Affiliate Disclosure: This post may include affiliate links. If you click and make a purchase, I may earn a small commission at no extra cost to you.
Test Bench Specifications
- CPU: Intel Core i5 12600K
- Memory: Crucial DDR5 4800 MHz 16GB
- Motherboard: MSI Pro Z690-A
- Graphics Card: NVIDIA GeForce GTX 1650
- Storage: Silicon Power UD90 1TB
- Operating System: Windows 11 Pro (Version 10.0.26100 Build 26100)
Benchmark Scores
These benchmarks are performed with the drive 20% filled. All the tests are done multiple times, and either the average of the maximum scores has been published. The drive has been installed on the main M.2 NVMe Gen 4.0 slot in the motherboard (nearest to the CPU).
PCMark 10
PCMark 10 tests your system’s overall performance using trace-based benchmarks. It simulates real-world tasks like browsing, video editing, file copying, office work, etc. You can run either a full or a quick benchmark, but we are seeing the results for the Full System Drive benchmark only.
| PCMark 10 Score | Bandwidth | Latency |
|---|---|---|
| 2956 | 475.09 MB/s | 57µs |
The PCMark 10 score of 2956 is an average performance, but still enough for basic to medium content creation and gaming tasks. Bandwidth is the amount of data the SSD can read or write in a second. In this case, 475 MB/s is a decent speed but is somewhat lower compared to top-end SSDs that typically achieve sequential read/write speeds of over 1,000 MB/s or more.
Latency is how quickly the SSD can respond to read or write requests. A latency of 57µs is quite high as compared to other SSDs in the competition, but still acceptable.
CrystalDiskMark 8
We have done Sequential 1M Q8T1 and Q128T1 tests to test the sequential performance. For random tests, we have selected 4KB Q8T16 and Q1T1. So, we have tried to test the sequential speed in a single thread to simulate tasks like game installation, and then tested with a much deeper queue of 128 to stress the SSD’s controller and interface.
For random tests, 4KB at Q8T16 is used to simulate heavy I/O scenarios and multitasking. However, 4KB at 1QD and single thread simulate every data task, like launching programs and OS boot. You can see the results below.
AS SSD Benchmark 2.0.7
AS SSD Benchmark is special because it doesn’t use data compression during tests. It gives a true picture of the real-world performance in this way. The parameters work similarly. The sequential read/write tests come first. Then comes the 4K test for everyday OS tasks, then comes the 4K at 16 threads for simulating heavy multitasking. Access time measures how long it takes (in milliseconds) to read or write a small data block. Then this software gives an overall performance score, which is good for comparisons.
I have also done the ISO, Program, and Game speed and duration test.
ATTO Disk Benchmark 4.01.0f1
ATTO Disk Benchmark is a great tool to test a drive’s read and write performance using a wide range of block sizes, from very small (e.g., 512 bytes) to large (up to several megabytes). It provides a clear picture of how a drive performs across different file sizes, which is crucial because storage devices often behave differently depending on the size of the data being processed.
Anvil Pro
Anvil’s Storage Utilities (Anvil Pro) is designed to evaluate SSD and HDD performance across a wide range of scenarios. It measures both sequential and random read/write speeds, as well as response times, IOPS, and compression handling. One of its key strengths is its detailed IO test suite, which simulates real-world usage patterns such as multitasking and varying queue depths. At the end, you get both the read and write scores with the overall score.
AJA System Benchmark
High-Throughput Sequential Test
This benchmark is ideal for understanding how the SSD handles large, continuous video files like those used in professional editing. It helps gauge performance under heavy media workloads.
Mixed Workload Test
These AJA setting depicts a general editing or media workstation environment. It gives insight into performance under mixed workloads and medium file sizes.
Small File/Low Latency Test (Proxy Editing or OS Tasks)
These AJA settings focus on low-latency performance and how well the SSD deals with small files and quick access patterns. Useful for evaluating how the drive might perform when used for system tasks, cache storage, or proxy editing.
Temperatures and Thermal Throttling
I wrote around 400GB of data on the drive using the heatsink that came with the motherboard. Then I tested without any heatsink at all. The room temperature was around 28°C. With the heatsink, the SSD idles at around 48°C and goes up to 52°C maximum.
Without a heatsink, the drive reached the maximum temperature of 75°C and an average of around 72 °C. It didn’t reach the throttling point even though I tried a lot by blocking the airflow to the case as well.
However, this is SMART data, and I don’t have access to a thermal gun. But, I got pretty assured that the throttling point is very hard to reach. Maybe in summer, with very high ambient temperatures, the throttling point can be reached.
