Oh crap, the listing says "Internal SAS Ports!". That was on my To Do list to get to the bottom of those connectors.
As an alternative, there are LSI OCP 2.0 controllers. I have some 40 GbE QSFP OCP 2.0 cards coming as well. It will allow me to test multiple OCP cards, PCIe limitations, etc. There was one request this thread asking if an dual OCP slot connector card will work in a single OCP slot. Oddball things like that for a FAQs.
There is a 2nd OCP 2.0 port with a single Type A x8 PCIe port. At least, via the pictures I can see it on the left, unoccupied.
I have never seen an OCP 2.0 device or port inperson - until this arrives this weekend. But to me, it may interfere with the left-side pci expansion card (if you install one). That's another thing to test, OCP height limitations with PCIe cards because the OCP specification allows for 3 different heights.
Btw, I'm updating the OP with findings and links to posts as we go (it's a WIP)
Of course, which is why I mentioned it.
So the 6126F processors either run REALLY hot, or I damaged the heatsink.
root@ubuntu-server:~# sensors
coretemp-isa-0000
Adapter: ISA adapter
Package id 0: +93.0°C (high = +91.0°C, crit = +101.0°C)
Core 0: +26.0°C (high = +91.0°C, crit = +101.0°C)
Core 1: +25.0°C (high = +91.0°C, crit = +101.0°C)
Core 3: +25.0°C (high = +91.0°C, crit = +101.0°C)
Core 4: +25.0°C (high = +91.0°C, crit = +101.0°C)
Core 5: +26.0°C (high = +91.0°C, crit = +101.0°C)
Core 6: +26.0°C (high = +91.0°C, crit = +101.0°C)
Core 8: +24.0°C (high = +91.0°C, crit = +101.0°C)
Core 9: +25.0°C (high = +91.0°C, crit = +101.0°C)
Core 10: +24.0°C (high = +91.0°C, crit = +101.0°C)
Core 11: +25.0°C (high = +91.0°C, crit = +101.0°C)
Core 12: +25.0°C (high = +91.0°C, crit = +101.0°C)
Core 13: +24.0°C (high = +91.0°C, crit = +101.0°C)
power_meter-acpi-0
Adapter: ACPI interface
power1: 136.00 W (interval = 1.00 s)
pch_lewisburg-virtual-0
Adapter: Virtual device
temp1: +27.0°C
or for those of us who don't natively speak Metrique:
root@ubuntu-server:~# sensors -f
coretemp-isa-0000
Adapter: ISA adapter
Package id 0: +199.4°F (high = +195.8°F, crit = +213.8°F)
Core 0: +78.8°F (high = +195.8°F, crit = +213.8°F)
Core 1: +77.0°F (high = +195.8°F, crit = +213.8°F)
Core 3: +77.0°F (high = +195.8°F, crit = +213.8°F)
Core 4: +75.2°F (high = +195.8°F, crit = +213.8°F)
Core 5: +78.8°F (high = +195.8°F, crit = +213.8°F)
Core 6: +77.0°F (high = +195.8°F, crit = +213.8°F)
Core 8: +75.2°F (high = +195.8°F, crit = +213.8°F)
Core 9: +77.0°F (high = +195.8°F, crit = +213.8°F)
Core 10: +75.2°F (high = +195.8°F, crit = +213.8°F)
Core 11: +77.0°F (high = +195.8°F, crit = +213.8°F)
Core 12: +77.0°F (high = +195.8°F, crit = +213.8°F)
Core 13: +75.2°F (high = +195.8°F, crit = +213.8°F)
power_meter-acpi-0
Adapter: ACPI interface
power1: 136.00 W (interval = 1.00 s)
pch_lewisburg-virtual-0
Adapter: Virtual device
temp1: +78.8°F
I'm pretty sure that's mostly accurate due to the system fans running at full blast. As in, loud enough that I can't hear the AC unit over them, nor can I hold a conversation in the same room without yelling. LOUD.
for reference, this is my second box- the first is running DUAL 5120s, and runs quiet enough that I can easily lose track of the fact that it's even running. I'm going to try replacing the heatsink and see if that makes a difference. Also, as far as accuracy goes... this is a 12 core proc, but it looks like lmsensors picked up 14... <shrug?>