Some folks like having shiny hardware and it’s partly a hobby rather than a need in almost certainty barring specific niche professional needs. I upgrade my wor

Some folks like having shiny hardware and it’s partly a hobby rather than a need in almost certainty barring specific niche professional needs. I upgrade my workstation setup quite rarely because my work provided laptop is what I use 90%+ of the time. Top of the line is one thing (Xeons and Threadrippers or EPYC) but pretty powerful cost effective CPUs like a 3900X or 9900k are another. Even though I do ML stuff occasionally I still do it as a useless hobby and it’s not worth bumping up to a Threadripper or a crazy GPU, especially when I can use a cloud machine for a fraction of the price and as a bonus force myself to make it somewhat repeatable and deployable for someone else. If one’s been building machines in the past year or so, we’ve just had the biggest performance gains in CPUs in literally 10 years with the latest AMD CPUs. I went a long time with an E3-1230 but got really bored honestly after 5 years and built an i7-4790k ITX machine which wasn’t as big of a change as going from 8GB of RAM to 16 GB or to my first SSD. I got a 3900X this past year and the substantial performance gains are not a big enough reason for me to upgrade yet again compared to something like the new Ampere GPUs for even hobby machine learning purposes like mine. In the future a bigger reason for me to build again would be based around space efficiency reasons where a small case could fit on top of my desk and drive 4K AAA games, encode HEVC video day and night, and all with a 600w SFX PSU.