Hello everyone,
I’m new to the QNAP community and am looking forward to exchanging ideas with you all.
Currently, I’m having an issue with the performance of my 10 GbE network connection at home. I hope that the experienced members here can give me some pointers or tips for troubleshooting.
Thanks a lot in advance.
My setup:
PC:
Processor 11th Gen Intel(R) Core™ i9-11900K @ 3.50GHz, 3504 MHz, 8 cores, (RAM) 32.0 GB
BaseBoard product ROG STRIX Z590-A GAMING WIFI
Network: Marvell AQC113 10Gbit network card.
Windows 11 Pro 25H2.
Switch: Unifi USW Pro XG 10 Poe with Cat6A cabling.
NAS: NAS-TS-765eU with QXG-ES101T 10gb network card; 2 x Samsung 990 Pro SSD RAID1 (Thick Volume) system pool without cache; 2 x Seagate Ironwolf Pro RAID1 (Thin Volume); QTS 5.2.9.3499.
PC > Iperf3 > NAS-TS-765eU
Iperf3 -c -P 4 > 9.45 Gbit/s - multi stream
Iperf3 -c -P 4 -R > 9.52 Gbit/s - multi stream
Iperf3 -c > 9.29 Gbit/s - single stream
Iperf3 -c -R > 5.05 Gbit/s - single stream
CrystalDiskMark 8.0.5
- MB/s = 1,000,000 bytes/s [SATA/600 = 600,000,000 bytes/s]
- KB = 1000 bytes, KiB = 1024 bytes
[Read]
SEQ 1MiB (Q= 8, T= 1): 700.984 MB/s [ 668.5 IOPS] <11942.08 us>
SEQ 1MiB (Q= 1, T= 1): 556.966 MB/s [ 531.2 IOPS] < 1881.48 us>
RND 4KiB (Q= 32, T= 1): 211.597 MB/s [ 51659.4 IOPS] < 615.48 us>
RND 4KiB (Q= 1, T= 1): 37.642 MB/s [ 9189.9 IOPS] < 108.09 us>
[Write]
SEQ 1MiB (Q= 8, T= 1): 1128.085 MB/s [ 1075.8 IOPS] < 7416.88 us>
SEQ 1MiB (Q= 1, T= 1): 616.287 MB/s [ 587.7 IOPS] < 1699.42 us>
RND 4KiB (Q= 32, T= 1): 196.406 MB/s [ 47950.7 IOPS] < 664.47 us>
RND 4KiB (Q= 1, T= 1): 35.197 MB/s [ 8593.0 IOPS] < 116.11 us>
Profile: Default
Test: 1 GiB (x3) [Y: 12% (157/1313GiB)]
Mode: [Admin]
Time: Measure 5 sec / Interval 5 sec
Date: 2026/06/14 13:29:47
OS: Windows 11 Pro 25H2 [10.0 Build 26200] (x64)
My question is: What do I need to configure so that the single stream performance (-R) improves? Is that even technically possible?
In multi-stream tests, I’m already reaching nearly the full 10 Gbit/s speed. However, single stream performance is much lower. So I’m interested to know whether there’s still potential for optimization here or if these values are normal for my setup.