It seems that a lot of GPU problems revolve around specific versions of drivers. Though NVidia has their own support structure, you can often learn from information reported by others who fold.
To make a long story short, Folding doesn't work with dissimilar cards. They need to be identical in chip type (not necessarily board manufacturer) and therefore have the same number of shaders. And even in that case, when you install multiple cards, be sure to do a clean uninstall of the old drivers first. For that, Display Driver Uninstaller (DDU) is currently the favorite, but there are others.
Hardware configuration: Intel i7-4770K @ 4.5 GHz, 16 GB DDR3-2133 Corsair Vengence (black/red), EVGA GTX 760 @ 1200 MHz, on an Asus Maximus VI Hero MB (black/red), in a blacked out Antec P280 Tower, with a Xigmatek Night Hawk (black) HSF, Seasonic 760w Platinum (black case, sleeves, wires), 4 SilenX 120mm Case fans with silicon fan gaskets and silicon mounts (all black), a 512GB Samsung SSD (black), and a 2TB Black Western Digital HD (silver/black).
In the example above, the point change has nothing to do with the addition of a second GPU. The reason the 970 PPD is down is because it pulled a core 15 work unit that has no bonus.
Fah does not require like chips to run well on multiple GPUs. However, running like families of chips (like all Keplers) helps to avoid potential driver conflicts.
Let the core 15 WU finish and see if the PPD goes back to normal on the next core 17 WU.
JimF somewhat overstates the issue, if he had said there can be problems with using two different cards that would be closer to the actuality. First, if you just added the 460 to an already existing F@H installation, it is possible that the client has misidentified which card is associated with each GPU folding slot. It is recommended that a full reinstallation of the folding client be done when there are hardware changes to GPU slots. In the meantime you can confirm whether or not the correct GPU is listed for each slot by pausing one and seeing which GPU is actually being used with a GPU monitoring app.
Second, as shown in your first picture, one of the WU's being processed is using Core_15. Wu's from those projects do not get the Quick Return Bonus, so the PPD will be lower on many GPU's. Also given the estimated time to complete, I suspect the WU is actually being processed on the 460, not the 970.
There are some how-to's posted on adjusting the GPU index values to get the right one associated with each slot, but I would recommend a full reinstall of F@H first after finishing your current WU's. Then if the index values are still wrong, follow one of the guides to get them set correctly.
iMac 2.8 i7 12 GB smp8, Mac Pro 2.8 quad 12 GB smp6
MacBook Pro 2.9 i7 8 GB smp3
I have reinstalled the client to no use.
Both cards running at near 100% and it is the 970 that has got the slow PPD.
Wanted to play some Battlefield yesterday and I paused the 970 client to do so.
It looks like the client has still got the slots mixed (I've never had it work on windows in 20+ installations), but at least the assignment for the 460 is working as intended (core-15) - it's just a pity that it then gets run on the 970
Personally, I'd not try and fold on an old fermi and save the electricity - it will draw pretty much the same power as the 970 and at best do 1/20th of the work. Even just having it in the system will likely slow the 970 down by more than it could produce itself.
Hardware configuration: Intel i7-4770K @ 4.5 GHz, 16 GB DDR3-2133 Corsair Vengence (black/red), EVGA GTX 760 @ 1200 MHz, on an Asus Maximus VI Hero MB (black/red), in a blacked out Antec P280 Tower, with a Xigmatek Night Hawk (black) HSF, Seasonic 760w Platinum (black case, sleeves, wires), 4 SilenX 120mm Case fans with silicon fan gaskets and silicon mounts (all black), a 512GB Samsung SSD (black), and a 2TB Black Western Digital HD (silver/black).
