Before we jump into the real-world workloads, we'll run a handful of synthetic tests. This helps us understand each product's strengths and the areas where performance could be improved.
In all but one benchmark, we linked a single gigabit Ethernet port from the device being tested to a Netgear S3300-52X 100/1000/10,000 switch. We used a Quanta MESOS CB220 server on the other end to administer the workloads. The MESOS CB220 connects to the network via 10GbE.
With all of tested platforms attached through a single GbE connection, performance is hampered by available network bandwidth.
The fact that appliances like these use such tightly integrated SoCs with host processing, memory control and I/O on one die, and still operate under low power ceilings is nothing short of amazing. Asustor's AS7004T employs Intel's desktop-class Core i3, giving us a good point of reference to compare more modest platforms.
Sequential Mixed Workloads
The line graphs show the percentage of read operations in each benchmark. Our results start with 100 percent reads and transition to 100 percent sequential writes. We highlight the 70 percent read/30 percent write combination specifically because it's regarded to be a mix typical of professional workstations.
Using the taxing mixed workloads, we see some separation between the products we're comparing. QNAP's TS-563 rips through these metrics with higher performance than the more expensive AS7004T.
The TS-563 also dominates the random read tests at low queue depths. Asustor's AS7004T picks up the pace after a queue depth of two thanks to aggressive cache algorithms. But other than the random read test at high queue depths, which the AS7004T wins, QNAP's TS-563 is faster.
It also performs really well in the random data write tests, achieving performance that's higher than the other products in this discipline as well.
Random Mixed Workloads
With a performance lead in both 100 percent random reads and random writes, the TS-563 also leads the other products in random mixed workloads.