M10 Intel Optane Memory: Theoretical estimate SSD performance upgrade?

Hello PCSpecialist Team,

Last month I've had brief contact with Martijn den Burger from the support team, who has been kind enough to give me a proper explanation on my previous questions on the i9 9900KS. Though it is too late for the KS now, I'm looking forward to the i9 10900k with the latest Z490 chipset motherboards now. But all that aside, the explanation on Optane did not fully cover my question. So I'll ask it again with an example build down below.

In this build I am hypothetically trying to accelerate the SATA SSD #4 with M10 Intel Optane 64GB.

This is a concept build of a new setup on an Asus ROG Maximus XII Extreme motherboard with 4× M.2 slots:
M.2 SSD #1: 2TB Samsung 970 EVO Plus R:3500MB/s, W:3300MB/s (OS Drive)
M.2 SSD #2: 2TB Samsung 970 EVO Plus R:3500MB/s, W:3300MB/s (Game Drive)
M.2 SSD #3: 2TB Samsung 970 EVO Plus R:3500MB/s, W:3300MB/s (Game Drive)
M.2 Optane: M10 Intel Optane 64GB (Accelerates SATA SSD #4)
SATA SSD #4: 4TB Samsung 860 QVO 6 GB/s R:550MB/s, W:520MB/s (Game Drive)
SATA SSD #5: 4TB Samsung 860 QVO 6 GB/s R:550MB/s, W:520MB/s (General Media & Work Drive)

So my first question here is: Is this SSD setup technically possible?
If yes: In theory, is the performance upgrade from the M10 Optane even noticeable on an SSD while loading heavy games?

I'd very much like to get a firm grasp on the cost-benefit ratio, with extreme performance in mind. If the increase in performance is too little, or barely noticeable, then I'd know to just replace it with another M.2 SSD if this motherboard allows it.

Much appreciation in advance!
 

ubuysa

The BSOD Doctor
No, I think the Optane will be a waste of money with that SSD. Optane is a disk cache, it's designed to reduce the seek and latency delays associated with HDDs.

To be effective you need the next block of data you read to be in Optane memory, so the effectiveness of Optane depends on your data access patterns and the Optane management algorithm. It's quite possible that your data access patterns for that drive don't suit Optane at all.

In any case, you'd be doing extremely well if Optane were able to accelerate an HDD to even close to the speeds of that SATA SSD, except under ideal conditions.

The limiting factor in that SSD is the SATA interface, not necessarily the SSD chipset itself. Optane won't help very much with that. I doubt you'll see much measurable difference except, perhaps with near perfect data access patterns such that the SSD was not accessed directly at all.
 
Last edited:
Top