Samsung 960 EVO Nvme SSD dead

My Samsung 960 EVO NVme 250GB SSD has died.
I got about a 1 minute warning. Windows became sluggish, so I decided to reboot.
The PC wouldn’t post, and stuck in the UEFI screen. I suspected an SSD had failed, but I was surprised it was the Samsung.
No data lost as it was the system drive, and backed up only yesterday with ‘Reflect’.

I’ve ordered an Samsung 970 EVO Plus to replace it. A 500GB version this time.

Scary, good that you had a backup. Which version of Reflect do you use? Will the free edition be sufficient for home usage?

1 Like

I use version 7 (free edition).
The free edition is more than enough for a home user. Just make sure you build a rescue media, preferably on a USB flash drive.
Reflect will backup and restore MBR or GPT partitions. It will even backup and restore Linux as well as Windows systems.

I thought Samsung SSDs have a health detector that give you a warning when the drive is going bad. Its been so long since any drive has died on me I am sort of out of the loop. I have been using an sata 3 Samsung 840 Pro for many years and its still going strong maybe you just got a bad one.

1 Like

The ware indicator only indicates the state of ‘flash memory’.
The last time I looked at it was probably a couple of months ago, and it was at 96% life left.

SSD’s have more than just flash memory though. You have an SSD controller, host interface, support components, and some NVRam. Anyone of those failing will also kill the SSD.

The same applies to an HDD, they too can fail suddenly, and with little or no warning.

1 Like

The new SSD has arrived and been fitted.

A quick benchmark from the Samsung Magician utility.


Click on the picture for a full size image

Drive info


Click on the picture for a full size image

A screenshot from the new Samsung Magician utility (version 6)


Click on the picture for a full size image

I’m happy with the performance of the new drive.

4 Likes

Glad you’re happy with the new one Dee.

How long had you been using the old one for?

The drive was almost exactly 3 years old.
It was a review sample with a locked firmware, so the firmware couldn’t be updated, nor the drive secure erased.

Anyway, there are no such restrictions with the new drive. :slight_smile:

That’s a bit scary actually as three years isn’t really all that long.

Is that normal or have you just been unlucky?

I think I was just unlucky.
Having said that. People are forever moaning about the price of SSD’s and PC components in general.

To make things cheaper, you have to use cheaper components. It’s as simple as that.

Keep in mind the 960 Evo that I had was a pre-production unit, with pre-production locked firmware.

1 Like

Yup, I was thinking about that fact Dee. The only thing I’d do is perhaps monitor the 5v and 3.3v rails on your supply just to be sure there aren’t any transient over or WORSE, under-voltages happening at all. Sometimes even the most knowledgeable folks forget there’s not much standing between there and some really violent and BIG “angry pixies.”

I have not bought a hard disk in many years and am terrified at the prospect.
When I did, I always bought 7200 RPM Server Grade drives. I figured if I was just using them as a consumer, I would get a long life from them.
If there is such a thing as server grade SSD’s, maybe it would be worth the extra money.
Just a thought.

Blackblaze publishes an quarterly report on its hard drive failure rates for its cloud backup data centre, such as this. They use consumer grade hard disks in their arrays, which really puts them to the test long term. Consumer grade hard disk reliability has improved quite a lot over the years as I still remember them listing several models with high failure rates. I happened to have the least reliable model, the Seagate 3TB ST3000DM001, which reportedly reached a 47.2% failure rate at Blackblaze before they took the remainder out of service. I retired mine around the time just in case even though it never had an issue. Now the highest failure rate they list is the Seagate 4TB ST4000DM000 with a 2.67% annualised failure rate. Most of the rest have 0.5-1% annualised failure rates.

As for SSDs, my early poor experience of them almost 10 years ago never put me off using one as the boot drive ever since my first SSD. The time I spent replacing the SSD and restoring the OS partition was a very small price to pay compared to the time it saved with having applications stat up much faster, no sluggishness with heavy multitasking, etc. The main thing I do is not store anything important just on the SSD, which simply involves redirecting my user profile folders to the hard disk. This means if I save anything to Documents, Pictures or even the desktop, this is all stored on the hard disk, so if my SSD fails, it’s just a matter of replacing it, restoring the OS and continue where I left off. I haven’t had a boot SSD fail in my main desktop or laptop in around 6 years.

1 Like

My setup is pretty near to what Sean does.
All the drives inside the PC (9 of them) are SSD’s.
The NVme drive is the system drive. Half of it is used as a scratch disk for audio and video editing, and the completed files are then stored on one of 3 SSD’s specifically used for audio and video editing.

Another one of the SSD’s (528GB) is used only for Macrium Reflect system backups. It’s great to be able to restore the system drive quickly.

All data files are then backed up to an external 8TB USB3 HDD which is encrypted using BitLocker.

Most of the SSD’s require specific software for maintaining them, such as secure erase. That required 4 or 5 USB flash drives to boot from, because in most cases you cant SE an SSD that is a system drive. Plus another USB flash drive for Macrium Reflect restores.

I got fed up with that, and made a USB3 (live) full Windows 10, booting and running from USB3. Now I have all my tools usable from one session. :slight_smile:

Having experienced defective hard disks, SSDs, optical discs and various types of removable media over the years, over the past few days I can add another one to my list - A defective cloud storage.

My workplace uses SharePoint and OneDrive through Microsoft Office 365’s, i.e. Microsoft hosts the storage. Over the past week, whichever server Microsoft hosts our SharePoint on is having serious issues as new files are failing to sync and it takes multiple attempts to open existing files. Microsoft is still trying to troubleshoot the issue. In the meantime, many files are unable to open with the following error:

SharePoint%20Cloud%20Operation%20Error

Now you’ve seen what a Microsoft cloud error looks like. So if you store data on the cloud, don’t forget to back it up to some other medium. :cloud_with_lightning: :wink: