Hi there!
My Zotac gtx 770 has a memory clock of 7 ghz and it is a PCI-e 3.0 card.
But I plan to run it on a PCI-e 2.0 slot.
The pci-e 3.0 clearly states that it has a 15.5 GB/s per lane capacity on x16.
Question - Is that memory clock of 7 ghz anywhere related to this 15.5 GB/s per lane capacity strentgh ? If I run this card on a pci-e 2.0 having 8 GB/s per lane capacity on x16 can it ever bottle neck?
I know this query has been asked in a nutshell but I wanted to know if there is a relation between the memory clock with per lane capacity is gb/s
Thanks.
You're fine with a single PCIe 3.0 GPU on a PCIe 2.0 motherboard.
PCIe 2.0 only gets a little saturated if you crossfire or SLI two high-end cards. So if you're running in x8 x8, there's about 1% drop in performance. I think you'll find that some AMD motherboards (990) run at full bandwidth x16 x16.
Thanks for the reply. So this means a 7 ghz memory clock running a single card can NEVER bottleneck on a pci-e right? I meant to ask that the gb/s of a lane is no where related to the memory clock of the card?
Can I get an answer?
According to this chart - http://www.hardwaresecrets.com/article/190
The pci-e 2.0 clock frequency is stuck to 5 ghz. However my card can run 7 ghz. 2 ghz is a huge drop IMO :(
So pci-e 2 has a mem clock limit to 5 ghz.my 770 can run upto 7 ghz.so i was curious if I am getting a bottleneck.so what I did was ran GPU-z and checked my mem clock.
( link - http://www.hardwaresecrets.com/article/Everything-You-Need-to-Know-About-the-PCI-Express/190/1 )
it was not succeeding 1.75 ghz,and i was like wtf? why cant i see 7000 mhz
Then I found this :-
"The really radical move on the GTX 770 comes with the memory clock. A figure here of 1.75GHz amounts to an effective 7GHz (or 7010MHz, to be precise) on GDDR5 (the RAM effectively quadruples the memory clock figure),
Read more: http://www.pcadvisor.co.uk/reviews/pc-components/3462990/nvidia-geforce-gtx-770-review/#ixzz2d51YDXmj "
So apparently nothing is getting bottlenecked and I am left satisifed after this research.lol. way to go for messing our heads Tech Brands.