"The first two versions of PCIe used 8b/10b encoding, and thus had lower processing overhead."
"Version 4.0 will use the same 128b/130b encoding..."
I was under the impression that going to 128b/130b was a rather significant reduction in overhead compared to 8b/10b or is this referring to the scrambling/descrambling that is applied to the data? After all the 2 bits in an 8b/10b encoding is 20% while 128b/130b is 1.5%.
Blog Doing Math in FPGAs Tom Burke 24 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...