SEAGATE EXOS X16 REVIEW

alt text

Since its inception, Seagate’s Exos X series was, is, and probably always will be, their premier Enterprise 7200RPM 3.5-inch hard disk drive line. As such every ounce of knowledge, experience, and trick Seagate’s engineers have learned since Seagate released their first enterprise hard drive is distilled into this special series. Put simply, the Exos X series is meant for massive server farms where hard drives are not counted in ones or tens, they are counted in the hundreds and even thousands… and the number of different duties they are expected to handle is nearly legion. The latest edition to the Exos X model is the Exos X16 and it should also come as no surprise that it also comes with an asking price that is more in line with a ‘everything including the kitchen sink’ design philosophy that the Exos X series it built around.

alt text

By the same token, an asking price of $629 is not all that outlandish for a 16TB capacity hard drive. It may be nearly $150 more than the $480, SOHO NAS oriented, IronWolf 16TB drive (or 3.9 cents per GB vs 3 cents per GB), but this model is the pinnacle of performance, engineering, durability, and even adaptability. So yes, a flagship capacity model that is part of the premier enterprise storage series is expected to cost more… but you get a lot more as well. On the performance front buyers can routinely expect empty drive performance in excess of 260MB/s, with small file random IO performance that we could hardly believe during testing.

Equally important, and as the name suggests, this hyperscale orientated hard disk drive model comes with a whopping 16TB of capacity, yet only consumers an average of 6.3 watts (which matters when talking about thousands of drives in a server room). A five year, 24/7/365 warranty, backstopped by a year work load rating of 550TB – or over three times that of the standard IronWolf 16TB model. It even boasts (optional) automatic hardware level encryption that promises to not impact overall performance. If all that was not enough it also comes with uber advanced features such as PowerChoice where you can configure how deep the idle state it enters is, when it enters this state (or if at all), and can even set up four different idle states to give you complete control over reduced power vs. response time.

While we honestly doubt many home users will ever have the privilege of using this drive series, this is perfectly acceptable. They are not meant for us mere mortals. Instead this drive series is meant for the likes of Amazon, Google, IBM… even the NSA. This is why the Exos X series is not part of the Seagate Guardian series. Instead it resides above that line-up at the very top of Seagate lines. This is however a great way to showcase precisely how far Seagate has come, and how far they have pushed the performance boundaries of what was once thought impossible from a mere 7200RPM hard disk drive. So even if you are not the intended buyer, we hope you enjoy this in-depth look at what Seagate can offer when you give them a darn near unlimited budget.

Full Review

Facebook

Hello, I also have the Exos X16, and it makes this uncommon sound during startup. Did you have the same sound in your sample (at about 00:10)? Recording here.

Ours were mostly silent on our open bench test rig, and are silent in our server. The only time they get loud is with random small file, deep queue depth IO requests where the arms have to move back and forth quickly to keep up.

Few things it could be. Firstly I would shut down the server and gently, but firmly, press down on the back-plane. Could simply be not perfectly in place and randomly throwing data/power errors. HP servers are notorious for the backplane board vibrating out of alignment. Next I would make sure your raid controller is not misconfiged with a wonky staggered startup routine. Could be as simple as it is powering up, then told to shut down and then quickly told to start up again.

IF you have it in a PC or server without hotswap backplane... change the data cable and power supply. HDDs will get noisy if there is dirty power coming in on the 12v or 5v line (ripples/spikes/etc).

Though the first thing I would do is ask what is being accessed on startup on that drive? Sounds like it could simply be aggressive sector seeking with random IO requests (eg OS drive). If its just a 'data' drive... run the long diagnostics on it. Could have a batch of bad sectors that have been reallocated. This would cause the drive arms to swing from inside edge to outside edge... aggressively, and makes more noise. Still might be in spec, at which point... defrag the drive. Get all the parts of the files 'together'.

This is what I do with a new HDD. I run SMART test, write down all the raw data numbers. Then I do a long format (expect it to take a long time). Then I re-run smart. If the numbers skyrocket... RMA. You got a bad drive. HHDs follow a bathtub curve of failure. Most happen in 90 days or then after 3 years. This reduces the 90 day wonders from YOUR system.

last edited by GaK_45

Thank you so much for your detailed response!

A few notes, to clarify:

  • I am using it in a PC, not a server.
  • This sound is resulting from the HDD power up sequence, and it sounds exactly the same even if no data cable is attached at all, about 10 seconds after power has been applied.
  • I tried with a different power supply, no difference whatsoever.
  • It sounded like this since the first time I powered on the drive, with it being new/empty. So I doubt it is relocated bad sectors.
  • The sound/vibrations heard in my recording, are probably being amplified a bit by the case, which is rather small.

From what I understand, your sample doesn't exhibit any similar seek (or whatever that sound is), which starts about 10 seconds after it has been powered and lasts for about 2 seconds? Think of something similar, but less amplified?

Thanks again!

last edited by asava

Just to make sure I just power cycled both and both were pretty darn silent on POST. Typical hum from any big hdd, but not noticeable unless really, really listening for it.

BUT I then yanked and placed each in a cheap 5.25 to 3.5 adapter (startech 20 dollar job) and the drives did become audible. Did this with a couple other hdds I had kicking around ( a wd 8tb, couple ironwolf pro 10tbs, the iron worlf 16tb)... all become noticeable.

Sooo could be as simple as the case you have them in. Try shutting down the rig. Pulling the drive BUT keeping both sata and power cables connected (leave your side panel off). Then place it on a neoprene 'mouse pad' w/ anti-static bag it came in between the pad and the drive (unlikely ESD will happen but better safe than sorry). See if the noise decreases or increases. If it decreases or changes tone/pitch... its your case acting as an echo chamber.

IF you can not do that, do you have an external 3.5 to usb adapter you can try?

Either way if you think something is wrong with the drive run the extended seagate diagnostics on it. It will tell you one way or the other!

Hope that helps 🙂

last edited by GaK_45

Thank you once again for your reply!

Well, I tested your suggestion, noise does decrease, but the noise it produces on power up is still noticeable and uncommon based on what I heard before... But then again, I did not encounter any other 8GB+ HDD, so maybe it's just me...

Would it be too much if I ask you to add also a recording of the sound emitted from your Exos X16 when it powers up? If it is too much, I totally understand, but if you can, it can be a good base for comparison 🙂 Even if recording it only in an open rig.

All the best and thanks again for all your help!

No problem. Happy to help.
One more quick question. Have you downloaded AS-SSD or Crystal DiskMark and run the small file portions of the tests? It would not surprise me IF the sounds you are hearing on POST you will hear during stress testing. If so, what I would suggest is phoning Seagate support and getting them to listen to it. I would first do the long seagate diagnostics on the drive - they are going to ask for this. Tell them its on POST and during random IO work loads as well.
That way you can get THE experts opinion on it.

Personally if it passes all the tests, is working, and only happens on POST... I probably would not care too much. It does not sound like a bucket of bolts being dropped down a flight of stairs. Just a touch louder (based on the recording... ie I don't know what it sounds like 'in person'... maybe you have better hearing in certain frequencies than myself!) than you would like. It is the highest performance 7200RPM drive Seagate makes so they are tuned for speed/performance not noise.

If they say the same thing they probably can walk you through modifying the AAM (Auto Acoustic Management) settings which will reduce noise at the expense of performance. EG walk you through changing it from 254(perf) to 128(low noise).

Yes, the noise happens only on POST (and when the HDD spins up again after waking up from sleep state).
It does not sound like that when reading small files.

I actually mailed Seagate already, their reply was run the long test - if it passes then it's ok... By the way, I already started the long test yesterday, but it seems to take over 24 hours for this drive, so it's still not finished. They did not offer anything regarding AAM.

Thanks again, and waiting forwards to hear your recording on spinup, when you have time for it.

All the best!

If they didnt mention AAM adjustments... ask them. Keep asking until you get someone who will help you. Explain the noise problem and they will step you through the process. 🙂

This sounds like the next upgrade to my parity HDD in my home server.

IF you can afford the asking price. Yeah they are the best of the best of the best.
IF not, the Wolf Pro and Wolf std are pretty darn sweet and overkill for 99 percent of home users.

YMMV