High Precision Event Timer (HPET) vs TSC (Time Stamp Counter) Synchronization: Advanced Implementation and Real-World Architecture

[Mia Campbell]

Has anyone actually benchmarked QueryPerformanceCounter overhead on modern Zen 4 or Intel 14th-gen rigs when running a custom high-res timer loop? I’ve been chasing micro-stutters in competitive shooters and I’m increasingly convinced the default Windows clock source switching is the culprit—my theory is the OS silently flips between invariant TSC and HPET mid-session, and I can’t isolate it reliably in testing.

🎮 Battlestation & Rig: Ryzen 7 7800X3D | RTX 4080 Super TUF | 32GB DDR5 6000 CL30 | 2TB 990 Pro | ASUS ROG Swift OLED PG27AQDM 27″ 240Hz 0.03ms | Razer Huntsman V3 Pro TKL + Razer Viper V3 Pro + Pulsar Superglide Glasspad | Beyerdynamic DT 770 Pro (80 Ohm) + Focusrite Scarlett Solo Gen 4

📍 Toronto, ON (Canada)

"Smooth frame time pacing matters infinitely more than peak FPS."

[Zoe Edwards]

Great topic. The short version is that TSC almost always wins for per-frame timing in a game loop, but only if you handle it correctly on modern silicon. The critical detail most people gloss over is the invariant TSC — Intel introduced it with Sandy Bridge and AMD with Excavator/Ryzen, meaning the counter now ticks at a constant rate regardless of P-states, Turbo Boost, or C-states. If you’re on anything older, TSC frequency scales with the core and you’re basically asking for jitter. Once you confirm invariant TSC (check the CPUID leaf 0x80000007 EDX bit 8, or just look up your CPU’s spec

"⚠️ When asking for FPS/stutter troubleshooting, please include your exact RAM specs and PSU model!"

[Trevor King]

Great topic, and I want to push past the usual “HPET is slow, TSC is fast” talking points because the reality on modern hardware is more nuanced. The key thing everyone undersells is Invariant TSC.

"[Motherboard: B650] • [CPU: i5-13600K] • [RAM: 64GB DDR5 6000MHz]"

[Author]

Posts