Intel PL1 and PL2 Power Limits Explained: Sustained vs Boost Power Behavior

Every modern Intel desktop CPU runs under two separate power ceilings that most people never touch. PL1 sets what the chip can pull indefinitely; PL2 sets what it can pull in short bursts. Motherboard vendors quietly override both, which is why a "125W" CPU can sit at 250W under sustained load.

Intel's power limit system has three numbers that matter: PL1, PL2, and Tau. Intel publishes a base spec for each CPU SKU, but almost no consumer motherboard ships with that spec enabled by default. Understanding what each value does explains why identical CPUs behave so differently across motherboards, and why "unlocking" a chip is often just a BIOS checkbox rather than an overclock.

PL1: the number that caps sustained power

PL1 is the long-duration power limit, expressed in watts, that the CPU is allowed to draw once it has been running under load for longer than the Tau window. On a stock Core i7-13700K, Intel's base spec lists PL1 at 125W. Once the chip has been under sustained multi-core load for more than the Tau period, the power controller throttles clock speed until package power settles at or below 125W. This is the number that determines your all-core Cinebench or video-encode clock speed on a spec-compliant board.

PL1 matters most for workloads that hold all cores busy for minutes at a time: rendering, encoding, compiling, and sustained all-core benchmarks. Gaming rarely triggers PL1 throttling on modern CPUs because most games don't load every core to 100% continuously, so PL2 and the boost window dominate the gaming experience instead.

PL2: the short-term boost ceiling

PL2 is a higher power ceiling the CPU is allowed to draw for a limited time, defined by Tau, before it must drop back to PL1. For that same Core i7-13700K, Intel's spec lists PL2 around 253W. The idea is that most real workloads are bursty rather than sustained, so letting the chip pull significantly more power for a short window gives better responsiveness without raising the sustained thermal load the cooler has to handle all day.

Tau (the boost duration) is typically specified around 28 seconds at stock, though this varies by SKU and board. During Tau, the CPU can run at its full turbo clock as long as thermal headroom allows. Once Tau expires, the power controller steps power down toward PL1 and clock speed drops with it, which is the "boost decay" you'll see in a long benchmark run's frequency graph.

Why your motherboard already ignores the spec

Nearly every enthusiast motherboard from ASUS, MSI, Gigabyte, and ASRock ships with "MCE" (Multi-Core Enhancement) or an equivalent setting enabled by default, which raises PL1 to match PL2 and extends or removes Tau entirely. In practice this means the CPU runs at its boost clock indefinitely rather than stepping down after 28 seconds, provided cooling and VRM temperatures allow it. This is why review sites frequently note that "stock" performance numbers from board partners already exceed Intel's own published spec — the board is silently out of spec by default, not through user tuning.

You can check whether your board is running Intel spec or an extended profile using HWiNFO64, which reports both the current package power draw and the configured PL1/PL2 values under the CPU sensor group. If package power sits well above the CPU's rated PL1 during a long sustained load, MCE or an equivalent feature is active.

Setting limits manually

In BIOS, look for "Long Duration Power Limit" (PL1), "Short Duration Power Limit" (PL2), and "Package Power Time Window" (Tau) — naming varies by vendor but the three values are always present under the CPU or "Extreme Tuning" section. Setting PL1 and PL2 to the same value and extending Tau effectively removes the sustained throttle point, letting the chip run at boost clocks continuously as long as thermals permit. Setting PL1 back to Intel's spec value restores the intended sustained-load behavior and noticeably lowers sustained power draw and heat output for workloads like video encoding, at the cost of lower sustained clocks.

Interaction with cooling and VRM design

Raising PL1 to match PL2 only helps if the cooler and motherboard VRM can actually sustain the higher power draw without thermal throttling kicking in through a different mechanism — core temperature. A 125W-spec air cooler running a chip held at 250W indefinitely will hit TjMax and throttle anyway, just through the thermal limiter rather than the power limiter. On boards with modest VRM cooling, sustained higher power draw also raises VRM temperature, which can trigger its own throttling independent of CPU package power. Before extending PL1, confirm your cooler is rated for the higher sustained wattage and that core configuration and scheduling behavior on your specific chip match what you're tuning for, since hybrid architectures distribute that power budget across P-cores and E-cores differently depending on workload.

When to leave it alone

If your system is already running at target temperatures with acceptable noise, there's little reason to touch these values. The main cases where manual tuning helps are: a system with strong cooling that's being held back by an overly conservative Tau, a laptop or SFF build where you want to manually cap PL1 below default to control fan noise, or diagnostic situations where you're trying to isolate whether a performance or stability issue traces back to power delivery rather than clock or voltage settings.