Success stories

[p0358]

Please share about the devices you've successfully overclocked. Use a top comment for device entry and sub-comments to comment on individual entries.

Failure stories on the other hand should go here: #2

Please include:

• whether this was a success or not
• the device name (can check with lsusb, or use the name you know, as sometimes the name from USB descriptor is ridiculous and doesn't tell you anything)
• the VID:PID (check with lsusb)
• the USB speed as reported by lsusb -d VID:PID -v | grep '^Negotiated speed:'
• the USB version as reported by lsusb -d VID:PID -v | grep bcdUSB
• the bInterval of the device and corresponding Hz
• the original bInterval of the device (on Interrupt endpoints) and corresponding Hz, can be checked with lsusb -d VID:PID -v | grep -A5 'Transfer Type *Interrupt' | grep -E 'Transfer Type|bInterval'
• what software was used to benchmark/validate the final polling and what was its result

As for benchmarking, for controllers you can use Gamepadla Polling (gamepadla-polling^AUR) and for mouse you can use evhz (evhz-git^AUR).

I'll include the template below, but feel free to adjust it if you feel the need (or add any comments at the end):

- **Overclock success**: ✅
- **Device name**: 
- **VID:PID**: ``
- **USB Negotiated speed**: 
- **bcdUSB**: 
- **bInterval and Hz**: 
- **Original bInterval and Hz**: 
- **Benchmarked with**: 
- **Benchmark score**: 

[p0358]

• Overclock success: ✅
• Device name: Sony DualSense
• VID:PID: 054c:0ce6
• USB Negotiated speed: High Speed (480Mbps)
• bcdUSB: 2.00
• bInterval and Hz: 1-4, 1000 Hz
• Original bInterval and Hz: 6, 250 Hz
• Benchmarked with: Gamepadla Polling + evhz
• Benchmark score: 986.19 Hz

=== Polling Rate ===
Average: 986.19 Hz
 
=== Refresh intervals ===
Minimal interval:   0.96 ms
Average interval:   1.01 ms
Maximum interval:   2.01 ms
Jitter:             0.11 ms
 
=== Outliers Report (based on IQR method) ===
Number of outliers: 815
Outliers percentage: 13.58%
Average outlier delay: 1.23 ms

bInterval 4 to 1 seemed to poll 1000 Hz. Only with evhz I could see the frequencies of 2000, 4000, 8000 being reported, but only for the motion sensors.

The evhz measurements are as follows:

=== bInterval 1 "8000 Hz"
Average for Sony Interactive Entertainment DualSense Wireless Controller:   999Hz
Average for Sony Interactive Entertainment DualSense Wireless Controller Motion Sensors:  6133Hz
Average for Sony Interactive Entertainment DualSense Wireless Controller Touchpad:   199Hz

=== bInterval 2 "4000 Hz"
Average for Sony Interactive Entertainment DualSense Wireless Controller:   999Hz
Average for Sony Interactive Entertainment DualSense Wireless Controller Motion Sensors:  3447Hz
Average for Sony Interactive Entertainment DualSense Wireless Controller Touchpad:   199Hz

=== bInterval 3 "2000 Hz"
Average for Sony Interactive Entertainment DualSense Wireless Controller:   999Hz
Average for Sony Interactive Entertainment DualSense Wireless Controller Motion Sensors:  1999Hz
Average for Sony Interactive Entertainment DualSense Wireless Controller Touchpad:   199Hz

=== bInterval 4 "1000 Hz"
Average for Sony Interactive Entertainment DualSense Wireless Controller:   992Hz
Average for Sony Interactive Entertainment DualSense Wireless Controller Motion Sensors:  1000Hz
Average for Sony Interactive Entertainment DualSense Wireless Controller Touchpad:   200Hz

=== bInterval 5 "500 Hz"
Average for Sony Interactive Entertainment DualSense Wireless Controller:   499Hz
Average for Sony Interactive Entertainment DualSense Wireless Controller Motion Sensors:   499Hz
Average for Sony Interactive Entertainment DualSense Wireless Controller Touchpad:   209Hz

So it seems the actual polling rate of sticks doesn't go above 1000 Hz. Might still be worth to operate the controller with bInterval 3 for 2000 Hz, as that seems to offer more stable 1000 Hz. It seems that any higher modes are pointless, unless you care about the motion sensors polling.

[p0358]

• Overclock success: ✅
• Device name: Victrix Gambit Tournament
• VID:PID: 0e6f:02d6
• USB Negotiated speed: Full Speed (12Mbps)
• bcdUSB: 2.00
• bInterval and Hz: 2, 500 Hz
• Original bInterval and Hz: 4, 250 Hz
• Benchmarked with: Gamepadla Polling
• Benchmark score: 500.00 Hz

=== Polling Rate ===
Average: 500.00 Hz
 
=== Refresh intervals ===
Minimal interval:   1.99 ms
Average interval:   2.00 ms
Maximum interval:   2.02 ms
Jitter:             0.01 ms
 
=== Outliers Report (based on IQR method) ===
Number of outliers: 20
Outliers percentage: 1.01%
Average outlier delay: 5.84 ms

The controller could be overclocked to 1000 Hz (and it would indeed even sometimes show round 1000.00 Hz on benchmark), but it seems to be problematic, weirdly lagging in games from time to time during intensive scenes. And somehow 500 Hz overall just seemed to feel better. Also on my laptop, it'd poll just 517 Hz when I tried bInterval 1. It did the same on my PC under Windows, despite technically polling 1000 Hz under Linux, but having these aforementioned problems anyway.

=== Polling Rate ===
Average: 999.00 Hz
 
=== Refresh intervals ===
Minimal interval:   0.99 ms
Average interval:   1.00 ms
Maximum interval:   1.28 ms
Jitter:             0.01 ms
 
=== Outliers Report (based on IQR method) ===
Number of outliers: 20
Outliers percentage: 1.00%
Average outlier delay: 1.84 ms

[Tsune-Reient]

• Overclock success: ✅
• Device name: Sony Corp. DualShock 4 [CUH-ZCT2x]
• VID:PID: 054c:09cc
• USB Negotiated speed: Full Speed (12Mbps)
• bcdUSB: 2.00
• bInterval and Hz: 1, 1000 Hz
• Original bInterval and Hz: 5, 250 Hz
• Benchmarked with: Gamepadla Polling
• Benchmark score: 1000 Hz

Left Stick:

=== Polling Rate ===
Average: 853.24 Hz
 
=== Refresh intervals ===
Minimal interval:   0.99 ms
Average interval:   1.17 ms
Maximum interval:   4.00 ms
Jitter:             0.48 ms
 
=== Outliers Report (based on IQR method) ===
Number of outliers: 823
Outliers percentage: 13.72%
Average outlier delay: 2.40 ms

Right Stick:

=== Polling Rate ===
Average: 677.97 Hz
 
=== Refresh intervals ===
Minimal interval:   0.99 ms
Average interval:   1.48 ms
Maximum interval:   5.01 ms
Jitter:             0.76 ms
 
=== Outliers Report (based on IQR method) ===
Number of outliers: 179
Outliers percentage: 2.98%
Average outlier delay: 4.73 ms

[afdw]

• Overclock success: ✅
• Device name: GIGABYTE M5050 (Pixart Imaging, Inc. Optical Mouse), an old basic mouse
• VID:PID: 093a:2510
• USB Negotiated speed: Low Speed (1Mbps)
• bcdUSB: 1.10
• bInterval and Hz: 1, 1000 Hz
• Original bInterval and Hz: 10, 125 Hz
• Benchmarked with: UFO Test: Mouse Poll Rate Test / Browser Mouse Performance and evhz
• Benchmark score: 1000 MHz

Thank you very much for this module! I would remark that it works perfectly (everything is stable and very smooth) even though I have a low speed device, and this is contrary to what is written in the README. Here is what I get in practice for this device:

Value of bInterval	Polling Period	Frequency
1	1 ms	1000 Hz
2 to 3	2 ms	500 Hz
4 to 7	4 ms	250 Hz
8 to 15	8 ms	125 Hz
16 to 31	16 ms	62.5 Hz
32 to 63	32 ms	31.25 Hz
64 to 127	64 ms	15.625 Hz
128 and above	128 ms	7.8125 Hz

[p0358]

Oh, that's very interesting then, it seems that the Microchip KB article I based these tables off of is way more misleading than I originally thought, and I didn't have and low speed device to test. I assume the overclock works properly then and the mouse is not bandwidth-starved and thus lagging or anything despite the speed (I mean you said it's stable and very smooth so yeah). And that your table is based on real testing up to these lowest values?

Thank you for the input! I will update the README later then, and have just this single table for both USB 1.1 speeds. Did lsusb show 1 Mbps rather than 1.5 though? It's only a pity I don't have constant access to an USB 2.0 device either to test the lowest possible rates for that one now...

[matejbrezner]

• Overclock success: ✅
• Device name: Sony Corp. DualShock 4 [CUH-ZCT2x]
• VID:PID: 054c:09cc
• USB Negotiated speed: Full Speed (12Mbps)
• bcdUSB: 2.00
• bInterval and Hz: 1, 1000 Hz
• Original bInterval and Hz: 5, 250Hz
• Benchmarked with: wireshark and libreoffice
• Benchmark result: "image below"

I OC'd my dualshock4 controller and tested its polling rate with wireshark by intercepting the packets sent from the PC to the controller. I then exported the timestamps of the packets and made a frequency chart in libreoffice. It works!