What TDP actually is
MakeUseOf's writer says he picked a PSU based on a misreading of TDP, and it later proved too small. His conclusion is that TDP is a cooling figure and has "nothing to do with power draw". The first half is mostly right. The second half goes too far.
Intel's own support documentation describes TDP as a wattage figure. It says TDP refers to power consumption under the maximum theoretical load. Intel also says the purpose of the number is to give system designers a power target for choosing a thermal solution. So TDP is a thermal-design rating expressed in watts. It is not unrelated to power, and it is not a ceiling on what the chip can draw.
The same Intel page is explicit about the ceiling. Under a steady workload at published frequency, power is roughly the TDP. During turbo, or with certain workloads such as AVX, the chip can exceed it. That only lasts for a limited time, or until the processor hits a thermal throttle temperature or a power-delivery limit.
That gives you a working definition:
- TDP is a design target that tells cooler and board designers what to plan for.
- Actual draw depends on the workload, the chip's configured limits and the motherboard or system it sits in.
- A higher TDP class generally means a hungrier design, but TDP isn't a promise about draw in any specific workload.
Intel: TDP became PBP and MTP
Intel retired the TDP label on newer parts. Its documentation says the TDP terminology was replaced with Processor Base Power from 12th Gen Core onward. To check a chip on Intel's ARK site, read the TDP field for processors older than 12th Gen. For 12th Gen and newer, read the Processor Base Power field.
The Core i9-14900K shows why this matters. Intel's specification page lists it with:
- 125 W Processor Base Power
- 253 W Maximum Turbo Power
- 24 cores (8 Performance and 16 Efficient) and 32 threads
- a 6 GHz maximum turbo frequency
- a 100°C junction temperature limit
Treat the 125 W figure as the base design number. Intel's own page puts the maximum turbo number at roughly double that. The MakeUseOf piece describes PL1 as simply "base clocks" and PL2 as "boost clocks". That oversimplifies. PL1 and PL2 are configurable power limits, and boards often ship with different settings. Some set the two equal. Intel's PBP and MTP are the published reference values, and your motherboard's firmware decides what actually happens.
AMD: TDP, PPT and the 1.35 rule
AMD also publishes TDP, but its useful number for CPU power is PPT, or Package Power Tracking. PPT is the socket power limit. The Socket AM5 figures from Tom's Hardware's Ryzen 7000 review show how the two relate:
| TDP class | PPT (socket power) |
|---|---|
| 65 W | 88 W |
| 105 W | 142 W |
| 170 W | 230 W |
The TDP × 1.35 formula in the MakeUseOf piece has independent support. Tom's Hardware, TweakTown and HWCooling all describe PPT as roughly 1.35 times TDP for Zen-era sockets. HWCooling adds a qualifier: the ratio is "usually (there are exceptions)" 1.35×. Wccftech's list shows the pattern holds across several classes (45 W to 60 W, 65 W to 88 W, 105 W to 142 W). Rounding explains the 170 W to 230 W class, since 170 × 1.35 is 229.5.
My take: the multiplier is a decent shortcut for stock-limit estimates, but check the exact model's published limits. Motherboard firmware can also raise PPT. Tom's Hardware reported an MSI BIOS option that lifted the Ryzen 5 9600X and Ryzen 7 9700X from a 65 W limit to 105 W. With that setting, the stock-TDP arithmetic no longer describes the system.
I couldn't retrieve AMD's own product page for the Ryzen 7 7800X3D. I'm therefore not citing its specifications here.
Why you can't compare an "80 W" Intel chip to an "80 W" AMD chip
The two vendors define and use the number differently. Intel's current practice splits it into base and turbo figures. AMD keeps TDP as a class label and attaches a separate socket limit. I couldn't verify a precise, standardized cross-vendor measurement method, which is itself a reason not to treat the numbers as directly equivalent. Compare chips within a vendor's family, and use independent power testing for cross-brand comparisons.
Practical buying guidance
Choosing a cooler
- Look up the CPU's turbo or socket limit: Intel's Maximum Turbo Power, or AMD's PPT.
- Check the cooler's rated capacity against that figure, not the base TDP.
- Look for independent reviews of the exact CPU and cooler pairing, since no single number guarantees temperatures.
- Remember that firmware can change the limits. Intel's documentation names thermal and power-delivery limits as constraints on turbo.
Sizing a PSU
CPU power is only one line in the budget. Add the graphics card, drives, fans and the rest of the system. Use the CPU's maximum figure (MTP or PPT), not the base number, for the CPU line.
The article recommends 20% to 30% headroom above calculated draw. That is a reasonable rule of thumb, and it's commonly used. It isn't a manufacturer requirement, and I haven't found primary-source evidence that headroom extends PSU life. Treat it as sensible engineering margin for transient spikes and future upgrades.
Laptops and prebuilts
Laptop and OEM designers choose the power and cooling limits of the finished machine. A thin heatsink may force lower sustained limits. The sound takeaway is that a CPU name alone doesn't describe the performance of a laptop or prebuilt. Intel's documentation on thermal and power-delivery limits supports the principle. I haven't verified a specific laptop comparison or measured how much performance varies. Check independent reviews of the exact model and look at its sustained-load results.
The bottom line
"TDP is just cooling" is a good slogan and an imperfect fact. TDP is a thermal design target in watts. It is not the maximum a chip can pull, and it isn't comparable across brands. For real planning, use the figure that matches your question:
- Cooler: Intel's Maximum Turbo Power or AMD's PPT, plus reviews.
- PSU: the whole system's draw, with a margin.
- Laptop or prebuilt: the specific model's tested behavior.
A TDP number alone won't tell you any of those.