Five key themes that dominate the discussion
| # | Theme | Representative quotes |
|---|---|---|
| 1 | RAM & memory price inflation | “I still regret not buying 1 TB of RAM back in ~October…” – epistasis “If I had a decent sales channel I might be speculating on DDR4/DDR5 RAM and holding it because I expect prices to climb even higher in the coming months.” – epistasis |
| 2 | On‑prem vs cloud – staffing & operational cost | “The main cost with on‑prem is not the price of the gear but the price of acquiring talent to manage the gear.” – justsomehnguy “If you can’t bill a customer for it, and it’s not scaling regularly, then it shouldn’t be in the public cloud.” – zbentley |
| 3 | CPU core density vs single‑core performance | “E cores didn’t just ruin P cores, it ruined AVX‑512 altogether.” – mort96 “Without the hyperthreading (E‑cores) you get more consistent performance between running tasks.” – hedora |
| 4 | Target workloads for many‑core machines | “A large number of fast, low‑power cores would indeed suit such a application, where large numbers of network nodes are coordinated in near real time.” – topspin “For a Yocto build, the only metric that really matters is total integer compute per dollar and per watt.” – mort96 |
| 5 | Power, noise, and data‑center constraints | “I personally feel like I will downscale my homelab hardware to reduce its power draw.” – MayeulC “The noise these things make… You better have a separate garage.” – speed_spread |
These five themes capture the bulk of the debate: the rising cost of memory, the trade‑offs between owning hardware and using cloud services, the architectural shift toward many‑core CPUs, the kinds of workloads that benefit from such density, and the practical limits imposed by power, noise, and data‑center compatibility.