In a 24/7 NAS environment, caching SSDs face relentless random I/O workloads that consumer-grade benchmarks fail to represent. Unlike desktops or client systems, NAS caching involves continuous small-block read/write activity under concurrent access, placing heavy demands on both sustained performance and endurance. To meet these challenges, Synology M.2 NVMe SSDs are designed not only for performance consistency, but also for enterprise-grade endurance and seamless system integration, ensuring reliable cache acceleration across long-term deployments.
Uncovering true SSD performance and longevity
Despite these real-world demands, SSD product datasheets often highlight impressive peak performance figures and high TBW ratings. The problem is that both are typically derived from consumer-oriented workloads that reflect short bursts or light-duty conditions, not the continuous pressure of NAS caching. This creates a misleading picture of capability. The true measure of an SSD is revealed under enterprise-grade workloads, where sustained performance and endurance are tested by continuous high-frequency 4K random I/O operations running 24/7 without pause.
Staggering performance gaps under stress
To accurately evaluate real-world performance under sustained workloads, we conducted rigorous testing following Storage Networking Industry Association (SNIA) Performance Test Specification (SSS PTS) methods for preconditioning and steady-state performance measurement. With this methodology, we benchmarked our SNV5420-1600G against competitors using rigorous enterprise workloads. As shown in Chart 1, the SNV5420-1600G offers significantly higher sustained 4KB random write IOPS, outperforming the Seagate Nytro 4350 1.92T enterprise drive by 3.7×, the WD Red SN700 2T NAS-grade drive by 16.9× and the WD Blue SN5000 2T consumer-grade drive by 57.1×1,4.
Chart 1: Sustained 4KB random-write IOPS measured after SNIA SSS PTS preconditioning to steady state.
Many drives maintain respectable random IOPS during initial Fresh-Out-of-Box (FOB) phases, but their performance quickly drops once the SLC cache2 is saturated and background activities such as garbage collection are triggered. For instance, both the NAS and consumer drives from Western Digital saw a 90% decline in IOPS after a single full-device write, revealing limitations for continuous-use caching scenarios. Even the Seagate Nytro 4350 1.92T, an enterprise-tier SSD, showed little resilience, with an 87% decline in IOPS as the workload intensified1,4.
Chart 2: IOPS over write size. SNV5420 maintains high performance, while others drop rapidly after SLC exhaustion. X-axis represents cumulative written data as a percentage of the drive’s capacity. For example, 100% means the equivalent of one full drive write, 200% means two full drive writes, and so on.
By contrast, the SNV5420 drive shows a controlled and gradual decline before settling into a stable high-performance plateau, maintaining consistent random IOPS under sustained write workloads. This level of performance stability is critical in enterprise NAS environments, where caching must remain responsive over time to support real-world access patterns.
While frequently emphasized in marketing materials, FOB results only reflect a brief period of peak performance and fail to capture long-term behavior under continuous stress. In contrast, analyzing sustained performance with realistic enterprise workloads provides a more accurate evaluation of an SSD’s suitability for NAS use. This approach helps users distinguish between drives optimized for marketing benchmarks and those truly engineered for operational endurance and consistency.
TBW ratings: Key differences in enterprise vs. consumer testing
Beyond sustained performance, endurance is a crucial factor for the long-term viability of SSDs in NAS cache environments. However, not all endurance ratings are created equal. Consumer-grade SSDs often advertise high Terabytes Written (TBW) ratings based on consumer-oriented workloads. These figures may look impressive, but they fail to represent the sustained, high-intensity stress of real-world NAS caching.
Relying on these figures is risky, as internal testing reveals a stark contrast between how drives perform under consumer versus enterprise conditions:
- A consumer-grade SSDs reported TBW can be 3 to 5.4 times higher under lighter consumer workloads, as typical desktop usage masks the true rate of wear (Chart 3).1,3
Chart 3: Consumer-grade SSD endurance (TBW) under enterprise vs. client workloads.
- When subjected to an identical, demanding enterprise workload, an enterprise-grade SSD can withstand 5 to 10 times more write volume than a consumer-grade SSD before failing (Chart 4).1,3
Chart 4: Endurance comparison of consumer vs. enterprise-grade SSDs under enterprise workloads (TBW).
Using a drive that’s not designed for continuous 4K random IO workloads can lead to accelerated wear, performance degradation, and a much higher risk of premature drive failure and data loss.
Synology M.2 NVMe SSDs are tested with the JEDEC JESD219A (Enterprise) workload profile and evaluated under JESD218A endurance conditions (e.g., 24×7 @ 55 °C, full-space utilization, no TRIM). This better reflects the sustained random read/write mix and concurrency typical of NAS caching. Table 1 below summarizes the difference between enterprise and consumer (client) workloads in accordance to JEDEC.
Enterprise workload | Consumer workload | |
Operational window | 24h/day @ 55°C | 8h/day @ 40°C |
UBER | ≤ 1e-16 | ≤ 1e-15 |
Coverage | Full drive (100% user LBA) | Localized/partial footprint |
Idle block reclamation (TRIM/UNMAP) | Disabled for sustained performance test | Enabled for typical client usage test |
Table 1: Major testing differences between enterprise and consumer testing workloads.
System-level integration
Our system-level integration provides tangible advantages for management and reliability. Supported directly within DSM, our SSDs offer:
-
Unified health analytics: Monitor drive health and receive automated error reports without installing third-party tools.
-
Seamless firmware updates: Receive and apply firmware updates directly through DSM, ensuring your drives are always optimized.
-
Verified compatibility: Eliminate guesswork with drives that are tested and guaranteed to work with your Synology system.
To further protect your data against all eventualities, Synology 5400 series SSDs are also equipped with hardware-level power-loss protection, safeguarding data integrity during unexpected power failures.
Another aspect that is often neglected is thermal stress. Intense workloads generate heat, which affects performance and longevity. Synology performs comprehensive thermal testing on our SSDs inside every compatible Synology system. This validation ensures predictable performance and longevity under rated operational conditions.
A trusted foundation for demanding environments
For businesses and professionals who prioritize durability and consistent performance, Synology M.2 NVMe SSDs deliver a reliable foundation. Engineered for demanding environments, these drives are a trusted part of a unified system, validated through extensive testing to handle sustained workloads and scale reliably.
Synology M.2 SSD lineup
SNV3410-400G | SNV5420-400G | SNV5420-800G | SNV5420-1600G | |
Capacity | 400 GB | 400 GB | 800 GB | 1600 GB |
Form factor | M.2 2280 | |||
Interface | NVMe PCIe 3.0 x4 | |||
Sustained sequential read (128 KB, QD32)4 | 3,000 MB/s | |||
Sustained sequential write (128 KB, QD32)4 | 750 MB/s | 650 MB/s | 1,000 MB/s | 1,000 MB/s |
Sustained random read (4KB, QD256)4 | 225,000 IOPS | 225,000 IOPS | 400,000 IOPS | 660,000 IOPS |
Sustained random write (4KB, QD256)4 | 45,000 IOPS | 45,000 IOPS | 70,000 IOPS | 120,000 IOPS |
Terabytes written (TBW)5 | > 490 TB | > 700 TB | > 1,400 TB | > 2,900 TB |
Power loss protection | No | Yes | Yes | Yes |
Warranty6 | 5 years | |||
Recommended models7 | Plus series | SA / XS+ / XS / Plus series | SA / XS+ / XS / Plus series | SA / XS+ / XS / Plus series |
Notes:
- All M.2 SSD performance & endurance results are based on single-drive results obtained from Synology’s testing environment using a DS3018xs+ with E10M20-T1 adapter card.
- SLC cache (Single-Level Cell cache) is a technique used in many modern SSDs to temporarily store data in a faster portion of the NAND flash before writing it to the main storage area.
- The endurance differences are based on tests conducted in accordance with the JEDEC JESD218A endurance methodology, using both Enterprise and consumer workload patterns as defined in JESD219A.
- All performance measurements refer to sustained performance achieved in a steady state, as defined by test specifications in accordance with the Storage Networking Industry Association (SNIA).
- The TBW specifications denote the baseline endurance levels for Synology SSDs and are calculated using the JESD219A enterprise workload.
- 5-year limited warranty provides coverage until the end of the warranty period or until the endurance usage of the drive has been reached, whichever comes first. Learn more
- Supported drive interfaces and compatible drive models may vary by Synology system. Please refer to the product specification page and compatibility list for details.