Crucial P2 NVMe SSD Review

The Crucial P2 is the latest low-cost addition to the consumer portfolio of the firm’s NVMe SSD. Certainly, these types of drives are one of the bare-bone solutions on the consumer market at NVMe, as it is hardly a step up from the slower and less durable QLC NAND drives. Having said that, its price point makes it one of the most affordable out there, making the P2 perfect for those who are looking for an ultra-affordable way to upgrade their rig to an NVMe system.

As for what the P2 is actually capable of, Critical quotes up to 2,300MB / s read and 940MB / s write for the 500 GB model; this marks a potential increase (in reads) of 300MB / s over the previous P1 line, though writes remain the same. Life Expectancy (MTTF) is set at 1.5 million hours, whereas Endurance is set at 150 TB TBW for both capacities, the latter a noticeable increase over the 500 GB model of its predecessor.

The Crucial P2 SSD also comes with handy performance optimization management software, data security, and firmware updates, as well as Crucial’s Acronis True Image, the latter allowing users to clone drives, and backup operating systems, applications, settings.

Specifications of Crucial P2 NVMe SSD

Interface PCIe G3 1×4 / NVMe
Form factor M.2 2280
Capacities 250GB, 500GB
Performance Sequential Read (up to) – 2300 MB/sequential Write (up to) – 940 MB/s
Data Transfer Software Acronis True Image for Crucial cloning software
Advanced Features Active Garbage Collection

Integrated Power Loss Immunity

Dynamic Write Acceleration

Redundant Array of Independent NAND (RAIN)

Multistep Data Integrity Algorithms

Adaptive Thermal Protection

TRIM Support

Error Correction Code (ECC)

NVMe standard Self-Monitoring and Reporting Technology (SMART)

NVMe Autonomous Power State Transition (APST) Support

Life Expectancy (MTTF) 1.5 million hours
Endurance 250GB: 150 Total Bytes Written (TBW)500GB: 150 Total Bytes Written (TBW)
Warranty Manufacturer’s 5 Year Limited

Crucial performance at P2 NVMe SSD

Testbed

A Dell PowerEdge R740xd server is the test platform leveraged in those tests. We measure SATA performance inside this server by means of a Dell H730P RAID card, although we set the card in HBA mode only to disable the RAID card cache effect. NVMe is checked natively via an adapter card M.2 to PCIe. The approach used best represents the workflow of end-users with the testing of accuracy, scalability, and versatility within virtualized application offerings. A major focus is placed on drive latency across the drive ‘s entire load range, not only at the smallest QD1 (Queue-Depth 1) levels. We do so because many of the common consumer benchmarks do not adequately capture profiles of end-user workload.

Houdini by SideFX

The Houdini test is explicitly designed to measure the efficiency of the storage as it relates to CGI rendering. The testbed for this application is a variant of the core server type Dell PowerEdge R740xd we use in the laboratory with dual Intel 6130 CPUs and 64 GB DRAM. In this case, we have installed a bare metal running Ubuntu Desktop (Ubuntu-16.04.3-desktop-amd64). Benchmark output is measured to complete in seconds, with fewer being better.

The Maelstrom demo represents a section of the rendering pipeline that highlights storage performance capabilities by demonstrating its ability to make effective use of the swap file as an extended memory form. The test does not write out the result data or process the points to isolate the wall-time effect of the latency impact on the storage component underlying them. The test itself consists of five phases, three of which we run as part of the benchmark, the following are:

  • Loads of Disk Points packed. This is the fastest time to read. This is a single-threaded operation, which can limit the overall performance.
  • Unpacks the points into a single flat array, so they can be processed. If the points have no dependency on other points, the work set will be modified to stay in-core. That is a multi-threaded step.
  • (Does not run) Process points.
  • Repacks it into bucketed blocks suitable for back-to-disk storage. That is a multi-threaded step.
  • (Not run) Write back to the disk the bucketed blocks.

VDBench Workload Analysis

Application testing is best when it comes to benchmarking storage devices and synthetic testing comes in second. Although not a perfect representation of real workloads, synthetic tests provide reference storage devices with a repeatability factor that makes a comparison between competing solutions easy to do apples-to-apples. These workloads offer a range of different test profiles ranging from “four corners” testing, common size transfer testing of databases, to tracking captures from different VDI environments. All these tests leverage the common vdBench workload generator, with a scripting engine to automate and capture results over a large cluster of computing tests.

This enables us to repeat the same workloads across a wide range of storage devices, including flash arrays and storage devices individually. Our testing process fills the entire drive surface with data for these benchmarks, then partitions a drive section equal to 5 percent of the drive capacity to simulate how the drive might respond to application workloads. This is different from maximum entropy tests that use 100 percent of the drive and take it into a stable condition. Those figures will, therefore, reflect higher-sustained write speeds.

  • 4K Random Write: 100% Write, 64 threads, 0-120% iorate
  • 4K Random Read: 100% Read, 128 threads, 0-120% iorate
  • 64K Sequential Write: 100% Write, 8 threads, 0-120% iorate
  • 64K Sequential Read: 100% Read, 16 threads, 0-120% iorate

Final Thoughts:

The Crucial P2 is an entry-level drive designed for end-users who want to make very affordable use of the NVMe technology, specifically those who build a new system or upgrade their old SATA-based one. The P2 is more in line with the daily computing needs as compared to the new P5 model (which focuses more on the graphic designer and gamer demo). Furthermore, although its predecessor shipped with QLC, by equipping the P2 with TLC 3D NAND technology, Micron decided to go more on a performance-focused route It currently only comes in 250 GB and 500 GB capacities.

While the P2 is one of the weaker TLC-based consumer SSDs we’ve seen, Crucial ultimately achieved exactly what it set out to do: build an ultra-affordable NVMe drive that’s priced at the cost of writing efficiency. In any case, if you mount your OS on the Critical P2 NVMe SSD, you will surely see an overall increase in device efficiency compared to a SATA SSD or an HDD. If you value the cost above all else, or if you are on a strict budget, the P2 is definitely a viable choice as a starter drive.