What is the expected temperature of Intel Optane P1600x at idle and underload?

I recently built a TrueNAS box in a not well-ventilated chassis and noticed the Optane P1600x is operating at an “unusually high” temperature, but still under 70-degree maximum operating temperature. I could not find discussions or reviews regarding this matter. That said, the temperature stays consistent even at idle or under load (if they are even used at all).

The following are their temperatures over a 1-day period (max/mean/min) in celsius.
SLOG (stripe x1): 60/60/58 (under heatsink, near CPU - might be bad installation?)
Meta VDEV (mirrored): 44/43/41 (near case fan, no heatsink), and 53/50/48 (on PCIE to NVME adapter, no heatsink, no airflow)

Meanwhile, here are the temperatures of my 4 SATA SSD over a 1-day period (max/mean/high) in celsius. I did try copying (for testing purposes) over 100 GB at once during this period, in other words, they were under load at some point.
SATA 1: 30/26/24
SATA 2: 33/27/26
SATA 3: 34/29/27
SATA 4: 35/29/27

I know Optane tends to be hotter than regular SSD, as shown in the picture below (HWINFO54: current/min/max/average, in celsius). In this case, the Optane is installed on a PCIE to NVME adapter on my main rig, with direct airflow to the stick (bottom PCIE slot, with bottom to top case fan installed). Note, I used the Optane solely for Linux duo-boot setup, so while in Windows, the Optane is not used at all.

1678565135534.png

Compared to a regular QLC SSD (used for OS installation, at another location, with worse airflow due to under the GPU).

1678565237052.png

With that said, is this a normal behavior of the Optane? When talking about such high temperatures as 60 degrees, is it something to be concerned about? This question is more about understanding the nature of Optane because I do have a workaround to reduce these temperatures. The NAS box is my playground for testing purposes only, so no worry about production downtime or anything along the line.

Running a couple back to back btrfs scrubs, I see around 50 on a U.2 905p with direct airflow from a fan. Maybe the slog is just a heavier workload?

Hey @devemia. Did you ever get this resolved / any update on temp experiences? I have 4xP1600x on a LQD3000 riser card that I put into a RAID 0 array via BIOS (TRX 40). I know, inadvisable and unnecessary etc. etc. I wanted a fast boot drive and wanted to see if I can do better than a single 900p with this crazy approach… I have quite a few of the LQD 3000s I bought of someone second hand a while back.

Anyway, performance is/was mixed vs. a single 900p. Overall about equal it seemed. Better on some metrics and worse on others (looking at CrystalDiskMark RWP R&W+Mix results).

BUT… then I started seeing some TERRIBLE benchmarks and I just figured out why. CDInfo tells me drives all 4 drives are @ ~70C. The card is right next to my 3090 in the case, so I’m sure it’s getting a decent amount of heat from that as well but… these drives seem to be running REAL hot, even when I’m not hammering them.

Going to try some alternate placement configs and some fans blowing directly onto the LQD card to see if I can get the temps under control. Any tips or advice appreciated.