PCI-Express 1.0 x16 vs. PCI-Express 2.0 x16 - is there a tangible difference?Asked by amber lassiter
We all know that the difference in the speed of transmission over the pci-express bus has doubled, but a number of issues are related:
1) Are all the bus capabilities used by modern video cards or will there be no visible difference if you insert, for example, GeForce GTX460 in PCIe 1.0 and 2.0?
2) How to check the exact throughput? I found the PCIe Speed Test v0.2, but it only works on ATi video cards
3) Are there any motherboards with honest pci-express 2.1 or is there no difference between 2.0 and 2.1?
All this interests me both in the context of games and in the context of CUDA
There is a difference. But it is very minimal and appears only in high resolutions and sli / crossfire tandems.
2. In what sense is accurate? 1 line - 500 megabytes per second. Do you suspect that you have "Chinese" lines with a low PSP?
How much you can actually pump data in terms of CUDA is another matter entirely in practice.
3. There is no difference in speed, there is a difference in features (more interesting to the developers of the boards, not to their users). There are no implementations.
Regarding the "exact" question arose due to the fact that I have a Gigabyte EP35C-DS3R motherboard based on the P35 chipset. So the manufacturer claims that it works on pci-express 2.0, the documentation for the chipset says that the chipset supports only version 1.1, and Everest shows that pci-express is generally the first version. So, in connection with divergent data, the question arose where the wrong was written, and why Gigabyte did not receive a stream of criticism due to the deception in the description. I thought that maybe they somehow perished and did something mean, but how to check my assumption, I do not know - yitades
I got it in due time, but it was a little bit thin, this stream =) they were mistaken on the box, there is not 2.0 there, of course. - michelle baretela