Inquiries about USB ports and controllers on a motherboard
Inquiries about USB ports and controllers on a motherboard
Hello, thank you for your message. I’ve attached a screenshot of my motherboard specifications (ASRock Z790 Nova WiFi). I’m planning to link four devices: a mouse, a keyboard, a USB microphone, and an external DAC/amp. My goal is to achieve the lowest jitter and latency possible, especially prioritizing the mouse.
Regarding the mouse, I’ve faced some problems with USB 3.0 ports in the past—they can cause a sluggish feel when tracking on the screen. While USB 3.0 promises faster speeds and lower latency, does it really matter if the port is USB 2.0? Since USB 2.0 offers the most stable connection, it might be better for consistent performance. I’ve read similar discussions on forums, which align with my experience.
I also want to clarify the difference between ports that connect directly to the chipset and those managed by a USB controller. A “redriver” isn’t a USB controller itself but a driver that stabilizes the high-speed signal from a USB 3.0 port. Without it, a mouse could suffer from noticeable jitter. In short, using a proper redriver helps maintain accuracy and smoothness.
As for the setup, I’m considering:
- Mouse via USB 2.0 port
- Keyboard via USB 3.2 Gen 2 (USB32_4)
- Microphone via USB 3.2 Gen 1 (USB32_6)
- External DAC/amp via USB C (USB32_TC1)
The motherboard manual suggests connecting the mouse and keyboard to USB32_3 and 4, but I’m not sure about those exact ports. The explanation about the controller ports (ASM3042) is helpful—it’s meant to improve signal stability, especially for devices like mice.
Please let me know if my understanding is correct or if I missed something.
For optimal performance, using a premium PCIe→USB adapter would be ideal. Simply connecting it directly into the M.2 slot on the CPU should work without issues. However, you may need multiple adapters, and standard USB ports might suffice if you use just one.
If I were to proceed, I wouldn't have a direct M2 drive connected to the CPU (there are two, but the top one if I link something changes the GPU PCIe slot from x16 to x8). For the best game performance, I'd choose the drive for gaming with minimal latency. If I connect my mouse on the second CPU's M2 slot, it would likely offer the lowest latency and jitter. Just plug in the M2 adapter and use a USB male-to-female cable.
Which one are you asking about? You’re referring to the single unit connected directly to the CPU, the one passing through a redriver and switch—this introduces some delay. Lower bandwidth for your GPU seems less ideal too. A smaller bandwidth for your GPU definitely feels suboptimal. It’s hard to say this setup is completely pointless; every system has its inefficiencies or slowdowns. Also, which games are you playing that don’t keep their main assets in RAM? That could shift performance priorities toward faster drives.
M2_2 would likely be my best choice. However, I'm not sure I'd want to follow that path. It seems to be the top recommendation for a mouse connection, though I doubt it's feasible with the M2 ports you mentioned. I think connecting devices via USB 2.0 or 3.0 would be more practical. The idea of using M2 directly to power the CPU isn't something I'd pursue, as it might cause too much latency. For gaming and other tasks, a direct CPU connection through a mouse adapter could actually be faster.
I think you're being too detailed, just switch the devices to separate USB buses and stop worrying. I've got a wireless mouse connected to my monitors via a hub, sharing with the keyboard and even a webcam—no lag issues there. My sensitivity makes me pay close attention to mouse responsiveness, so it's mainly about the mouse's quality and polling rate.
Several motherboards come with multiple ports, which explains the presence of both 5Gbit and 10Gbit USB connections. From what I grasp, USB 2.0 operates separately from USB 3.0. You’d only consider using a M.2 slot for USB if you’re facing specific issues that can’t be fixed otherwise.
I was testing the Universal Serial Bus interface, but it seems logical that varying speeds correspond to different buses. It’s possible they operate through an internal switch, allocating specific ports with a minimum bandwidth of X. I also learned that USB distributes its bandwidth across the whole bus, which could explain why 2.5Gbit adapters might appear slower when connected to the same bus as storage devices.
Thank you both for your input. I'm still processing this carefully. However, since I'm sensitive to delays and want the optimal result, testing each port could help. I'll maintain the m2 option as a fallback. It might be worth trying it just for the experience, regardless of whether I decide to keep it or not.