At the heart of every laptop (or desktop) computer is a central processing unit (CPU), commonly called a processor or just a chip, that’s responsible for nearly everything that goes on inside. The CPUs you’ll find in current laptops are made by AMD, Intel, Apple, and Qualcomm. The options may seem endless and their names byzantine. But choosing one is easier than you think, once you know a few CPU ground rules.
This guide will help you decrypt the technical jargon that haunts every laptop specification sheet—from core count to gigahertz and from TDP to cache amounts—to help you pick the one that suits you best. With almost no exceptions, a laptop processor can’t be changed or upgraded later as some desktops’ can, so it’s essential to make the right choice from the start.
First Up: Some Basic CPU Concepts
The CPU is responsible for the primary logic operations in the computer. It has a hand in everything: mouse clicks, the smoothness of streaming video, responding to your commands in games, encoding your family’s home video, and more. It’s the most important piece of hardware.
Before we get into specific CPU recommendations, let’s build an understanding of what differentiates one from another by focusing on the central traits that all laptop processors have in common.
Processor Architecture: The Silicon Underpinnings
Every processor is based on an underlying design called an instruction-set architecture. This blueprint determines how the processor understands computer code. Since software operating systems and applications are written to work most efficiently—or sometimes only—on a certain architecture, this is probably the most important decision point for your next processor.
Broadly speaking, today’s laptop processors use either the ARM or x86 architecture. The latter was created by Intel in 1978 and dominates the PC industry, with Intel and AMD battling for market-share supremacy. ARM-based chips, on the other hand, are produced by hundreds of different companies under license from the British firm ARM Limited, owned by Softbank. (Its planned sale to Nvidia is currently under scrutiny by the FTC and other regulators.)
Found in billions of devices from smartphones to supercomputers, ARM chips had been seen only in some Chromebooks and a very few Windows laptops (based on Qualcomm CPUs) until Apple switched from Intel to its own ARM-design M1 processors in late 2020. Apple’s changeover is a leading reason that ARM chips are seeing wider acceptance as an alternative to x86 for mainstream computing.
Your architecture choice is preordained if you’re an Apple user, with a few Intel-based Macs still available but M1 chips dominating (with rumors of a beefier M1X and all-new M2 on the horizon). But Microsoft Windows, Chrome OS, and many Linux operating systems are compatible with both ARM and x86. Based on our reviews of today’s handful of Qualcomm-powered Windows systems like the Microsoft Surface Pro X tablet and HP Elite Folio convertible, x86 remains our recommended architecture for Windows until more apps are written to run natively on ARM.
Apps written for x86 can operate on ARM chips through software emulation, but the translation layer slows performance compared to code written to run on ARM in the first place. Similarly, the occasional ARM CPUs (notably from MediaTek) seen in budget Chromebooks have proven much less peppy than the Intel and AMD processors in midrange and premium Chromebooks.
Core and Thread Count: Firing on All (CPU) Cylinders
Today’s laptop CPUs are composed, in part, of two or more physical cores. A core is essentially a logic brain. All else being equal, more cores are better than fewer, although there’s a ceiling to how many you can take advantage of in any given situation. A much simplified analogy is to the number of cylinders in a car engine.
For basic tasks like internet surfing, word processing, social media, and video streaming, a dual-core processor is today’s bare minimum. (Indeed, you can’t buy a single-core laptop today.) Multitaskers will be much better off with a quad-core CPU, now found even in many budget notebooks. For gaming, video editing, and other processor-intensive applications, look for a six- or eight-core processor. These are typically found in larger notebooks, since they require extra cooling. (They also tend to be a higher tier of CPU; more about that stratification in a bit, when we talk about Intel and AMD chip specifics.)
Then there’s the issue of thread count. We’re not talking about linens and sheets here, but processing threads. A thread is essentially a task, or a portion of a task, for the computer to perform. Computers routinely juggle hundreds or thousands of them, though a processor can work on only so many threads simultaneously. That number equals its thread count, which is often double its core count.
In olden days, CPU cores could process only one thread at a time, but today’s processors frequently (but not always) have thread-doubling technology that allows one core to work on two threads simultaneously. A quad-core chip with this technology, for example, can handle eight threads at a time. Intel calls this Hyper-Threading; the generic term is simultaneous multithreading.
At the minimum, look for a processor that can process four threads. Users working on heavy media creation and conversion tasks will want the ability to handle eight or more. Core count trumps thread count; all else being equal, a quad-core CPU without multithreading will generally outperform a dual-core processor with it. Of course, in the processing world, all else is seldom equal; that’s why so many varieties of chips exist. The next item, clock speed, is another key differentiator.
Clock Speed: The CPU Stopwatch
Measured in megahertz (MHz) or more often gigahertz (GHz), a processor’s clock speed is its operating frequency—a driver of how many instructions (basic operations) the processor can crunch through per second. Higher clock speed is generally better, though things get muddy when comparing clock speeds between different brands or even between chips within the same brand. That’s because some CPUs are more efficient than others, able to process just as many instructions in a given slice of time despite operating at a lower clock speed. Still, clock speed can be telling when comparing chips within a single vendor’s family line.
To complicate things further, today’s processors typically have two advertised clock speeds: a base (minimum) clock and boost (maximum) clock, sometimes dubbed turbo speed since Intel refers to the duality as Turbo Boost technology. When handling light workloads, the CPU runs at its base clock, typically between 1GHz and 2GHz for laptop chips though sometimes higher depending on the processor’s rated wattage. (More on that variable in a minute.) When more speed is needed, the CPU temporarily accelerates—often to 3.5GHz to 5GHz or so—until the task is done. Processors don’t run at their boost clock all the time because they might overheat.
Some low-end laptop processors lack a boost clock altogether, limiting their performance under pressure. Laptop CPUs’ boost clocks are often as high as their desktop counterparts’, but usually not sustained for as long before ramping down due to power or thermal limitations. This concept is called throttling, a safety measure built into the processor to keep it running within its rated specifications.
Thermal Design Power (TDP) Rating: Watt’s Up
Just as important as clock speed in determining a processor’s overall performance is its thermal design power (TDP) rating. This number, measured in watts, is often misinterpreted as a chip’s power consumption. Actually, it tells computer designers how much thermal energy the cooling solution they use must be able to dissipate in order for the processor to operate effectively.
Dissipating heat within the confines of a notebook chassis is a challenge. Putting a desktop processor into a laptop is a recipe for overheating, which is why mobile CPUs are a class of their own, designed with TDP ratings appropriate for a laptop environment. Laptop TDP ratings range from just a few watts for ultra-compact notebooks to a desktop-class 65 watts for some gaming rigs. (See our picks for the top desktop processors.) You won’t shop for a CPU strictly by TDP, but it’s good to know what the number signifies about the class of chip you’re looking at.
Most laptop CPUs are rated between 15 and 28 watts. They have a low enough thermal profile to work in slim notebook designs, yet sufficient power to reach desktop-like boost clocks for at least a short period. Notebooks with these chips almost always require active cooling—that is, the presence of one or two small onboard fans. Laptops with passive cooling—fanless designs, appealing because they’re silent—are restricted to processors rated for just a few watts, fine for everyday tasks but ill-matched for demanding jobs such as video editing.
Both AMD and Intel put the letter “H” at the end of their model numbers for chips at the top of the mobile-CPU TDP roster, rated between 45 and 65 watts and found in gaming laptops, mobile workstations, and other desktop replacements. They’re suited for the most demanding apps and most intense multitasking. (More about model numbers and letters later.)
A very few outliers like the Alienware Area-51m gaming rig and some laptops from specialists like Eurocom actually use socketed desktop CPUs, cooled by multiple fans in a big, thick chassis. The only notebooks that may allow you to swap out their processors at a later date, they’re extremely expensive specialty items that defy portability with huge, heavy AC adapters (or even pairs of them).
Cache: You’ve Probably Got Enough
A processor’s cache is a small memory pool, usually just a few megabytes, that is separate from the system’s main memory (RAM). It helps the CPU manage its workflow by providing a lightning-fast way to retrieve data. More cache—often subdivided into Level 1 through Level 3 (L1 through L3) cache depending on its closeness to the core logic—means quicker performance, but you can safely ignore this spec; gone are the days when processors were sent out into the world with too little cache to perform effectively. We only mention it because you’ll see it listed as you dig into processor specifications.
Integrated Graphics: The GPU on the Chip
Gaming laptops and mobile workstations depend on dedicated or discrete graphics processing units (GPUs) to accelerate 2D or 3D rendering, just as high-end desktops rely on AMD Radeon RX or Nvidia GeForce or Quadro graphics cards inserted into motherboard PCI Express slots. Laptops made for office productivity often don’t need a separate GPU, and can handle drawing the onscreen display with the integrated graphics processor (IGP) built into most of today’s CPUs.
We’ll dive deeper into integrated graphics performance a little later. For now, just know that while the latest processors can handle light or casual gaming—Intel especially has made considerable strides since the molasses-like graphics of its older CPUs—hardcore gamers will unquestionably want a laptop with a discrete GPU under the hood.
Which Way to Go: Intel or AMD?
With the basics covered, let’s start on specific processor brands. This section will focus on the x86 processors available from AMD and Intel, since most of Apple’s MacBooks have transitioned to the company’s own ARM-based M1 chips. (The 16-inch and cheapest 13-inch MacBook Pro have Intel inside at this writing, but we expect the whole lot to be based on Apple silicon before long.)
AMD and Intel are fierce competitors for laptop CPU market share as of 2021. This was not the case during the 2010’s, when Intel dominated the market with better-performing and more power-efficient processors, mostly relegating AMD to entry-level budget notebooks.
The last few generations of AMD Ryzen mobile CPUs have made the former underdog a formidable competitor. (See our benchmark tests of the first Ryzen 5000 series laptop processors.) That said, Intel still enjoys the favor of some manufacturers and corporate IT managers, which can force your processor choice depending on the laptop.
The ABCs of the Archrivals: Pentium, Core, Ryzen and More
AMD and Intel differentiate their laptop processors according to all of the basic concepts discussed earlier, but their top-level branding is most visible to casual shoppers. Here are their basic product lines by intended market.
Intel’s mainstream laptop CPU brand is Core, while AMD’s is Ryzen. They clash at every level—AMD’s Ryzen 3 competes with Intel’s Core i3, Ryzen 5 with Core i5, and Ryzen 7 and Ryzen 9 with Core i7 and Core i9.
Among laptops and Chromebooks retailing for just a few hundred dollars, AMD’s Athlon chips vie with Intel’s Celeron and Pentium lines. AMD has no direct alternative to Intel’s Xeon for flagship mobile workstations, though its Ryzen 7 and Ryzen 9 can offer similar performance. Xeons are essentially Core i7 or Core i9 processors with extra features, designed to work with exotic error-correcting-code (ECC) memory and to guarantee smooth operation with specific professional applications.
Most shoppers will find the middle members of the Core and Ryzen families offer the best mix of performance and value. The Ryzen 5 and Core i5 are particularly well-rounded. Supporting multithreading across the board in their latest generations, they are more powerful than the Ryzen 3 and Core i3 but cost less than the Ryzen 7 and Core i7. The latter will tempt power users and gamers, while users with cash to burn for whom media-rendering or number-crunching wait time means money can spring for a Core i9 or a Ryzen 9.
Generations and Codenames: You’ll Need a Decoder Ring
Just as car companies go by model years, AMD and Intel differentiate their chips by generation, identified at the start of their part numbers. For example, Intel’s Core i7-1065G7 and Core i5-1135G7 respectively belong to its 10th and 11th Generation families of mobile CPUs with integrated graphics. (Yes, there are exceptions, see “Special Intel Naming Conventions” below.) AMD indicates the generation after indicating the family or performance level (3, 5, 7, or 9): the Ryzen 7 5800H is a fifth-generation or Ryzen 5000 series chip.
Tech sites like PCMag also indulge in the codenames AMD and Intel use while chips are in development, such as “Tiger Lake” for Intel’s 11th Gen Core processors and “Cezanne” for AMD’s Ryzen 5000 series mobile chips. These inside-baseball terms are industry lingo more than consumer marketing terms, but they get used aplenty even after a chip is released. Confusingly, Intel sometimes uses multiple codenames within one generation (such as “Comet Lake” and “Ice Lake” for different subsets of its 10th Gen CPUs).
(Pro tip: Intel’s ARK site lets you drill down into processor generations and codenames. We often reference major Intel and AMD codenames before chips are released, and sometimes after; you can winnow our coverage by searching our site for a given codename.)
Knowing a CPU’s generation and/or codename is helpful to determine when it was released and to locate specific performance data on it. The two rivals typically refresh their processors every 12 to 18 months. Unless there’s some financial incentive to getting a laptop with an older chip, we advise buying the most recent generation to ensure you’re getting the newest features and the most longevity from your purchase. There’s more on chip lines in detail later in this guide, but here’s a cheat sheet to the laptop-CPU codenames of the last five years:
Processor TDP Ratings: It’s All in the Name (Well, Sometimes)
As mentioned earlier, AMD and Intel subclassify their processors by TDP rating. Both chipmakers denote their most desktop-like laptop chips with an H suffix, such as the Core i7-11800H and Ryzen 7 5800H, which each have a TDP rating of 45 watts. AMD further has an HX suffix for chips over 65 watts and an HS suffix for 35-watt chips, though Intel generically uses an H for mobile chips with TDP ratings between 35 and 65 watts.
Most consumer and business laptops use chips rated for 15 to 28 watts. Until recently, both vendors put a U suffix on 15-watt processors, though Intel has discontinued the practice. Starting with its 10th Generation “Ice Lake” chips, Intel switched to a G suffix plus a number indicating the integrated graphics performance.
To complicate matters, laptop makers can customize a chip’s TDP to suit their designs. For instance, an Intel “Tiger Lake” chip can be restricted to 12 watts and an AMD Ryzen 5000 U-series to 10 watts, though this is typically only done for ultra-compact and fanless laptops and tablets where heat output must be minimized.
The TDP rating is important since it determines a processor’s clock speed and thus its performance. Low-TDP chips such as the Ryzen 5000 U-series and Intel 11th Gen chips have the lowest base clocks (usually between 1GHz and 2GHz) and can maintain their high boost speeds only for short bursts; H-suffix processors can stay at their boost clocks longer. For tasks that use CPU power in spurts, however, low-TDP and high-TDP chips can perform similarly.
Cores and Thread Count: Breaking It Out by Line
The core and thread counts of Intel and AMD CPUs vary by product line and TDP rating. Intel’s Core i7, Core i9, and Xeon lines and AMD’s Ryzen 7 and Ryzen 9 chips have the highest, while Intel’s Celeron and Pentium and AMD’s Athlon have the lowest. As the following table shows, some brands have models with different core counts; this can vary by generation as well. We’ve mapped it out for CPUs released from 2019 onward.
Core counts generally increase with TDP rating. Intel’s U-series chips range from two to six cores while AMD’s go up to eight cores, though quad-core chips are most common for both. Thread count varies, too; Intel’s laptop Core CPUs fully support multithreading as of its 10th Generation, as do AMD’s Ryzen 5000 series. Celerons and some earlier Ryzen 3s, however, do not. You’ll want to look at any given chip’s specifics to verify whether it can handle twice as many threads as it has cores.
Special Intel Naming Conventions
Intel occasionally introduces different naming conventions for new features or special silicon. Its “Lakefield” CPUs, such as the Core i5-L16G7 in the Lenovo ThinkPad X1 Fold, forego the two-digit generation number after the brand identifier since they’re a hybrid of different generations. Fortunately, Intel’s most popular processors stick to its traditional naming scheme.
As mentioned above, most late-model mainstream Intel processors now end with a G plus number, with higher numbers indicating better-performing integrated graphics. (See the next section.) Last, a few low-end Intel chips such as the Pentium Gold 7505 have no suffix, leaving you to look up their specs on Intel’s site. C’est la vie.
Integrated Graphics Performance
As we said, most laptops other than gaming rigs and workstations rely on the integrated graphics built into the CPU. (Most systems with discrete GPUs can also switch to integrated graphics to save battery power when maximum 3D performance is not required, automatically making the switch without interrupting you.)
Until lately, most Intel mobile CPUs included what the company called UHD integrated graphics, sometimes accompanied by a performance rating such as UHD Graphics 600 or UHD Graphics 620. This silicon provided sufficient performance for desktop display connectivity, smooth onscreen animations, video streaming, and browser-based gaming, but fell far short of the oomph required for serious games, even for relatively undemanding titles such as Fortnite.
But graphics solutions aren’t just about gaming. They can also improve performance for photo and video editing and live streaming. AMD’s and Intel’s latest, better-performing integrated graphics are capable of all of that and even some gaming at low (720p) resolution. Intel’s current IGPs are called Iris Xe and Iris Xe Max, the latter technically a discrete GPU; AMD uses the moniker AMD Radeon Graphics for its current integrated silicon.
The AMD Radeon Graphics in today’s Ryzen 4000 and 5000 series chips perform much better than Intel UHD integrated graphics. (See, for example, the benchmarks in our HP Envy x360 13 review.) Intel’s answer to that challenge is Iris Xe, found in 11th Generation Core processors with a G7 suffix. (Chips with a G4 suffix still use Intel UHD Graphics, as do most of the company’s H-class chips.) Our Lenovo Yoga 9i review shows the G7 solution matching or outperforming AMD integrated silicon.
That said, AMD notebooks with Radeon integrated graphics tend to cost less than Intel laptops with Iris Xe. This can be a major point in favor of opting for an AMD CPU in your next notebook.
Business Considerations (Intel vPro and AMD Pro)
Home users can skip this section, but corporate buyers should know that the x86 duo offer remote management technologies—AMD Pro and Intel vPro—to help business IT personnel deploy and manage their computer fleets, including remote updates, repairs, and enhanced security features. The mixtures of services differ with each generation; check out the details at their websites.
AMD indicates whether a CPU has AMD Pro by simply including it in the product name, as with the Ryzen 7 Pro 5850U. Oddly, Intel is subtle about vPro support, leaving it out of product names though it’s listed on specific CPU product pages accessible via the invaluable ARK online database.
Laptop CPU Overclocking
Nearly all laptop CPUs are incapable of overclocking—that is, they don’t let users crank their clock speeds beyond factory ratings as some gaming desktop processors do. Intel’s rare mobile Core processors with a K suffix are exceptions.
The K means that the processor has unlocked multipliers, which can be used to modify the clock speed. (See our How to Overclock Your Intel CPU feature for desktop details; the process is much the same but with less thermal leeway on a laptop.) Intel’s only recent K-series mobile processor is the Core i9-11980HK.
Why not widely allow laptop CPU overclocking? The main reason is that laptops are built around strict thermal limitations. Increasing clock speed hikes power draw and generates more heat, which can cause overheating and instability or at least unwanted throttling. All told, laptop overclocking is a novelty found only on a few bleeding-edge gaming rigs with Intel K-series chips and ample cooling.
Summing It Up: Which Processor Should You Get?
The good news for consumers is that today, even despite well-publicized silicon shortages, is an excellent time to buy a laptop of any kind. Though an ultra-low-cost laptop might use a sluggish entry-level CPU, nearly all $500-and-up models will feature a responsive processor more than suited for everyday usage. There’s no shortage of power on the gaming, content creator, and workstation side of things, either; Apple, AMD, and Intel all have competitive offerings. (One note: Check out our gaming laptop guide for much more on picking a processor and the complex interactions among the CPU, GPU, and gaming performance.)
If you’re an Apple-laptop shopper, your choice is already made since the company started the switch to its in-house ARM chips in late 2020, unless you must stick with a legacy Intel MacBook for specific software reasons. At a minimum, M1 MacBooks are competitive with AMD- and Intel-based Windows laptops, and for specialized applications they can be even faster.
Windows and Chrome OS laptop buyers face a much greater choice with CPUs from AMD and Intel and even a few ARM chips thrown into the mix. Chromebooks generally deliver a smooth computing experience with any processor, though we’ve found ARM chips a bit more sluggish than x86. If you go for an AMD Chromebook, opt for one of the recent Ryzen C Chromebook-specific chips rather than one of the aged A-series. Similarly, an Intel Core CPU will serve Chromebook users better than a Pentium or Celeron if you tend to keep many tabs open at once.
AMD’s Ryzen 5000 and Intel’s 11th Generation Core lines are the mainstays of today’s Windows consumer and business market. They’re highly competitive on features and low in power consumption, though AMD often wins in raw CPU performance for core- and thread-hungry programs like content creation apps. AMD’s Radeon integrated graphics also top Intel’s UHD Graphics, though Intel’s 11th Gen Iris Xe silicon is roughly on par.
Outside of specific usage scenarios and benchmarks, however, like-priced Intel and AMD laptops will offer similar user experiences for most jobs. Video streaming, office productivity, and other everyday tasks are well within the reach of almost any Intel or AMD CPU. Even gamers can choose either brand; Ryzen 7 and Core i7 chips are competitive (though the latter will be easier to find). All this gives you the freedom to focus on the laptop’s design and features first and on the CPU second, though specific usage scenarios can dictate doing things the other way round.
Down and Dirty: A Look at the Very Latest CPUs
We haven’t tested every laptop CPU on the market—likely no one outside of Intel or AMD has, and maybe not even them. But with our general advice behind us, let’s wrap up with more specific laptop processor recommendations for various usage scenarios in the x86 aisle.
Beyond that general specification guide, you can get more granular with cheat sheets for the most common current-generation Intel and AMD laptop CPUs, along with their suggested usages and the kinds of systems in which you’ll find them. These final two tables should be very helpful when shopping the latest-model laptops.
You’ll still see plenty of notebooks on sale with last year’s or earlier chip generations, so an exhaustive list would be impossibly unwieldly. But if you look at AMD’s or Intel’s older CPU families, it’s usually easy to identify the parallel previous-gen versions of the chips listed below. You can safely assume in most cases that, if the core and thread count is the same, they’ll offer slightly lesser performance than the latest parts, but fit in the same relative place in the company hierarchy. If the price is right, don’t dismiss an almost-new CPU.
First, a look at the Intel lineup…
As you can see, the mainstream Intel processors are now dominated by the “Tiger Lake-U” line, though you’ll still find equivalent 10th Gen “Ice Lake-U” and “Comet Lake-U” chips for sale. Don’t ignore those if a laptop is a good buy. Intel released 45-watt, H-class 11th Gen processors for power-user notebooks in mid-2021, though you’ll still see many 10th Gen chips around.
Now for AMD…
As we said, you’ll see these processors in fewer laptops overall. Only fanatic users need to worry about the differences between the company’s “Lucienne” (Zen 2) and “Cezanne” (Zen 3) core architectures; there’s a much bigger difference between AMD’s U-series and H-series chips. The latter can get you an eight-core, 16-thread laptop that competes with many leading-edge desktop PCs.
Best of luck with your laptop hunt! As always, for nitty-gritty details, you can check out our endless stream of laptop reviews and list of our current favorites among overall, ultraportable, gaming, and workstation notebooks (with links to many reviews). Any given laptop’s performance in our CPU benchmarks won’t always reflect the results you’d get from the same chip in a different system—other factors like memory and thermals will come into play. But our detailed performance tests will get you close enough in your decision that you won’t be able to tell the differences without a stopwatch. Leave that stuff to us.
