Little issue I'm having. 8 Sticks on 2x Gekkoscience Hubs shared between 2x PI4 8gb versions because I couldn't get 8 to play on one Pi correctly. I seem to have an issue the minute the PI has 2x USB Hubs going to it. I remember someone in here had the same issue with over 6 sticks on 1x PI. CPU/Memory usage is completely fine with 8 (Its "high" but never above 60% CPU).
This is more like what i am experiencing and where my questions originate from.
In summary, he has 8 sticks split across 2 hubs which is 2x 4 sticks per hub. then he has trouble connecting the 2 hubs with 4 sticks each on them to a single Pi4 as the problem starts the moment he plugs in the 2nd hub to the Pi4 which is exactly same thing im experiencing.
But then again, running one hub per Pi4 or PC works perfectly. So he ended up getting a 2nd Pi4 to get the 2nd hub running on its own without issues.
Is that the only solution ?? or are there some other things to try ? It doesn't seem to be a hub power issue but more about connecting multiple hubs to a PC or Pi. My PC isnt resource constrained yet.
My PC Specification:
Windows 11
Processor AMD FX(tm)-6100 Six-Core Processor - 4.12 GHz (24.7GHz total)
Installed RAM 12.0 GB
System type 64-bit operating system, x64-based processor
USB ports: 8 behind, 4 Infront (12 in total )
Suggestions ...
With that many ports, it's likely you have two distinct USB controllers in the machine. If you connect one hub to each controller, they won't be fighting each other for bandwidth. When we test at the factory, we got a bunch of machines with 6 ports in the back, 4 on one controller and 2 on another (shared with the 2 up front) so we connect one hub to each bus and let 'em eat their fill.
Also, performance on Windows is gonna suck more than *nix anyway just because of how Windows handles traffic.
Yes, it feels more like a bandwidth battle i agree.
I looked at the device manager side & found 4x OHCI USB 2.0 controllers + 1 xHCI USB 3.0 controller.
Tried swapping them around to be on separate controllers but i cant tell when they're on different controllers as im not getting the normal feedback.
any tips on identifying which controller manages which port ?
Would a Raspberry PI solve this data transfer problem ?
Damn straight it will. If you
look back in this thread you find the results of me changing from W10 to a Raspi 3B -- immediate 10% increase in hash rate while running the same freq plus was able to run at a higher freq.
I don't know if the issue is how Windoze allocates resources to USB or if is related to the Zadig USB driver but it made a world of difference.
edit: A thought... I assume the 2nd hub is plugged into its own port on the PC right? (it should be)
If so, did you tell Zadig to use the 2nd USB connection? If you didn't, `Doze will not like it...
Yes, each hub goes into a different port on the PC. I'm not sure i know how to instruct Zadig to use 2nd USB connection, all i made sure was to install the zadig driver for the sticks.
how can i go about about doing that with zadig ?
1A per USB port may not be enough to run them at 500
The Sidehack USB hub provides up to 6A per port (shared between two ports)
Its a 120W hub with 10 ports, im only populating 3 ports for F + cooling fan. they are pretty steady at 550Mhz when a single hub is plugged to PC but as soon as i plug in a second hub to PC, the frequencies begin to flutter & the Fs on the 2nd hub wouldnt go past 300Mhz while affecting the Fs on the first Hub causing it to drop also ...
Yeah the 10 port 120W hub is better than the 7port one I thought you had, but its still only 2.1A per port. I've had no end of trouble testing different hubs and only the Gekkoscience ones work reliably in my experience.
The only other trouble I have had was when I didn't have enough cpu grunt to drive three sticks. I was using a lowly Pi1 and it just couldn't drive more than one stick at full speed.
Oh and I managed to melt the solder off one of the chips on one stick but that wasn't terminal
The hubs supplies ok power.. there enough PC resource to go round ..