Comment by rayiner
2 days ago
> Compiler people think aliasing matters. It very seldom does. And VLIW will never become practical for entirely unrelated reasons (read: OoO is fundamentally superior to VLIW in general purpose computing).
I love how long Linus has been right about this.
Linus worked for Transmeta in 2003. That company built a VLIW CPU.
I don’t think you were a visionary if you stated VLIW is not practical for general purpose computing after that time.
Also “VLIW will never become practical” is a stronger claim than “VLIW will never become practical for general purpose computing”, and may be up for debate. https://en.wikipedia.org/wiki/Very_long_instruction_word says
“Since the number of transistors on a chip has grown, the perceived disadvantages of the VLIW have diminished in importance. VLIW architectures are growing in popularity, especially in the embedded system market, where it is possible to customize a processor for an application in a system-on-a-chip.”
That may be untrue or no longer true, but clicking around looking a the Wikipedia pages for several of the mentioned VLIW designs, I couldn’t find conclusive evidence for that (but, as with that VLIW page, the pages I looked at may be out of date, weren’t clear as to whether chips still were in production, whether newer generations actually still are VLIW, etc.)
He's right about OoO but not about aliasing.