4 ms·
Your professor was 100% correct: chips haven't been getting faster, though, and for precisely the reasons they stated. However, parallelizing work onto separat
by pwinnski 4y ago
Your professor was 100% correct: chips haven't been getting faster, though, and for precisely the reasons they stated.
However, parallelizing work onto separate cores turned out to continue to extend performance, so it didn't matter so much.
That said, I feel like there were many things that were faster 25 years ago than now, so even multicore hyper-threaded CPU improvements are being chewed up by software doing more. Still, that's a separate issue.
- hotpotamus 4y agoI had to dig out an old Pentium 4 Dell machine this weekend to work with some old hardware. 1 DIMM and a SCSI card had died in the 12 years it was powered off, but it worked just fine with 256MB of memory. Mostly as responsive as a modern machine, but you could tell when the 20 year old spinning drive had to chug a bit to load things. And I seem to still be able to navigate WinXP with no issues - no weather or ad widgets loading under my mouse cursor; it was actually a breath of fresh air. I'm not sure if that's nostalgia/familiarity or if it was really more useable. I've also heard non tech people complain about how clear it is that each Windows version is just a shabby coat of paint over the last version and they wish Microsoft would just stop, so I don't think it's just me. I have to imagine XP would fly on a current machine with multiple gigs of RAM and an SSD.
- vl 4y agoNot at all. In 1998 there was 0.25 um, ie 250 nm process, now we have 5 nm process. 50x improvement. In 1998 processors were ~100MHz, now we have 2-4GHz. 20x-40x improvement. Modern chips are incredibly faster even on the single core. You couldn’t get anything close to modern smartphone on 1998 technology.