Itanium wasn't a great design in many ways, but it did have really neat little bits like VLIW which was supposed to make JIT work. Intel never delivered on JIT for Itanium but Transmeta had a really performant way of running IA-32 code on their 128bit VLIW processor during the same time.
We seem to be on the threshold of JITed regular programs being the norm. MacOS runs AMD64 code on their M1/M2 Arm CPUs faster than Intel/AMD CPUs(check out Apple's "Rosetta 2" that lets x86-64 apps run on M1/M2). Windows arm64 runs IA-32 and AMD64 apps pretty flawlessly through Microsoft's emulation layer.
Transmeta's patents all went to a troll when they went belly up, but those patents are now expired.
If Intel did a IA-128 that was a beefier Itanium design with a handful of flaws addressed I could see it doing well today. Microsoft is in a better position to do JIT(because it already works on Arm). A beefy VLIW is perfect for AI/ML. Intel's AI/ML chips are still VLIW and do very well.
https://www.tomshardware.com/n...