Direct answer

A dedicated CPU allocation can reduce noisy-neighbor interference, but it does not guarantee faster ticks. A fast shared vCPU may beat a slower dedicated one when host contention is low. Evaluate CPU generation, per-thread performance, throttling and actual server load.

Look for host contention

  1. Record /spark tps and a profile while the lag occurs; separate server-thread cost from network delay.
  2. On Linux VMs, compare CPU steal in top or the st field of mpstat (if installed). Persistent steal during lag suggests the host is not scheduling your vCPU immediately.
  3. If available, inspect cat /sys/fs/cgroup/cpu.max and repeated readings of cpu.stat for container quotas and throttled periods. These are different from VM steal time.

Choose the right upgrade

If steal or quota throttling is consistently high, ask the provider about guaranteed CPU time and allocation. If steal is low but the main tick thread is saturated, prioritize faster per-core performance or reduce the measured tick workload. Folia and modded servers can distribute work differently; test your actual workload.

Version and safety notes

VPS providers can use different meanings for dedicated and shared vCPU. Verify billing and scheduling guarantees rather than assuming the plan label specifies physical core ownership.

Verification

Run the same player count, world and profiler workload on both options and compare median and high-percentile MSPT, not only provider “core” counts. Keep enough RAM, storage performance and backup capacity consistent to avoid a misleading comparison.

References

Check documentation for the versions and hosting environment you actually run.