|There are no user reviews for this product. [Why Not Add Yours?]|
I tested the DS412+ with DSM 4.1-2647 firmware, using our NAS test process with RAID 0, 5 and 10 volumes. As is our standard practice, four drives were configured in each volume type.
Windows File Copy tests were very consistent for RAID 0, 5 and 10, with the lowest result of 102 MB/s for RAID 5 read. Highest filecopy throughput was 107 MB/s from RAID 0 write.
Synology DS412+ DiskStation benchmark summary
Intel NASPT File Copy results were about 20 MB/s higher for writes across the RAID modes, but about 10 MB/s lower for reads. Highest NASPT write throughput was 129 MB/s for RAID 0 write and lowest read was 90 MB/s in RAID 5. Note that the highest possible throughput from a Gigabit connection is 125 MB/s, so some write cache effects are in the results.
iSCSI performance of 95 MB/s for write and 84 MB/s for read places the 412+ among the top 10 of all RAID 5 capable NASes tested for these benchmarks.
Attached backup tests were run with our new higher performance Startech USB 3.0 eSATA to SATA Hard Drive Docking Station (SATDOCKU3SEF) containing a WD Velociraptor WD3000HLFS 300 GB drive. Since Synology NASes have built-in FAT, EXT3 and EXT4 formatting for external drives, I was able to run a full set of backup tests.
The 412+ produced some of the highest attached backup throughputs measured to date, especially with the USB 2.0 connection. USB 2.0 usually produces results in the mid 20 MB/s range. But 31 MB/s was the lowest measured in that group. Highest attached backup throughput of 109 MB/s was obtained using eSATA and an NTFS formatted drive.
By comparison, Network backup to a DeltaCopy target on our NAS Testbed system was pretty slow at only 35 MB/s.
For a competitive look, I ran NASPT RAID 5 File Copy charts filtered for dual-core Atom processors. Note that RAID 5 (and 10) tests are run with four drives so as not to give an unfair advantage/disadvantage to NASes with more than four bays.
As we have seen previously, the latest-generation Dual-Core Atoms (D2700, D2550) don't seem to provide a significant performance boost over previous-generation D525 Atoms. In the chart below, the top three products and Thecus N4800 use the D2700, the Thecus N5550 uses the D2550 and the rest use the older D525.
With a 10% performance difference from the Thecus N4200PRO to the Synology DS1512+, you would probably not notice the difference in real world use. The outlier is the D525-powered Iomega px6-300d, which significantly lags in performance, as Iomaga NASes have consistently tended to do.
Synology DS412+ NASPT RAID 5 File Copy Write comparison
RAID 5 NASPT read throughput is slower across the group, with a 14% performance spread among all products except, again, the Iomega.
Synology DS412+ NASPT RAID 5 File Copy Read comparison
Use the NAS Charts to further explore and compare the DS412+'s performance
If you're shopping for the latest-generation high-performance four-bay NASes, it looks like you're going to spend at least $600 and that's without drives. The closest in performance and price to the DS412+ that we've tested is Thecus' N4800, which is currently only best-priced around $30 cheaper. But pricing varies all over the map, so shop carefully.
If you're looking to spend less, you're going to need to step down in processor power or look for discontinued models. But you'll also step down below 50 MB/s for read/write performance, too.
If you're looking to stay with Synology, but want to spend less, the new DS413 would be the logical choice, priced around $100 less. But we haven't been able to get one from Synology yet, so can't say what you'll be sacrificing for performance with the dual-core Freescale QorIQ P1022 PowerPC processor that powers it.
Related Items:Synology DS413 Reviewed
Synology Releases Hot-Swappable Four-Bay Atom NAS
Synology DS1512+ Disk Station Reviewed
Synology DS213+ DiskStation Reviewed
QNAP TS-269 Pro NAS Reviewed
NOTE! Please post product reviews from actual experience only.
Questions, review comments and opinions about products not based on actual use will not be published.