Comments on: Solidigm D5-P5336 122.88TB NVMe SSD Launched Shipping in Q1 2025 https://www.servethehome.com/solidigm-d5-p5336-122-88tb-nvme-ssd-launched-shipping-in-q1-2025/ Server and Workstation Reviews Sat, 16 Nov 2024 11:27:47 +0000 hourly 1 https://wordpress.org/?v=6.7.1 By: Malvineous https://www.servethehome.com/solidigm-d5-p5336-122-88tb-nvme-ssd-launched-shipping-in-q1-2025/#comment-591004 Sat, 16 Nov 2024 11:27:47 +0000 https://www.servethehome.com/?p=82141#comment-591004 Just bought three 15TB Micron drives myself, to replace my Samsung QVO 8TB SATA SSDs. The second hand 15TB SSDs off eBay from China were the same price per TB as the Samsungs were brand new, before the recent price hikes. So now per TB they’re not too bad, but of course buying one is a sting to the wallet enough, let alone three. (I have four NVMe slots in an old Dell R720 to use them in, but I figure by the time I fill up these three drives the prices will have come down enough that I can save money on the fourth 15TB drive, or perhaps spring for a 30TB which are still quite a lot more expensive per TB on the second hand market at the moment.)

If you do this yourself make sure you upgrade the firmware first, apparently early Micron firmwares had a tendency to brick themselves for no reason. You could recover the drive by flashing the latest firmware but the data was lost. My drives came with an undocumented firmware version and Micron are pretty cagey about sharing details on it, so it’s not quite plain sailing dealing with second hand enterprise drives.

]]>
By: osmarks https://www.servethehome.com/solidigm-d5-p5336-122-88tb-nvme-ssd-launched-shipping-in-q1-2025/#comment-590895 Fri, 15 Nov 2024 12:16:30 +0000 https://www.servethehome.com/?p=82141#comment-590895 You can buy 15TB datacentre SSDs used now. Cost/TB is about the same as consumer drives, though.

]]>
By: MDF https://www.servethehome.com/solidigm-d5-p5336-122-88tb-nvme-ssd-launched-shipping-in-q1-2025/#comment-590709 Wed, 13 Nov 2024 21:18:22 +0000 https://www.servethehome.com/?p=82141#comment-590709 @Thomas Womack

They will want client content to train with tenant specific content …so they’re going to take your SPO/OneDrive data and feed it to the training models. And traditional storage models while fast enough for end user access, isn’t going to be fast or large enough for the training models to feed off.

Or

Content creation models…wanting to add more rich data (video/audio) to the mix. Feed it the Matrix so that kids can re-hash it back out into tiktok vids of Morpheus doing a HawkTuah line during Agent Smith’s interrogation.

]]>
By: James https://www.servethehome.com/solidigm-d5-p5336-122-88tb-nvme-ssd-launched-shipping-in-q1-2025/#comment-590684 Wed, 13 Nov 2024 16:19:42 +0000 https://www.servethehome.com/?p=82141#comment-590684 What Cynyr said. A couple of 60TB or 122TB SSDs in a NAS enclosure at a reasonable price would be ideal.

That being said: How many years until thousands of these hit the secondary used market? What is the average use case for an AI datacenter drive? Do they use them for 3 years and replace them with something newer/shinier/higher capacity? Or do they ride them into the ground because all the funding is going to GPUs?

]]>
By: Cynyr https://www.servethehome.com/solidigm-d5-p5336-122-88tb-nvme-ssd-launched-shipping-in-q1-2025/#comment-590678 Wed, 13 Nov 2024 15:21:24 +0000 https://www.servethehome.com/?p=82141#comment-590678 Can we please just get lower cost larger drives? I’d like so see ssds compete with the 10tb – 20tb hdds. As a home user sata3 speeds are plenty for my NAS.

Sidenote: why cant i use an email inbox (me+box@example.com) as an email?

]]>
By: Thomas Womack https://www.servethehome.com/solidigm-d5-p5336-122-88tb-nvme-ssd-launched-shipping-in-q1-2025/#comment-590668 Wed, 13 Nov 2024 14:06:18 +0000 https://www.servethehome.com/?p=82141#comment-590668 Llama 3 was trained on fifteen trillion tokens, which is about 30TB; what on Earth are the AI people storing on drives twice that size? Prodigious numbers of checkpoints so they can rethink training decisions mid-stream?

]]>