I have been doing HFT for many years now, so I have an acute appreciation for exactly what is likely to be involved in doing what they claim to be able to do. I

I have been doing HFT for many years now, so I have an acute appreciation for exactly what is likely to be involved in doing what they claim to be able to do. If you want to put an algorithm in control of large amounts of real money, then some very thorough testing is required to ensure that it's not going to blow up due to some corner case that no one thought of (of which there are typically a surprising number). I suspect there are some significant caveats to the claim, and even considering that I still think that's a far more impressive accomplishment than just making java fast enough.