I worked on a VLIW processor long ago and it had a theoretical peak of 700 MIPS (iirc) back in 2000. It was a neat architecture but required fairly low level knowledge to get the most out of it.
Sounds like the Intel i860 from the late 1990s. Frighteningly fast VLIW in theory, but in practice not so much. I think untweaked code ran at 3% of max speed.
The problem there is, Intel intentionally kills any non-x86 Intel arch. For example, the Itanium was brilliant, and then they repeatedly threw it under the bus until people hated it.
That's not really how the story goes, though. Intel spent a lot of effort on getting Itanium to succeed, and they really dragged their feet into x86-64. It was the market that decided that the price/performance of Itanium wasn't worth it.
Thats how a lot of people tell the story, but I just don't agree with it. There is zero evidence that the market killed Itanium when Intel was already trying to kill it because it was eating into Xeon sales on high end platforms.
