[Dylan16807]

I'd like to see the distribution on that 250-5000 nanoseconds claim. The low end of that competes quite well with Optane, but Optane under moderate loads can keep up similar latency for 99.99% of requests.