Like every other website on the planet, SmallNetBuilder uses cookies. Our cookies track login status, but we only allow admins to log in anyway, so those don't apply to you. Any other cookies you pick up during your visit come from advertisers, which we don't control.
If you continue to use the site, you agree to tolerate our use of cookies. Thank you!

Router Charts

Click for Router Charts

Mesh Charts

Click for Mesh Charts


The Ultra 4 was tested with our new NAS test suite running latest 4.2.13 firmware. As noted earlier, four Seagate Constellation ES 1 TB (ST31000524NS) drives were installed by NETGEAR. Tests were run with all four drives configured in RAID 0 for the default charts and X-RAID2 for the RAID 5 charts. All tests were run with 1000 Mbps LAN connections.

The Windows File Copy test RAID 5 write ranking for four-drive NASes is shown in Figure 6. The Ultra 4 measured 64.5 MB/s for write vs. 51.6 MB/s for the NVX, and 67.4 MB/s for the QNAP TS-439 Pro II.

Because the NAS benchmark system changed, the results for the Ultra 4 and TS-439 Pro II and NVX aren't totally apples-to-apples. The only thing I can do is note that all of the products in the chart, ranking above the Ultra 4, are D510 Atom based, except for the Sans Digital EN104L+, which uses an Atom 330. The NVX uses a 1.06 GHz Intel Tolapai SOC.

RAID 5 Windows File Copy write ranking - four drive NASes

Figure 6: RAID 5 Windows File Copy write ranking - four drive NASes

Figure 7 shows the RAID 5 File Copy read results. This time the NVX tops the charts with 98.6 MB/s, followed by the Ultra 4 with 94.9 MB/s and trailed by the QNAP at 90.3 MB/s.

RAID 5 Windows File Copy read ranking - four drive NASes

Figure 7: RAID 5 Windows File Copy read ranking - four drive NASes

Moving along to the new Directory Copy To and From NAS tests from the Intel NAS Performance Toolkit (NASPT), there is only one other four-drive NAS to compare, the D510 based Cisco NSS324, which is a QNAP TS-459 Pro in disguise. This test copies 236 MB consisting of 2,833 files in 44 folders to and from the NAS under test and causes a lot more head seeking than the highly sequential Windows File Copy test.

On the "To" test, the Ultra 4 lagged slightly behind the NSS324 at 13.5 MB/s vs. 14.7 MB/s. But for the "From" test, the Ultra 4 bested the NSS324 with 18 MB/s vs. 14.8 MB/s.

The NASPT Content Creation test is also a good NAS stress test. It simulates a user creating a web advertisement from various content sources including video and 3D modeling software. It consists of 95% writes with mix of 1k, 4k & little reads. There is also a wide range of writes up to 64 kB, mostly sequential.

Again, there is only data for the Ultra 4 and NSS324, with the latter turning in results almost 2X faster than the Ultra 4 at 10.3 MB/s vs. 6.3 MB/s.

Backup and iSCSI

I was able to run backup tests for a USB-attached Iomega UltraMax Pro Desktop Hard Drive configured in RAID 0 and formatted in FAT and NTFS. The Ultra 4's built-in USB drive formatter supports only FAT32 and EXT3 formatting, so I had to use a Windows system to do the NTFS formatting.

The Ultra 4 backed up at 24.8 MB/s for FAT, which is about the limit of the USB 2.0 connection. But it delivered only 16.5 MB/s with an NTFS formatted drive. Looks like the Ultra 4 could benefit from using the Paragon NTFS drivers that both QNAP and Cisco NASes now use.

Updated 9/21/2010

My attempts to back up to an EXT3 formatted drive were thwarted, however, by what appears to be a bug. I formatted the drive multiple times using the built-in formatter and made multiple attempts, even trying an unreleased firmware version supplied by NETGEAR. But each time, the Iomega drive sounded like the head was constantly seeking and I had to give up after 15 minutes had passed with only a few of the 1 GB VOB files copied.

NETGEAR was finally able to track down the source of the EXT3 backup problem. By default NETGEAR has EXT3 sync on every write vs. only after the file completion with FAT and NTFS. After enabling the fast USB writes option in the System->Performance tab and rebooting (there was no reboot prompt), EXT3 backup ran properly at 23 MB/s.

I also ran a network backup benchmark, using Delta Copy running on the NAS Testbed system as an rsync target. Again, I have only the NSS324 to compare, which with 39.9 MB/s, basically tied the Ultra 4's 39.2 MB/s.

Finally, I created a 10 GB iSCSI target on the xRAID2 volume and copied the 4.35 GB folder back and forth. iSCSI write ranking is shown in Figure 8, with the Ultra 4 at 64.5 MB/s doing much better than the NVX at 29.3 MB/s and a bit worse than the TS-439 Pro II at 71.4 MB/s.

iSCSI write ranking - four drive NASes

Figure 8: iSCSI write ranking - four drive NASes

iSCSI read ranking isn't shown, but the same relative rankings held, with TS-439 Pro II at 71 MB/s, Ultra 4 at 64.4 MB/s and NVX at 53.9 MB/s.

Do some more performance exploring with the NAS Charts.

Closing Thoughts

Updated 9/19/2010

I've pretty much covered the comparison points as I've gone along. But if you've skipped here just looking for the bottom line, here it is.

In a nutshell, if you're trying to decide between an Ultra 4 and NVX, choose the Ultra 4 unless you can get a really good deal on the NVX (the naked Ultra 4 is at least $50 cheaper) and can live without its business level features like NIC teaming / failover and AD support. And although I don't have the apples-to-apples data to support it, NETGEAR says its testing shows the Ultra 4 has a 10 to 15% edge on the NVX. But more importantly, the Ultra 4 (via the D410 Atom) has 64 bit support, which will allow for future volume support beyond 16 TB.

Support Us!

If you like what we do and want to thank us, just buy something on Amazon. We'll get a small commission on anything you buy. Thanks!

Don't Miss These

  • 1
  • 2