Transmeta: A CPU Revolution That Never Was [video]
youtube.com
youtube.com
https://en.wikipedia.org/wiki/OQO
good times
It was a super cool device but ultimately too slow to really do much. You could play videos on it if you compressed them down to ~800x480, but I had a Palm TX and an iPod video that could do the same and the battery lasted significantly longer on those devices.
Maybe if I had an IT career then I could have found a use for it. The transflective screen was pretty neat on it tho. In bright sunlight you could still see the screen perfectly.
But: (1) Getting that to work with the transistor budget available to a 250nm chip is hard, and the value is comparatively low: there just weren't that many execution units to dispatch to! In some sense they were ahead of their time; Sandy Bridge ended up making it work almost a decade later.
And more importantly: (2) Intel at the time was at the peak of its process ascendency. Transmeta might have made a competetive CPU at the circuit level, but not when their fabs were a full process node behind. NVIDIA and ATI, at the time, were having a ton of success in the industry because they weren't trying to compete with Intel.
I had a Compaq TC1000 for some time and I while I loved it for it's exotic nature and interesting approach, it was slow as molasses.
[0] https://datasheets.chipdb.org/Transmeta/pdfs/brochures/cruso...