Compute
Shared vs Dedicated CPU
Shared CPU plans borrow processing power from a common pool and can slow under contention. Dedicated CPU guarantees full-time access to your cores.
On a shared CPU plan, your vCPUs draw from a pool that other tenants also use. Most of the time there's plenty to go around, and shared plans are cheap for exactly this reason. But under sustained load, you can hit the ceiling and get throttled while a neighbor's workload spikes. Performance becomes a little unpredictable.
Dedicated CPU guarantees full, uninterrupted access to your cores. DigitalOcean, Vultr, and Linode all sell both tiers, with dedicated costing roughly two to three times more for the same core count. The premium buys consistency, which matters for anything latency-sensitive or CPU-bound around the clock.
The rule of thumb: shared for bursty, low-average workloads like a typical website, dedicated for databases, CI runners, video encoding, or anything that pegs the CPU for long stretches. Paying for dedicated on an idle web server is money lit on fire.