DEC's Alpha created HP panic, desperation

When HP first delivered the 32 bit HP-PA architecture in the early 1990s, they were the performance leader for a very brief period. But DEC's Alpha quickly followed and blew them out of the water in integer performance and floating point performance and it being 64 bits.

So HP management got desperate since their technical people didn't know how to design a 64 bit chip and since HP-UX was widely acknowledged as the worst unix ever sold (dubbed hp-sux on the company's own mailing lists).

So they made a deal with Intel circa 1994 that they would jointly work on a 64-bit chip. HP were the idiots who came up with EPIC (which is nothing but a stupid interpretation of the superscalar idea) that ruined Itanium 1 and they distracted Intel from doing a 64-bit version of IA32.

HP were the morons responsible for the original Itanium 1 which everyone agrees was a piece of crap.

They had TEN YEARS to out-design the Alpha and they couldn't do it. The Itanium 1 with EPIC at the time of its introduction was slower even than a three-year old Alpha chip fabbed with an obsolete process after DEC/Compaq had dropped their committment to the Alpha architecture. Even the Pentium 3 was a superior architecture, with far less cache! Look at spec.org if you don't believe it: can you even find another chip that performs worse than the Itanium 1 on an equal clock speed or equal fab level basis?

Now that the Itanium 2 doesn't suck (because they stole ideas from Alpha and gave up on their moronic EPIC idea), they are anxious to claim some credit to restore their ruined technical reputations. But where were the HP designers when their Itanium 1 was proven to be a piece of upbelievable crap? Or when most computer architects ridiculed their EPIC idea in technical discussions and in computer architecture courses ("the itanic")? Or when their HP-PA couldn't keep up with the Alpha or the RS/6000 or even the Athlon for cripe's sake? Or when it took them eight+ years to design a 64 bit architecture? I don't think I've every heard as many chip designers make fun of an architectural idea as they did for HP's EPIC idea, which is nothing but a bad interpretation of IBM's "superscalar" concept from 1981!!!

And lest you continue in your misunderstanding, the Alpha is a superior architecture to the Itanium. Compare their performance on an equal clock speed basis or on an equal fab size basis and you will see that the Alpha would be viable to 2025. If technical merit determined market dominance, we would all be running on Alphas. And if DEC/Compaq/HP had spent any money to get Alphas on a competitive fab process, the HP designers would be even more discredited than they currently are (if that is even possible).

Name, email address supplied µ