- May 25, 2009
- 2
- 0
- 0
I have a motherboard with PCI Express 1.0 x16 (1.1a apparently). I want to upgrade my video card and am wondering at which point cards used by themselves (no SLI or Crossfire) without overclock actually took advantage of PCI Express 2.0??
I read that the Geforce 9800GTX 512MB PCI-E 2.0 does not take advantage of PCI Express 2.0 when run alone so I could just as well get this card and it will run as per its normal specs even with my PCI Express 1.0 x16 slot (apparently 1.1a) ...
I don't understand this because on Wikipedia it says that with 2.0 you get a standard bus bandwidth of 0.5GByte/s and with an x32 connector you get up for 16 GB/s for both videocards. So I will assume that with only one, you can actually get up to 32 GB/s.
But then according to Techarp, the Geforce 9800GTX 512MB PCI-E 2.0 has a memory bandwidth of 70.40 GB/s...
But then why would they make a card that fast if the PCI Express 2.0 technology only goes up to 32 GB/s... so I guess it can handle more...
The bottom line is if I am shopping for and then using a single video card, and not overclocking, and have sufficient wattage in my power supply, at which point would a "PCI Express 2.0" being run on my motherboard (Capable of PCI Express 1.0 x16 ... 1.1a apparently) actually start to perform below the stock specs as seen on a site like Techarp (eg- According to Techarp, the GeForce 9800GTX 512MB PCI-E 2.0 has a memory bandwidth of 70.40 GB/s...)
Maybe some clarification on how much memory bandwith (in GB/s) the PCI Express 1.0 x16 (1.1 apparently) can handle assuming a single card?? That should help me when I scan the lists on Techarp... Or perhaps there are other factors which come into play.
Thanks for reading this and thanks for any help!!
I read that the Geforce 9800GTX 512MB PCI-E 2.0 does not take advantage of PCI Express 2.0 when run alone so I could just as well get this card and it will run as per its normal specs even with my PCI Express 1.0 x16 slot (apparently 1.1a) ...
I don't understand this because on Wikipedia it says that with 2.0 you get a standard bus bandwidth of 0.5GByte/s and with an x32 connector you get up for 16 GB/s for both videocards. So I will assume that with only one, you can actually get up to 32 GB/s.
But then according to Techarp, the Geforce 9800GTX 512MB PCI-E 2.0 has a memory bandwidth of 70.40 GB/s...
But then why would they make a card that fast if the PCI Express 2.0 technology only goes up to 32 GB/s... so I guess it can handle more...
The bottom line is if I am shopping for and then using a single video card, and not overclocking, and have sufficient wattage in my power supply, at which point would a "PCI Express 2.0" being run on my motherboard (Capable of PCI Express 1.0 x16 ... 1.1a apparently) actually start to perform below the stock specs as seen on a site like Techarp (eg- According to Techarp, the GeForce 9800GTX 512MB PCI-E 2.0 has a memory bandwidth of 70.40 GB/s...)
Maybe some clarification on how much memory bandwith (in GB/s) the PCI Express 1.0 x16 (1.1 apparently) can handle assuming a single card?? That should help me when I scan the lists on Techarp... Or perhaps there are other factors which come into play.
Thanks for reading this and thanks for any help!!
