• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

Seagate HDD with NVMe Interface Demoed

TheLostSwede

News Editor
Joined
Nov 11, 2004
Messages
18,887 (2.50/day)
Location
Sweden
System Name Overlord Mk MLI
Processor AMD Ryzen 7 7800X3D
Motherboard Gigabyte X670E Aorus Master
Cooling Noctua NH-D15 SE with offsets
Memory 32GB Team T-Create Expert DDR5 6000 MHz @ CL30-34-34-68
Video Card(s) Gainward GeForce RTX 4080 Phantom GS
Storage 1TB Solidigm P44 Pro, 2 TB Corsair MP600 Pro, 2TB Kingston KC3000
Display(s) Acer XV272K LVbmiipruzx 4K@160Hz
Case Fractal Design Torrent Compact
Audio Device(s) Corsair Virtuoso SE
Power Supply be quiet! Pure Power 12 M 850 W
Mouse Logitech G502 Lightspeed
Keyboard Corsair K70 Max
Software Windows 10 Pro
Benchmark Scores https://valid.x86.fr/yfsd9w
Back in June, NVM Express, Inc announced the NVMe 2.0 specification, which included support for mechanical drives, also known as hard disk drives. Seagate has now demoed the first implementation of NVMe 2.0 on a HDD and is targeting market availability sometime in 2024. You might wonder why anyone would want an NVMe equipped hard drive and the short answer is that it's intended for the server space.

One of the key reasons that Seagate is even considering this, is because of its multi-actuator drives, which the company is expecting to surpass the interface speed of 12 Gbps SAS in the future. In addition to this, it makes sense having a single storage interface on servers, which allows for a mixed storage drive environment, where some drive bays can be fitted with SSDs and others with NVMe based hard drives. Seagate is planning to sample key customers with early drives in September 2022 and final drives will be backwards compatible with SATA and SAS.



View at TechPowerUp Main Site
 
Well given the habit of board manufacturers to keep stripping SATA ports, this isn't that unexpected to me.
 
and all we need is a 100.000 rpm drive to saturate that bandwith lol
 
and all we need is a 100.000 rpm drive to saturate that bandwith lol
I guess you missed the part about Seagate's multi-actuator drives? It's like having two or more drives inside the same housing.
 
Cdr. Data about his quest to become moar human:

"An evolutionary step... in the wrong direction" (- from the ST-TNG episode Descent part 2"
 
I wonder, will we see direct storage support on faster HDD in the future? With HAMR/MAMR and TDMR, we might see substantial boost to HDD speed hitting consumer space.
 
Does this mean the DRAM cache on HDD's won't be bogged down by SATA bandwidth limitations!? You really only need like 1GB DRAM cache for every 1TB of storage.
 
I guess you missed the part about Seagate's multi-actuator drives? It's like having two or more drives inside the same housing.
OK, so they can finally get to almost full sata III speeds then?
 
But why? Shouldn't they just try to make a faster SATA interface for SSDs, I mean, SATA 3.0 has been around for over a decade.
 
yep, been there, done that, and IIRC, it's called nvme :fear::respect::roll:
You joke, but it was called sata express, not NVMe. Same speed as SAS 12Gbps.

Come to think of it, servers already have SAS 12Gb. No hard drive is saturating that bus anytime soon. Why bother with NVMe for such a slow medium. What's next, NVMe tape drives?
 
yep, been there, done that, and IIRC, it's called nvme :fear::respect::roll:
When I should laugh? On a typical motherboard you have 1 or 2 M.2 slots and you have 4-6 SATA connectors, it's cheaper to just get a new SATA SSD than replace those NVMe drives for more storage.
 
But why? Shouldn't they just try to make a faster SATA interface for SSDs, I mean, SATA 3.0 has been around for over a decade.
Remember SATA Express? Biggest failure since DFP when it comes to PC interfaces of any kind.
 
Remember SATA Express? Biggest failure since DFP when it comes to PC interfaces of any kind.
Barely, I remember those connectors on motherboards but I haven't never seen an actual device.
 
But why? Shouldn't they just try to make a faster SATA interface for SSDs, I mean, SATA 3.0 has been around for over a decade.
SATA is a dead end. They tried with SATA Express and failed miserably. Best to leave it as as is and adopt NVME.
Come to think of it, servers already have SAS 12Gb. No hard drive is saturating that bus anytime soon. Why bother with NVMe for such a slow medium. What's next, NVMe tape drives?
Im thinking signaling, latency and dropping support for SATA/SAS. NVME is not just about transfer speeds.
When I should laugh? On a typical motherboard you have 1 or 2 M.2 slots and you have 4-6 SATA connectors, it's cheaper to just get a new SATA SSD than replace those NVMe drives for more storage.
Why is it cheaper to get SATA? Atleast Gen3 NVME drives are dirt cheap and now with toolless M.2 slots coming to market swapping out M.2 is easier than SATA with it's power and data cables. Cleaner too as we more towards to having less individual cables dangling around.

Personally i've always had trouble with SATA as the ports are always blocked by a long GPU. So i have to remove the GPU to access the ports. Although depending on the cooling solution the same might be true for M.2
Atleast there's hope for getting rid of the stupid screws there.
 
Wouldn't it make more sense to call it the first U.2 3.5" HDD? This is not a M.2 drive, it still will need the same ports as a SATA drive would. The real reason for it would be to simplify motherboard manufacturing - no SATA controllers, just split the available PCIe lanes up a little more. Also makes expansion cards easier... One advantage I see is that you could use a x1 lane and still have similar performance to a SATA device with a cheaper motherboard. It also has compatibility (2 lanes) with SATA, and therefore has legacy support. I say it is the future.
 
Wouldn't it make more sense to call it the first U.2 3.5" HDD? This is not a M.2 drive, it still will need the same ports as a SATA drive would. The real reason for it would be to simplify motherboard manufacturing - no SATA controllers, just split the available PCIe lanes up a little more. Also makes expansion cards easier... One advantage I see is that you could use a x1 lane and still have similar performance to a SATA device with a cheaper motherboard. It also has compatibility (2 lanes) with SATA, and therefore has legacy support. I say it is the future.
Form factor vs data transfer protocol. NVME does not mean it has to be M.2 or even U.2 or U.3
 
I guess you missed the part about Seagate's multi-actuator drives? It's like having two or more drives inside the same housing.
Double the heads, double the clicks, double the fun?
 
Form factor vs data transfer protocol. NVME does not mean it has to be M.2 or even U.2 or U.3
True enough. However, based on my understanding, the form factor would be 3.5", the connector would be U.2 (from the picture and the description), and the supported protocols are NVMe, SATA, and SAS. I guess I just look at things a little differently.

Im thinking signaling, latency and dropping support for SATA/SAS. NVME is not just about transfer speeds.
I agree fully. Even a x1 link shouldn't be saturated by the fastest dual-actuator drives.
 
This is probably more to do with standardizing communication protocols and controllers, and possibly unifying storage under the NVMe protocol. Having a direct line to the CPU and faster SSD drives could help with the "hybrid memory management" idea that things like AMD's StoreMI 2.0 tries to do, without needing a SATA or SAS compatibility layer.
 
Back
Top