise the default the pic is rejecting it and
automatically setting the 840mV

You can take a stock image change the 0x366 to 0x370 at 49764 and will automatically get an increase over the 142k
You can also undervolt them this way but may lose speed depending on the clocking.

Now anyone with winSCP and a hex editor can try it for themselves and see what they get. They can also look for the speeds like the default at location 4985C change the first or second byte of that 4 byte block and the default is now changed....And so on


jr. member
Activity: 559
Merit: 4
The limit is in the PIC or elsewhere.
 
Any increase you see is randomness.

I have long since REd and replaced the entire code path around frequency management, handling, autotuning and scanning.

-j

I am holding a small increase at 143-144, but I have theory at how it is reducing over time after monitoring this last week. The hash rate showing as the average may be spoofed. With the changes I did my hashrate went from 450-455 to 455-463 for all my miners. The z11's went from 142 to 143-144 but the miner didn't show as high of an increase like the server has in shares.

Spoofing the rate was started for gpu software so the hacker didn't know if they were getting more speed or not because no matter what was done to the software the rig showed the same speeds.

edit update...

This is my most recent hash rate  from the flypool server for 1xZ11
147.3 KH/s   5522 (100%)   0 (0%)   9 minutes ago

same miner right after update

146.5 KH/s   5493 (100%)   0 (0%)   2 minutes ago

Randomness. Hash rate pool side is derived based on share submission (and share difficulty) which is impacted by luck and other attributes associated with cryptography.

-j

When the machine is showing 144-145 that is not randomness when it has been steady
member
Activity: 504
Merit: 51
The limit is in the PIC or elsewhere.
 
Any increase you see is randomness.

I have long since REd and replaced the entire code path around frequency management, handling, autotuning and scanning.

-j

I am holding a small increase at 143-144, but I have theory at how it is reducing over time after monitoring this last week. The hash rate showing as the average may be spoofed. With the changes I did my hashrate went from 450-455 to 455-463 for all my miners. The z11's went from 142 to 143-144 but the miner didn't show as high of an increase like the server has in shares.

Spoofing the rate was started for gpu software so the hacker didn't know if they were getting more speed or not because no matter what was done to the software the rig showed the same speeds.

edit update...

This is my most recent hash rate  from the flypool server for 1xZ11
147.3 KH/s   5522 (100%)   0 (0%)   9 minutes ago

same miner right after update

146.5 KH/s   5493 (100%)   0 (0%)   2 minutes ago

Randomness. Hash rate pool side is derived based on share submission (and share difficulty) which is impacted by luck and other attributes associated with cryptography.

-j
member
Activity: 504
Merit: 51
Hello Jason.

When wait 2.2 for z9mini with NiceHash support?

Thank you.

Hey man, it will be at least a week. I have other non-crypto related things that are priority at the moment.

-j
newbie
Activity: 17
Merit: 0
Hello Jason.

When wait 2.2 for z9mini with NiceHash support?

Thank you.
jr. member
Activity: 559
Merit: 4
The limit is in the PIC or elsewhere.
 
Any increase you see is randomness.

I have long since REd and replaced the entire code path around frequency management, handling, autotuning and scanning.

-j

I am holding a small increase at 143-144, but I have theory at how it is reducing over time after monitoring this last week. The hash rate showing as the average may be spoofed. With the changes I did my hashrate went from 450-455 to 455-463 for all my miners. The z11's went from 142 to 143-144 but the miner didn't show as high of an increase like the server has in shares.

Spoofing the rate was started for gpu software so the hacker didn't know if they were getting more speed or not because no matter what was done to the software the rig showed the same speeds.

edit update...

This is my most recent hash rate  from the flypool server for 1xZ11
147.3 KH/s   5522 (100%)   0 (0%)   9 minutes ago

same miner right after update

146.5 KH/s   5493 (100%)   0 (0%)   2 minutes ago
member
Activity: 504
Merit: 51
The limit is in the PIC or elsewhere.
 
Any increase you see is randomness.

I have long since REd and replaced the entire code path around frequency management, handling, autotuning and scanning.

-j
jr. member
Activity: 559
Merit: 4
Z11 image is currently in testing, but due to some family commitments, I will not have time to release it until Monday, unfortunately.

On-going work will continue to further unlock these units. I have been able to see > 180kSol for brief periods of time on Z11s, but I have not been able to get it to stick more than a few minutes at a time until I find what else is acting as a limiter (my testing was done with 3x PSU to ensure that available power was not an issue).

This is the change log for the Z11 image when it does release.
==============================================
This is Fuddware 2.2 for the Z11 series miner.

* Full control is given back to the user.
    - SSH enabled
    - Frequency setting
    - No signature checking for future firmware updates
* Per-Hashboard frequency setting is enabled
* Support for NiceHash
* Proper #xnsub support
* Protection against known antminer malware

This image will not have any dev-fee for the next two weeks. After two weeks, dev-fee will be enabled once a day for 25 minutes.

NOTE: This firmware does allow you to set the frequencies as you see fit, however at this point in time, it is unlikely you will exceed 142kSol on average. I will continue to work on this highly desired feature. Setting higher frequencies just lets the unit get up to speed faster. Also, Bitmain's auto-tuning tends to be a bit too conservative and will artificially limit your hashrate.
==============================================

Thank you,

Jason



Dear sir,

About Z11,
1, Should the Z11 can be overclocked by simple methods, such as simple update the custom firmware(no need sig), or simple app through ssh?
2, The 142k is almost the same as 135k, would you able to oc it at least about 10%?
3, If you have any update, please let me know, thanks a lot!

I am working on a custom fw for the Z11 and have increased from 142 to 143-145 I think know where the holdback may be on the speed and am trying to work around it. Bad news is it may not be able to work around because the limit may be coded into the pic chip. I have 2 possible workarounds I am testing.
jr. member
Activity: 82
Merit: 1
Cooldragon
Z11 image is currently in testing, but due to some family commitments, I will not have time to release it until Monday, unfortunately.

On-going work will continue to further unlock these units. I have been able to see > 180kSol for brief periods of time on Z11s, but I have not been able to get it to stick more than a few minutes at a time until I find what else is acting as a limiter (my testing was done with 3x PSU to ensure that available power was not an issue).

This is the change log for the Z11 image when it does release.
==============================================
This is Fuddware 2.2 for the Z11 series miner.

* Full control is given back to the user.
    - SSH enabled
    - Frequency setting
    - No signature checking for future firmware updates
* Per-Hashboard frequency setting is enabled
* Support for NiceHash
* Proper #xnsub support
* Protection against known antminer malware

This image will not have any dev-fee for the next two weeks. After two weeks, dev-fee will be enabled once a day for 25 minutes.

NOTE: This firmware does allow you to set the frequencies as you see fit, however at this point in time, it is unlikely you will exceed 142kSol on average. I will continue to work on this highly desired feature. Setting higher frequencies just lets the unit get up to speed faster. Also, Bitmain's auto-tuning tends to be a bit too conservative and will artificially limit your hashrate.
==============================================

Thank you,

Jason



Dear sir,

About Z11,
1, Should the Z11 can be overclocked by simple methods, such as simple update the custom firmware(no need sig), or simple app through ssh?
2, The 142k is almost the same as 135k, would you able to oc it at least about 10%?
3, If you have any update, please let me know, thanks a lot!
member
Activity: 504
Merit: 51
Z11 image is currently in testing, but due to some family commitments, I will not have time to release it until Monday, unfortunately.

On-going work will continue to further unlock these units. I have been able to see > 180kSol for brief periods of time on Z11s, but I have not been able to get it to stick more than a few minutes at a time until I find what else is acting as a limiter (my testing was done with 3x PSU to ensure that available power was not an issue).

This is the change log for the Z11 image when it does release.
==============================================
This is Fuddware 2.2 for the Z11 series miner.

* Full control is given back to the user.
    - SSH enabled
    - Frequency setting
    - No signature checking for future firmware updates
* Per-Hashboard frequency setting is enabled
* Support for NiceHash
* Proper #xnsub support
* Protection against known antminer malware

This image will not have any dev-fee for the next two weeks. After two weeks, dev-fee will be enabled once a day for 25 minutes.

NOTE: This firmware does allow you to set the frequencies as you see fit, however at this point in time, it is unlikely you will exceed 142kSol on average. I will continue to work on this highly desired feature. Setting higher frequencies just lets the unit get up to speed faster. Also, Bitmain's auto-tuning tends to be a bit too conservative and will artificially limit your hashrate.
==============================================

Thank you,

Jason

member
Activity: 504
Merit: 51
Folk,

I am pleased to announce the release of Fuddware version 2.2 for the Z9 with NiceHash support! (sorry Mini users, I hope to resolve this for you soon).

This release is for the Z9 unit only and is available at https://releases.broked.net

Change log for version 2.2:

  • Adds Support for Nicehash (yay!)
  • Adds proper #xnsub support
  • Adds protection against known antminer malware.
  • NOTE: If you are a paid user, you will need to re-apply your license after upgrade.
  • SHA256: f77310d7d0cb7ea42ed84c8c5f91749767faf18af04cabab332caea26e308356  Z9_2.2.tar.gz
  • MD5SUM: 35c7a7c9bbfb321ed8603657fa36710d  Z9_2.2.tar.gz

EDIT: There was a small mistake in the image -- it has since been fixed. If you downloaded and have issues installing, please download it again.

Thank you,

Jason
jr. member
Activity: 559
Merit: 4
I would love to see the pic code from the Z11 …. Could be a rewrite and that is where the holdup is behind the scenes
member
Activity: 504
Merit: 51
No to both.
jr. member
Activity: 559
Merit: 4
I kind of have, I just can't get it to hold it. I can get some ASICs stable at 800Mhz, and all they end up doing is getting up to 142kSol faster (30 seconds from startup vs. 5-10 minutes).

I don't have a Z11 myself this time around so I'm only able to test remotely on other peoples machines; the progress is going to be slow at best. Something I have yet to find appears to be limiting the maximum hash rate on the system (it is not the auto-tuning, I can turn that off).

-j


Do you think maybe that they are throttling the voltage to control the 142k/sol and not throttling it by the freq. I have only looked at the disassembly a little bit and not enough to come to some of the conclusions you have already. Or are they using the network delay to slow down the data transfer. Cgminer can add a delay between its packets and was normally used for slow routers.

I also wonder if they are using the cgminer 4.10 source for the base since the autotuned wasn't implemented into cgminer until version 4.10 if I am correct.
member
Activity: 504
Merit: 51
I kind of have, I just can't get it to hold it. I can get some ASICs stable at 800Mhz, and all they end up doing is getting up to 142kSol faster (30 seconds from startup vs. 5-10 minutes).

I don't have a Z11 myself this time around so I'm only able to test remotely on other peoples machines; the progress is going to be slow at best. Something I have yet to find appears to be limiting the maximum hash rate on the system (it is not the auto-tuning, I can turn that off).

-j
jr. member
Activity: 97
Merit: 5
https://releases.broked.net/Z11StockSSHNoSignature.zip

This is the stock Z11 firmware dated Tue Mar 12 19:47:06 CST 2019
+ SSH has been enabled
+ Signature checking has been removed for future firmware updates
+ Protection against the known antminer malware has been added
+ No Dev-Fee/Call-Home
+ No additional features or changes exist.

This can be used to enable AwesomeMiner to manage your Z11 ASIC.


Hopefully, now you can find that extra 40k/sols.
member
Activity: 504
Merit: 51
https://releases.broked.net/Z11StockSSHNoSignature.zip

This is the stock Z11 firmware dated Tue Mar 12 19:47:06 CST 2019
+ SSH has been enabled
+ Signature checking has been removed for future firmware updates
+ Protection against the known antminer malware has been added
+ No Dev-Fee/Call-Home
+ No additional features or changes exist.

This can be used to enable AwesomeMiner to manage your Z11 ASIC.

SHAH256: 4c034f823a68d4a05254630a8082bad13136b3ae85399a3cfae2aac00227fb70  Z11StockSSHNoSignature.zip
MD5SUM : 393430f23dd7ac3db0740ef7acabcc6f  Z11StockSSHNoSignature.zip

That will need to be installed via SD card:

1) Format SD card 32GB or less as FAT32
2) Extract zip file contents to root of SD card
3) Power off miner.
4) Insert SD card, power on miner.
5) WAIT (30-60 seconds) for lights to go off to the right of the ethernet port.
6) Unplug miner, remove SD card.
7) Plug miner back in.
Cool [optional] Install firmware of your choice.
member
Activity: 504
Merit: 51
Z11 Status:

I was able to test with a Z11 tonight on using a dev image. Like for like functionality with the Z9/mini thankfully worked well. That said, there appears to be a limit in the maximum hash rate independent of frequency, somehow.

It also appears that the auto-tuning functionality may end up setting a board lower than it is actually capable of performing and stay there.. I'm not quite sure what conditions cause it to reset.

It has been pointed out that awesomeminer requires SSH access to set the API configuration and with SSH disabled in the factory image, that is a sticking point. I had taken care of this by default in my Z9 and Mini images and had forgotten about that.

I'll get a baseline image out (no dev fee, blah blah ) tomorrow with SSH enabled and after some more testing, I'll add the parity-with-Z9 functionality in so you can choose your own adventure rather than bitmain selecting the wrong clock speeds for you automatically.

Past that, I'm going to have to keep trying on understanding how/why the hash rate appears limited independent of frequency and see if that can be overcome. I think these units have another 40kSol or more in them.

Thank you,

Jason

jr. member
Activity: 559
Merit: 4
Jason

I did get the controller stripped but there is a middle layer...I will upload the pics a I have later tonight…. I don't need the board if you want it pm me your address and I can mail it to you. I would rather leave it unshaved to view the middle section, this way it don't destroy the top or bottom layer stripping it off

Edit.... Here are the links to the stripped Z9 controller. It was a version 1.1... as I said if you want it just tell me where to mail it.

https://drive.google.com/file/d/1dd6lY58zxRDpSGisMmRGGckdwIZP7yGK/view

https://drive.google.com/file/d/1FwGoCVHDsASiCNyxZelKNzdwemdYGpEW/view

https://drive.google.com/file/d/1-sJb5X7qpWnHfNP6PLDcgSazEloQydSp/view

https://drive.google.com/file/d/10Pu1y3Ct9pWeWcUT7A9AAsbfyU9Tu1ST/view

https://drive.google.com/file/d/1agdgTwADJwWfLGy7J2J5qeerNG1wkgp7/view
Pages:
Jump to: