4 ms·

FX!32 was good compared to what was available elsewhere (including for Itanium IIRC). I am not saying it had to compete with fully with x86 on consumer desktops

by p_l 2mo ago

FX!32 was good compared to what was available elsewhere (including for Itanium IIRC). I am not saying it had to compete with fully with x86 on consumer desktops from beginning, but it would have been more stable position, especially with known upcoming work on 64bit-clean NT5. EV7 I am interested in more details because I want to know if the slippage was part of the original development run, or the post-Itanium one when it was restarted by HP. EV8 was admittedly dead in the water but I would say not so much for technical reasons as much as for direction options available (vs. following the samsung partnership to bring more AlphaPC-like systems) The motherboard difference was slight between K7 and EV6 parts from my understanding if there was a will to commit to large scale production (arguably EV67 could be probably made into pin-compatible packaging even). But admittedly small part.

stinkbeetle 2mo ago
> FX!32 was good compared to what was available elsewhere (including for Itanium IIRC). I am not saying it had to compete with fully with x86 on consumer desktops from beginning, but it would have been more stable position, especially with known upcoming work on 64bit-clean NT5. I meant not good compared with what was available elsewhere, i.e., actual x86 CPUs. It didn't emulate privileged modes, so none of your operating systems worked, it required Windows NT, and it only provided emulation for Win32 x86 binaries, and it did not emulate all facilities either. Independent benchmarks are not easy to come by, but even the few results published by DEC were not fantastic. A 500MHz Alpha (likely EV6) would achieve maybe little better performance than a 200MHz Pentium (P5, not PPro)! EV6 did 2.66x specint95/MHz that Pentium MMX 200 achived, and 2.5x the frequency. Natively it would havehave been 5.32x faster on those benchmarks, but it was a little over 1.x. This is not the 40-60% performance claimed of the trivial byte benchmark result, this is 20% performance at best. https://www.researchgate.net/publication/3214997_FX32_a_profile-directed_binary_translator https://www.researchgate.net/publication/3214997_FX32_a_prof... The slip was the original development AFAIK. HP later restarted a failed process shrink I think, but I can't remember if they finally got that done in the end.

by p_l 2mo ago

stinkbeetle 2mo ago

> FX!32 was good compared to what was available elsewhere (including for Itanium IIRC). I am not saying it had to compete with fully with x86 on consumer desktops from beginning, but it would have been more stable position, especially with known upcoming work on 64bit-clean NT5. I meant not good compared with what was available elsewhere, i.e., actual x86 CPUs. It didn't emulate privileged modes, so none of your operating systems worked, it required Windows NT, and it only provided emulation for Win32 x86 binaries, and it did not emulate all facilities either. Independent benchmarks are not easy to come by, but even the few results published by DEC were not fantastic. A 500MHz Alpha (likely EV6) would achieve maybe little better performance than a 200MHz Pentium (P5, not PPro)! EV6 did 2.66x specint95/MHz that Pentium MMX 200 achived, and 2.5x the frequency. Natively it would havehave been 5.32x faster on those benchmarks, but it was a little over 1.x. This is not the 40-60% performance claimed of the trivial byte benchmark result, this is 20% performance at best. https://www.researchgate.net/publication/3214997_FX32_a_profile-directed_binary_translator https://www.researchgate.net/publication/3214997_FX32_a_prof... The slip was the original development AFAIK. HP later restarted a failed process shrink I think, but I can't remember if they finally got that done in the end.