BIWIN M560 1TB

The BIWIN M560 1TB is a DRAM-less PCIe 5.0 SSD. In PyNVMe3 testing, the sample delivered approximately 11.6–13.7K IOPS in full-drive, high-queue-depth random writes across five checkpoints. The first read after each of 100 sudden-power-loss recovery cycles took about 7 seconds. In the small-model, multi-queue read test, loading time increased at higher queue depths.

BIWIN M560 1TB | Test Report

Test Platform

Item Test configuration
Drive BIWIN M560 1TB; M.2 2280, PCIe Gen5 ×4, NVMe 2.0, DRAM-less
Sample One sample; serial number BSBK46150201505
Firmware BS3563VN
Platform NUC, Core Ultra 9 285H, 16 GB RAM; Ubuntu 26.04.1, Linux 7.0.0-38
Tools PyNVMe3 26.11; Quarch PPM for power measurement
PCIe Link 32 GT/s ×4
HMB 64 MB

Test Procedure

PyNVMe3 was used for TP1 protocol tests and TP2 benchmarks covering performance, stress, power-on time and related behavior. Power was measured with a Quarch PPM throughout.

Category Test ID Scope
Protocol and reset TP1 Core I/O, HMB, registers, TCG and controller reset functions.
Low-power states TP2.a Power consumption and I/O response in low-power states.
Performance and power TP2.b Sequential, random and mixed I/O performance, latency and supply power at different write-workflow checkpoints.
Stress and data verification TP2.c Sustained stress followed by readback verification of known data across the drive.
Sudden power loss and recovery TP2.d Sudden power loss during random I/O, recovery timing and data verification.
Normal power cycles TP2.e Normal shutdown and power cycles, with recovery-to-access timing.
Advertised performance TP2.f Peak sequential and random read/write performance in a small region.
Link speed switching TP2.g Dynamic Gen2–Gen5 link-speed changes under a read workload.
Model loading TP2.h Read synthetic model images of different sizes at various worker counts and queue depths.

Product Radar

Product Radar

This chart shows five aspects of the tested sample. Protocol covers functions related to NVMe and PCIe specifications; performance covers read and write capability; latency concerns response time; power covers electricity consumption; and other includes stress, power cycling and recovery. Each dimension is scored from 1 to 5, with a higher score farther from the center.

Performance Positioning

Performance Positioning

Each black dot places one drive according to two tests. Farther right means higher sequential write speed, such as when saving a large file; higher up means stronger random read performance, such as when reading data from different locations.

The rectangle around each dot represents longer response times in two workloads. Its width corresponds to mixed reads and writes during sequential writing; its height corresponds to random-write latency at the 90% workflow checkpoint. Smaller dimensions indicate shorter measured tail latency in the respective tests. Color represents power consumption during testing: blue is lower and red is higher.

Model Loading Time

This test simulates reading model images from an SSD and compares loading times at different queue counts and queue depths.

Model Loading Time

The four panels show small (4.7 GiB) and large (40 GiB) model images, each with a single queue and multiple queues. Queue depth on the horizontal axis is the number of read requests that each queue can issue at the same time; the vertical axis shows average read time. A lower line means the read completed sooner under those test conditions. Each point summarizes multiple measurements; the vertical bars show the observed range.

Key Issues

ID Test ID Description
M560-01 TP2.b Full-drive high-queue-depth random-write performance was low at all five workflow checkpoints and substantially below the tested G3 QLC sample.
M560-02 TP2.d The first read after each of 100 sudden-power-loss recovery cycles took approximately 7 seconds.
M560-03 TP2.h Loading time increased at higher queue depths with multiple queues.

Related products