The processor core supply — VCore — is the most demanding rail on a laptop board. It runs at around one volt or less, delivers tens of amps under load, and changes its voltage hundreds of times a second at the processor's request. It is made by the biggest regulator on the board, and it is involved in a large share of "turns on, no display" and "shorted board" repairs. This guide explains how it is built, how it is controlled and how to diagnose it.
What VCore is
VCore is the main supply for the processor cores. Depending on the generation, the board may also make separate rails for the integrated graphics (often called VCCGT or similar) and for other parts of the processor (system agent, I/O, auxiliary rails), or it may provide one input rail that the processor divides internally. The schematic's power map shows which rails your processor needs; names vary between Intel and AMD and between generations. The general term is described under vcore.
The building blocks
| Block | Job | What it looks like on the board |
|---|---|---|
| Multi-phase controller | Decides the voltage, balances phases, reports status | A square chip with many pins near the processor |
| Power stages or driver + MOSFETs | Switch the system rail into the inductors | Small packages next to each inductor, one per phase |
| Inductors | Store and smooth energy, one per phase | Large square blocks in a row |
| Output capacitors | Hold the voltage steady during load changes | Banks of polymer and ceramic capacitors around the processor |
| Control bus | Processor tells the controller what voltage it needs | Clock, data and alert lines between processor and controller |
Each phase is a buck converter; several phases run in turn so the current is shared and ripple is lower. The structure is covered in detail in VCore Phases, Inductors and Output Capacitors, and the term under multiphase vrm.
How the voltage is set
Unlike most rails, VCore is not set by a fixed resistor divider. The processor sends the controller a voltage request over a serial bus — Intel calls it SVID, AMD uses its own SVI interface. At start-up the controller usually ramps to a boot voltage defined by its configuration; afterwards the processor requests different voltages as its load and frequency change, and can ask for very low voltages or turn phases off when idle.
This has two consequences for repair:
- The VCore voltage you measure varies. There is no single "correct" number; a meter shows a moving average.
- A missing VCore can be a communication problem, not a power problem — if the controller never receives a valid request, or never gets its configuration right, it may not regulate.
Enable and power-good
The VCore controller does not start on its own. Typical sequence:
- Its supply rails are present (the system rail and a small logic supply).
- An enable signal — often named
VR_ONor similar — arrives from the EC or chipset late in the power sequence. - The controller ramps VCore to its boot voltage.
- It raises a power-good signal (
VR_READY,VR_PWRGDor similar) that the chipset waits for before releasing the processor.
If the enable never arrives, the problem is earlier in the sequence. If the enable arrives and VCore does not appear, the controller, its configuration, its power stages or a short on the output are suspects. Power-good behaviour is covered in Power Good Signals and Why They Matter.
Diagnosing VCore
VCore missing
With the board powered and the power button pressed:
- Check the controller's supply pins and its enable input.
- Check that the bus lines to the processor have their pull-up voltage.
- Measure each phase's switch node: no switching means the controller is not driving the stages.
- With power removed, measure VCore to ground.
VCore shorted
VCore normally reads a very low resistance to ground — sometimes only a fraction of an ohm, because the processor itself is a large load. That makes it hard to judge from a single reading. Compare with:
- the same rail on a known-good board, or values other technicians have recorded for the same board;
- diode-mode readings, which are often more telling than resistance at such low values.
A true short reads close to zero in both resistance and diode mode. Common causes are a failed power stage (high-side to switch node or switch node to ground), a cracked ceramic capacitor, and occasionally the processor itself. Isolating them uses the usual techniques: lift a phase's inductor to separate its power stage from the output, then inject a low voltage at limited current and look for the hot spot.
VCore present but no boot
VCore can be present while the processor still does not start — the power-good may be missing, another processor rail may be absent, or the clock and reset may be at fault. Treat VCore as one item in the checklist rather than the whole answer.
Inductors and the switch node
Each phase inductor gives you two convenient measurement points: the switch-node side and the output side. With power removed, a switch node shorted to ground points at the phase's low-side switch; a switch node shorted to the system rail points at the high-side switch. The output side is VCore itself, shared by all phases. Duty cycle and ripple of each phase can be estimated with Buck Converter Calculator (Duty Cycle and Inductor Ripple).
Probing VCore safely
The VCore area packs high-current switch nodes, gate drive pins and the system rail into a small space, all right next to low-voltage control lines. On a powered board, measure VCore on the large output capacitors or an inductor's output pad, never on controller pins. A probe slip from a switch node or the system rail onto the processor's supply or a control line can destroy the processor instantly. Use a current-limited supply where possible, so that a slip during testing limits the damage.
Example: no display after a liquid spill
A laptop powers on, fans spin, no display. Measured in sequence, every rail up to the processor's auxiliary rails is present, the controller's enable is active, but VCore stays at zero and its power-good never rises. The controller's bus lines show the clock line at its pull-up voltage but the data line at zero. With power removed, the data line reads near zero ohms to ground; a tiny corroded capacitor-sized ESD part on the line, close to the spill marks, is shorted. Once it is removed and the area cleaned, the controller receives its commands, VCore ramps up and the laptop boots.
FAQ
What voltage should VCore be?
There is no fixed value. It changes with load and processor state; at start-up it ramps to a boot voltage, then follows the processor's requests.
Why does VCore read almost zero ohms to ground?
The processor is a large, low-resistance load. Compare with a known-good board rather than judging the number alone.
Is a shorted VCore always a dead processor?
No. Power stages and ceramic capacitors fail more often. Isolate them before condemning the processor.
Can I measure VCore with a multimeter?
Yes, as an average. A meter cannot show fast changes or ripple; an oscilloscope can.
What is SVID?
Intel's serial bus that lets the processor tell the VCore controller what voltage it needs. AMD uses a similar interface with its own name.
Bottom line
VCore comes from a multi-phase controller, power stages, inductors and capacitors, enabled late in the sequence and steered by the processor over a serial bus. Check the controller's supplies, enable and bus first, then the phases and switch nodes, and judge VCore resistance only against a reference. Many VCore faults are power stages, capacitors or a single damaged bus line — not the processor.