I have owned my QNAP TS-509 Pro for about 9 months now, and it performs very well. It is more than capable of streaming simultaneous movies to two different media streamers. I had read that there was a performance gain to be had, by giving it a simple memory upgrade.

Memory is cheap, so what the heck, here goes....................

WARNING: This will invalidate your warranty, I take no responsibility for your actions. It is completely at your own risk if you attempt this modification.

Firstly, the memory kit I selected consisted of 4GB (2x2GB Sticks), DDR2, 667MHz, 200-Pin SO-DIMM from Corsair.

Logging into the NAS as admin, and snapped a screenshot of the before config (note: this is the V3.1.1 0815T firmware), and you can see the Total Memory and Free Memory values.

Shutting down the unit, and flipping off the cover, you can see the old 1GB memory single stick in its slot. This was removed and the two new 2GB sticks were inserted. Chassis case back on and the unit powered up.

Old single 1GB stick top left	old stick removed	1st new stick in	2nd new stick in

After it went through its boot up process, I logged back in again, and took a new snapshot of the memory values now. As you can see this only reports around the 3500MB mark. This is of course due to the firmware running 32Bit Linux OS, and the memory address range is thus limited.

Next step was to run some benchmarks on the unit. I have been using the freeware application .Net 2.0 Disk Benchmark II by Ives Heymans as it is simple interface, and has the option to turn of caching and use unbuffered IO.

I ran the benchark nine times (3 @ 20MB data set, 3 @ 100MB data set, 3 @ 500MB data set) and averaged the results for each data set size. This was done for both 1GB and 4GB memory. The NAS is configured in RAID 5, with a single volume spanning 5x 1.5TB disks.

As you can see for the 500MB Set below, there is a clear performance advantage as a result of the memory upgrade. The 20MB and 100MB also showed the same percentage increases.

As with any benchmark, your results may vary. This is of course based on my hardware, network configuration etc.

So that is the first of the tweeks to my box. The next may be a reformat and upgrade to EXT4 filesystem and then some more benchmarks, but will have to wait and see!