Linux, Nvidia GPU & PCIe 4x?

[cine.chris]

Can Linux run the FAHclient with an Nvidia GPU on a PCIe 4x slot? This was disabled on windows ~1yr ago, I`m wondering if Linux is currently an option?

[bruce]

Both Linux and Windows are able to run normal GPUs on x4 slots provided that's what your motherboard supports.

[cine.chris]

The Supermicro X9SCM I tried would work on both x8 slots but not the x4s. I`ll test a different mobo.

[cine.chris]

Both Linux and Windows are able to run normal GPUs on x4 slots provided that's what your motherboard supports.

Thx Bruce, as you said `what your motherboard supports`.
This system I`m surfing on, with an Asrock Z87 Extreme4 worked perfectly on Win10 w/ 3 GPUs, x8x4x4.
Perhaps the Win10 comments I gathered on an OC site referred to a mobo?
I have 3ea of the X9SCM mobos, so it would be rewarding to re-purpose them as a 4-slot folder in a bit-mining setup.

[markdotgooley]

Is this about PCIe 4.0 or about x4 bandwidth on PCIe 3.0 or 2.0?

[HaloJones]

Can't comment on Gen 2 but I have a 1070 running on a PCIE Gen 3 4x slot with no impact compared to other cards in other systems that have faster slots.

[cine.chris]

Is this about PCIe 4.0 or about x4 bandwidth on PCIe 3.0 or 2.0?

It`s about me trying to get FAH to run on some older Supermicro mobos. For this case, the problem slots were two 8x slots running at PCIe 2.0 x4.
I`ve scrapped the idea for now, too many problems.

[markdotgooley]

Is this about PCIe 4.0 or about x4 bandwidth on PCIe 3.0 or 2.0?

It`s about me trying to get FAH to run on some older Supermicro mobos. For this case, the problem slots were two 8x slots running at PCIe 2.0 x4.
I`ve scrapped the idea for now, too many problems.

Ah. Someone suggested a Supermicro on eBay complete with two 2GHz Xeons and memory, under US$200, because it has four PCI 3.0 x16 slots, provided both Xeons remain installed. Probably three GPU cards would fit with no worries about bandwidth. I think that’s over 90W per Xeon, not to mention heat from the GPUs...

[ajm]

That was me and it's still on offer: https://www.ebay.com/itm/Supermicro-Ser ... cf15f01599

The board: https://www.supermicro.com/products/mot ... -LN4F_.cfm

[aetch]

From another thread

I've already replied multiple times to this question.
It's up to you to find and accept those settings, or just go against it.
In most cases:
PCIE 1.0 x4 / 2.0 x2 / 3.0 x1 is not recommended for folding. You could fold on it, but will experience serious PPD penalties on budget to mid range GPUs.
PCIE 1.1 x16 / 2.0 x8 / 3.0 x4, good enough for GTX 1600 series GPUs under Linux (an RTX 2060 gets 975k PPD in Linux, which is slower than a Core 21 WU), or up to a GTX 1060 in Windows.
PCIE 2.0 x16 / 3.0 x8 is good enough for up to a 2080Ti under both Linux and Windows. Estimated upcoming 5000+ shader/core gpus might be limited in Windows, but should fold fine in Linux.
PCIE 3.0 x16, there's currently no GPU that would exceed this PCIE bandwidth, but if there were, it would be a GPU with 8000 cores or more.

viewtopic.php?f=38&t=35651#p338263

[markdotgooley]

That was me and it's still on offer: https://www.ebay.com/itm/Supermicro-Ser ... cf15f01599

The board: https://www.supermicro.com/products/mot ... -LN4F_.cfm

It's tempting to buy one and just see if I can get my two 2060s to run unrestricted on it, then try to add a third GPU card. A bit power-hungry and it probably wouldn't be worth running CPU WUs on it...

[ajm]

These are 6C/12T CPUs and you would have two of them: https://ark.intel.com/content/www/us/en ... l-qpi.html
Those 20 threads would certainly fold quite well!

[markdotgooley]

These are 6C/12T CPUs and you would have two of them: https://ark.intel.com/content/www/us/en ... l-qpi.html
Those 20 threads would certainly fold quite well!

Depends on my power bill. Also the board comes with heat sinks on the CPUs that don't look as if they have fans. Maybe perching a fan on each would suffice?

The board is cheap enough and I have enough old parts (adequate power supply, notably, everything but a case) that I could buy it and see how it behaves.

[ajm]

Do let us know!

[markdotgooley]

Do let us know!

If I go through with it, I will. Probably I’ll post blow-by-blow accounts.