A few years ago, we demonstrated some truly awesome capability in single racks and on single machines. We had one of our units (now at a customer site), specifically the unit that set all those STAC M3 records, showing this:
and a rack of our units (now providing high performance cloud service at a customer site)
for 8k random reads across 0.25 PB of storage on a very fast 40GbE backbone.
These were tremendous numbers 2 years ago, and they are still fantastic today.
Why am I saying then, that imitation and repetition is a form of flattery?
Well, we’ve had calls and discussions with people for the past few months that think they have the best things since sliced bread, and are only now approaching our two year old numbers.
It would be, how shall I put this … unwise … to assume we have been standing still this whole time. Nope, definitely not wise. More on this later, but I find it very affirming that most of market are now saying the same things we’ve been saying for a decade.
But far more important than that is, you can’t fake performance. Most benchmarks don’t reflect well on actual use cases, the ones that do (like STAC M3) paint a very … very different picture than the “best case, fastest possible” scenarios with people building RAID0’s on the flash (yeah, really, we’ve seen that, and their numbers were still not even in the same ball park as ours).
Its nice to see, yet again (as we did with accelerators and APUs many years ago), our viewpoints, our message, resonating and amplifying across the market.