QNAP TVS-h1288X CPU Performance
Since a big part of QNAP’s value proposition is the ability to run applications and virtual machines on its NAS units, we wanted to share why the selection of the Intel Xeon W-1250 is such a differentiated choice. Having a higher clock speed processor means we get a lot of performance that we can use for NAS/ storage, but also for running virtual machines, containers, and applications.
Python Linux 4.4.2 Kernel Compile Benchmark
This is one of the most requested benchmarks for STH over the past few years. The task was simple, we have a standard configuration file, the Linux 4.4.2 kernel from kernel.org, and make the standard auto-generated configuration utilizing every thread in the system. We are expressing results in terms of compiles per hour to make the results easier to read:
Overall, we wanted to highlight a major difference between what QNAP is doing here and some of the competitive offerings. For example, Synology’s $2500 12x 3.5″ bay DiskStation DS3617xs uses the Xeon D-1527. We can see a massive difference in CPU performance here.
7-zip Compression Performance
7-zip is a widely used compression/ decompression program that works cross-platform. We started using the program during our early days with Windows testing. It is now part of Linux-Bench.
There are other offerings, such as the ZFS-based iXsystems TrueNAS Mini X+/ XL+ that use the Atom C3758 and one can see the clear benefit of the higher speed/ performance cores here. QNAP is effectively using a higher-performing CPU to enable better virtualization performance rather than just providing storage services.
OpenSSL is widely used to secure communications between servers. This is an important protocol in many server stacks. We first look at our sign tests:
Here are the verify results:
As a quick note, the Xeon D-1500 series was launched in 2015 and the Intel Atom C3000 launched in 2017. As a result, the 2020 Xeon W with a higher TDP is going to be a higher-performing part. Even if you just need basic file services, the newer platform has USB 3.2 Gen2 where the older platforms do not.
The storage performance of these systems is very dependent on the disk configuration including drives used, cache SSDs, compression/ deduplication, and even RAM/ other applications running on the NAS. We still wanted to show the impact of moving to 10GbE 10GbE as the performance over 2.5GbE was quite eye-catching.
Here is another view looking at the file transfer speeds.
Overall performance was good. It seems as though our Thunderbolt connectivity was more limited by array speed rather than interface speed.
Power Consumption and Noise
The power consumption is dominated by disks and add-in cards. We debated whether we should test this in our 208V data center lab or in our 120V embedded edge data center lab. In the end, we are using 120V edge lab at 17.7C and 70% RH. Our testing window shown here had a +/- 0.3C and +/- 2% RH variance.
- Idle empty: 100W
- Max observed: 172W
It seems as though the 550W power supply configuration is ample for this unit. We did get a bit of variance at idle but the range was more like 98-102W so we are using the average. This is a bit different than many of the workstations and servers we test where idle variances are within 1W.
In terms of noise, we saw the following in a 25dba ambient environment as measured from 1m in front of the chassis:
- Idle: 27dba
- Moderate Disk Activity: 31dba
The disk activity is probably what would make one put this a few feet/ meters away rather than directly next to their chair. Still, this is not a unit that most will find needs to be located in a data equipment closet.
Next, we are going to discuss the market impact of this solution, along with our final words.