@Riedl I was a bit confused about the difference between "overall latency" and "end-to-end" latency. Since they didn't give the benchmark numbers for the "end-to-end" latency, it's hard to tell how big an improvement this really is. It looks like the upper bound, though, is about 2x.