
The Bottom Line
Pros
- + Mixed workloads
- + Random
- + Sequential
Cons
- - Ineffective Secure Erase
Should you buy it?
AvoidConsiderShortlistBuyIntroduction and Drive Details
A new contender has risen on the enterprise storage front. Seagate's newly minted Nytro 5550H is a PCIe Gen4 3-DWPD (Drive Write Per Day) mixed workload specialist, offering overall the highest sequential and random throughput we've encountered from any flash-based PCIe Gen4 SSD. To bring forth its first PCIe Gen4 enterprise offering, Seagate has unsurprisingly partnered with Phison Electronics.
The Nytro 5050 series is not only the first of its kind enterprise SSD for Seagate, but it is the first of its kind for Phison as well, in that that this drive (X1 for Phison/Nytro 5050 for Seagate) are both company's first PCIe Gen4 enterprise SSD. At the heart of the Nytro 5050 series SSDs is Phison's PS5020-E20, or E20, sixteen-channel controller. Phison's 16-channel E20 controller sits in front of a hynix 128L eTLC flash array.
Phison advertises that its E20 CPU complex is composed of two performance and power-efficient ARM R5 CPUs and dozens of small CPU co-processors that complete computationally heavy, redundant tasks at high speed with minimum power consumption.
The controller company further states that the X1/Nytro 5050 SSD platform was created in partnership with Seagate and Phison's engineering teams collaborating on the architecture, features, and development of the X1/Nytro 5050 SSD solution. Additionally, Seagate's drive validation lab performed extensive systems and environmental testing to ensure the SSD has world-class reliability.
Well, let's dive in and see exactly what the Nytro 5550H 6.4TB can do for you by the numbers.
Specs/Comparison Products


The Nytro 5550H is offered in the ubiquitous 2.5"x 15mm U.3 form factor. The drive is offered at five capacity points, with three variants at each capacity point. Additionally, the drive is dual port enabled.
Seagate Nytro 5550H 6.4TB NVMe PCIe Gen4 x4 U.3 SSD


Enterprise Testing Methodology
TweakTown strictly adheres to industry-accepted Enterprise Solid State Storage testing procedures. Each test we perform repeats the same sequence of the following four steps:
- Secure Erase SSD
- Write the entire capacity of SSD a minimum of 2x with 128KB sequential write data, seamlessly transition to the next step
- Precondition SSD at maximum QD measured (QD32 for SATA, QD256 for PCIe) with the test-specific workload for a sufficient amount of time to reach a constant steady-state, seamlessly transition to the next step
- Run test-specific workload for 5-minutes at each measured Queue Depth, and record results

Benchmarks - Random and Sequential
4K Random Write/Read

We precondition the drive for 16,000 seconds, receiving performance data every second. We plot this data to observe the test subject's descent into steady-state and to verify steady-state is in effect as we seamlessly transition into testing at queue depth. Steady-state is achieved at 2,100 seconds of preconditioning. The average steady-state write performance at QD256 is approximately 515K IOPS. We note the rather unusual pre-con pattern at the back half of the run. Unusual? Yes. Problematic? No.


We hit a max of 516K IOPS sustained, or 16% higher than the factory up to spec of 445K IOPS. In terms of pure 4K random write performance, our test subject is offering the best for a Gen4 SSD we've tested to date.


Seagate advertises up to 1.7 Million IOPS here, and we find that to be spot on, with our test subject delivering an average of 1.779 Million IOPS at QD256. This is overall the highest we've ever extracted from a flash-based PCIe Gen4 SSD. Impressive. The performance curve is excellent and is the best we've encountered for a PCIe Gen4 interfaced SSD at queue depths of 64 and higher.
8K Random Write/Read

We precondition the drive for 16,000 seconds, receiving performance data every second. We plot this data to observe the test subject's descent into steady-state and to verify steady-state is in effect as we seamlessly transition into testing at queue depth. Steady-state is achieved at 4,000 seconds of preconditioning. The average steady-state write performance at QD256 is approximately 242K IOPS.


We expect 8K random to track exactly the same as 4K random here, just at a lower rate because it's moving twice the amount of data. We find performance here to be very impressive, especially at queue depths of 4 and 8, where we are getting by far the best numbers we've encountered to date from a flash-based PCIe Gen4 SSD. In fact, our test subject is even able to deliver more at QD4 than Kioxia's PCIe Gen5 CM7-V. Impressive.


Pure 8K random read performance is again among the best we've encountered for a PCIe Gen4 SSD. Again, it is the best we've seen from an SSD of its kind at queue depths of 128 or higher.
128K Sequential Write/Read

We precondition the drive for 6,500 seconds, receiving performance data every second. Steady-state for this test kicks in at 0 seconds. The average steady-state sequential write performance at QD256 is approximately 6,867 MB/s.


Seagate specs our test subject as capable of up to 7,200 MB/s sequential 128K writes. We are getting close to 6,900 MB/s on average. We are of the opinion that 7,200 MB/s is likely a typo carried over from Phison's X1 SSD, which is, for all intents and purposes, the very same SSD. Nevertheless, our test subject again delivers the best we've encountered to date from a PCIe Gen4 SSD at queue depths of between 1-4, which is arguably where performance matters most. Impressive.


In terms of maximum throughput, our test subject delivers as advertised and then some, attaining a maximum throughput of 7,479 MB/s, or the third best we've encountered from a Gen4 SSD. We like what we are seeing at queue depths of 16 and higher, but we would like to see better at queue depths below 16.
Benchmarks - Workloads
4K 7030
4K 7030 is a commonly quoted workload performance metric for Enterprise SSDs.

We precondition the drive for 16,000 seconds, receiving performance data every second. We plot this data to observe the test subject's descent into steady-state and to verify steady-state is in effect as we seamlessly transition into testing at queue depth. Steady-state is achieved at 5,400 seconds of preconditioning. The average steady-state performance at QD256 is approximately 692K IOPS. Again, we are seeing a rather unusual IO pattern, but this time it's only for one drive write of our pre-con run. From 5,400 seconds on out, the drive settles into what we would characterize as an extremely high QOS IO distribution. This is exactly what we want to see from any enterprise SSD. Excellent.


Mixed workload performance is performance that matters, and overall we are impressed with what our test subject is able to deliver, especially at queue depths of 64-128, where we are yet again getting the highest performance we've encountered to date from a flash-based PCIe Gen4 SSD.
Email Server
Our Email Server workload is a demanding 8K test with a 50 percent R/W distribution. This application gives a good indication of how well a drive will perform in a write-heavy 8K workload environment.

We precondition the drive for 16,000 seconds, receiving performance data every second. We plot this data to observe the test subject's descent into steady-state and to verify steady-state is in effect as we seamlessly transition into testing at queue depth. Steady-state is achieved at 2,400 seconds of preconditioning. The average steady-state performance at QD256 is approximately 249K IOPS.


Our test subject is demonstrating above-average performance at low queue depths and, once again, the best we've encountered from a PCIe Gen4 SSD at QD32.
OLTP/Database Server
Our On-Line Transaction Processing (OLTP) / Database workload is a demanding 8K test with a 66/33 percent R/W distribution. OLTP is online processing of financial transactions and high-frequency trading.

We precondition the drive for 16,000 seconds, receiving performance data every second. We plot this data to observe the test subject's descent into steady-state and to verify steady-state is in effect as we seamlessly transition into testing at queue depth. Steady-state is achieved at 4,400 seconds of preconditioning. The average steady-state performance at QD256 is approximately 312K IOPS. As we observed with our 4K 7030 precon, the drive drops a few extreme outliers until it is filled once. After which, it settles into a steady-state pattern that indicates exceptional QOS.


Very similar to what we just observed with the previous test in that our test subject is delivering well above average at queue depths of up to 16, where it then shifts into overdrive and delivers the most we've encountered from a flash-based PCIe Gen4 SSD at QD32-QD64. Good stuff.
Web Server
Our Web Server workload is a pure random read test with a wide range of file sizes, ranging from 512B to 512KB at varying percentage rates per file size.

We precondition the drive for 16,000 seconds, receiving performance data every second. We plot this data to observe the test subject's descent into steady-state and to verify steady-state is in effect as we seamlessly transition into testing at queue depth. We precondition for this test with an inverted (all-write) workload, so no relevant information can be gleaned from this preconditioning other than verification of steady-state.


We consider this test to be the most taxing test we run. Our test subject delivers brilliantly here as well, by delivering what we would consider the second-best overall performance curve we've encountered from any flash-based PCIe Gen4 SSD. Additionally, Seagate's newest is delivering more PCIe Gen4 lab records at QD128-256.
Final Thoughts
As mentioned, Seagate's 5050 series or Phison's X1 SSD is the first of its kind for both storage companies. Overall, the 3-DWPD version has demonstrated itself to be one of the most performant SSDs of its kind we've ever encountered. At almost every turn, the Nytro 5550H 6.4TB delivered a lab best at one or more tested queue depths. That's impressive, especially considering this drive represents a first for both companies.

Now, we did encounter some never before seen pre-con patterns and even some extreme outliers, all of which turned out to be non-consequential in a negative way. It's just different than we've ever observed before. One other anomaly we encountered that we don't typically see is that secure erasing the drive proved completely ineffective at resetting the drive.
Only after filling the drive sequentially at least once did the drive reset to deliver the expected performance. Whether or not this is much, if any, issue at all as it relates to real-world applications, we are not sure but tend to think it is not.
The Seagate Nytro 5550H 6.4TB is one of the best-of-its-kind SSDs we've ever encountered, and as such, we find it worthy of one of our highest awards.