Think of processors as the brains of a laptop. They control practically everything: from how your laptop interprets hand movement on a mouse to the more abstract (maybe) complex stuff, like how a computer connects to wifi, transmits data and renders content from the internet.
Good processors make good laptops, every other thing being equal. But with so many processors on the market today and the sometimes cryptic naming and branding, it can be difficult to tell apart a great processor for your needs.
A basic primer on Processor technology
Processors are complex technology. Here’s an introduction so you get the hang of what really matters.
1. Clock speeds
Let’s use a medical anecdote to explain CPU clock speeds, your heartbeat. The average human heart beats 60 – 100 times per minute, so on average, 1 – 2 beats per second. With each beat, your heart attempts to pump blood to every other part of your body.
Clock speeds follow this same concept. To process complete computational tasks, to ‘pump blood in your PC,’ CPUs go through a cycle or “beat.” Much like how beats are recorded per minute, as in 50 beats per minute, computational cycles are measured in clock speeds, in hertz, which is a measure of how many of these cycles or beats a CPU goes through per second.
A CPU with a 2.8GHz clock speed rating, for instance, executes 2.8 billion cycles each second. So in every second, it’s performing 2.8 billion compute cycles to solve the computational tasks your PC needs to operate – it’s beating 2.8 billion times to deliver enough blood for your PC to survive. Depending on many other factors, a singular beat might be enough to get the blood you need for that second to your brain; or you might need several beats to get blood to your kidneys.
Same thing for CPU cycles measured in clock speeds. One cycle might be all that’s needed to compute your last keypress; other times, a CPU might need to go through extra cycles to compute a task like rendering images on a PC screen.
So far, I’ve tried to separate the CPU compute cycles from actual CPU performance. Faster CPU cycles don’t necessarily mean faster (better) CPU performance for the same reason that faster heartbeats don’t always mean better blood delivery to organs in medicine. Well-conditioned athletes get all the blood they need to all other parts of their body with fewer beats, so it’s not uncommon to see high-performance athletes with a heart rate of 40BPM, and that’s enough.
With each beat, they might theoretically deliver more blood, in turn supplying more tissues than the average joe. Unconditioned persons, like you and me, need more beats to get blood to where it’s needed. Performance is the actual blood that gets to the end organ. In computational terms, the number of computational tasks solved after each cycle or the IPC. CPU speeds are merely a measure of frequency, how much the CPU beats and not how much it actually computes per second.
While it might be beneficial to think that a rapidly beating heart is better to meet the needs of a body, in many cases, with proper optimization, fewer beats can be better. For CPUs, many factors go into determining the efficiency (performance) of each cycle – the processor architecture, additional performance add ons and base level optimizations. This is why it’s true that newer gen CPUs might perform way better than older-gen CPUs despite the latter having faster CPU speeds than the former.
Why’s this important
As you shop for a computer, you’ll notice that processor speeds are one of the most marketed laptop/computer spec. Every manufacturer brandishes it like it’s the holy grail of performance, but the reality is it’s just a segment and given the current trend of CPU manufacturing – with the production of more optimized or ‘conditioned’ processors – it is a less so important segment.
There are many other important CPU factors you should take into consideration, and we’ll talk about them. Processor speeds are cool, and all, and faster is usually better but don’t fixate on them when shopping for a CPU. Ideally, they should be the determining factor when you have the same CPU family and you’re looking to make a pick. Across different CPU types like between Intel’s different CPU configs or AMD CPU configs or between both manufacturers, you should factor in other determinants like core count above just plain CPU speeds.
2. Cores
Two heads are better than one, and with CPU cores, it’s the same logic, or more specifically, two or MORE cores are better than one. CPUs perform computational tasks sequentially – one after the other. So, unless one started computational task is completed or paused, a CPU cannot hop on to the next. The next will exist in a queue, waiting for its turn to be processed.
Now imagine a task on your computer with several segments, something like a scene in a video game with graphic components, sound components and physics components, this is overly simplified, but you get the idea. A CPU would need to process each of these segments one after the other and, in practice, would require more time to do so. It’s like unpacking a load of boxes from a delivery truck into a warehouse; no matter your work rate, there’s a limit to how fast you can churn through a load of one thousand boxes if you’re working alone.
When you have a partner, your speed automatically doubles; whether it’s a team of two lean guys or buff, hyperactive dudes, doubling the number of hands immediately steps up the speed.
This same logic follows through with CPU cores. You can think of a CPU core as a separate bloc in the CPU architecture capable of executing computational tasks independently. If it’s one core, it’s just one person unloading boxes; two cores, and it’s two people unloading, three cores, three people and so forth.
Why is this important
It follows that the more cores a CPU has, the better it is for processing ‘multi-threaded’ computational tasks (which is a major bulk of the tasks your CPU will process). And that’s largely true – all things being equal, you want more cores in your CPUs.
More is, however, usually more expensive, and while it’s possible to buy consumer-grade CPUs with up to 64 cores, anything above eight cores should be capable enough to churn through even the most resource-intensive applications.
At the very least, you want a CPU with four cores. I use a dual-core setup as my portable PC for writing articles like this, and despite it being a recent gen CPU mated to 2GB of graphics memory, there’s a more than visible lag when I run resource-intensive apps like Davinci Resolve and Adobe After Effects.
3, Threads
Through some smart engineering, chip manufacturers have been able to configure certain CPU cores to handle more than one task simultaneously at a given point in time. (Remember I said each core typically handles just one computational task in time, the next is placed on a queue waiting for the resolution of the preceding task).
The technology is called multithreading, and it’s basically a CPU’ hack’ that allows single cores in a processor architecture to virtually split themselves into separate processing units (cores). Each virtual core is known as a logical core as opposed to actual physical cores we talked about earlier. Performance-wise logical cores are not the same thing as physical cores, but they come close.
Imagine each CPU core to be a factory production line. Down the line, there are two technicians for each production station. A production station, despite having two staffers, can only complete its part of the production run after the preceding station completes its own part. This is the traditional way processors handle tasks – sequential computing, we discussed this earlier.
Now we can take one technician from each station and use them to form another production line in the same factory. Or, more accurately, we can choose a supervisor and ask him to organize the extra staff in each station to form an accessory production line. This new virtual production line is capable of handling other production runs in the factory. The selected staffers are still physically part of the original production line; the difference is they can operate out of sync and coordinate with other delegated staffers to handle other production runs.
Multithreading uses a special logic to round up unused parts of the CPU processing power to assemble a virtual parallel processing unit. The logic is the supervisor in our analogy; unused processing power are the extra staffers in the production queue, and the virtual processing unit or thread is the accessory production line that can run independently.
This example demonstrates why multithreading is not an exact substitute for a multicore architecture despite the obvious similarities. In the end, each thread is still running off resources from a single CPU core. When the CPU is operating at max load, the ‘free’ resources used to assemble accessory threads will be used up by the main CPU thread, thereby dropping performance. With a multicore setup, workload on one core has no effect whatsoever on the next core.
Think of it this way; multicore is setting up separate factories or distinctly separate production lines alongside each other. Multithreading is using one factory production line to run separate production runs through some smart man managing and scheduling techniques
Why is this important
Intel calls its multithreading functionality Hyper-Threading, on AMD CPUs you might hear the phrase Simultaneous Multithreading. As was the case with CPU clock speeds, both manufacturers market the hell out of this feature when it’s included in a CPU.
Do you need multithreading? It’s a nice feature to have. A CPU with multithreading support is obviously better than the same CPU without the feature. That said, you shouldn’t be paying more for a CPU just because it has multithreading support. You’re better off spending the money on more cores or on better clock speeds.
4. CPU cache
To perform all its processing magic, CPUs need a steady stream of information from various other PC components, most notably the RAM module. Sometimes waiting for the data feed from other PC components can slow down the CPU, much like how waiting for a supplier shipment can slow down the production run at a factory production line.
If you were a factory manager, one way to address this would be to order supplies in time before they’re needed and maybe store them in an in-house factory so you can retrieve them exactly when they’re needed.
The CPU-cache architecture does exactly that, provision and store needed data close to the CPU for speedy access. The cache is the warehouse in this framework. The CPU, on its part, uses some really smart algorithmic computation to determine which data it needs fetched based on its computational needs and the stream of data it’s currently working on. So, for instance, if the current task is to process navigation to house H on a street with houses numbered A – Z, the CPU can request data on house F, even while it’s still resolving the data for house B. Again, overly simplified, but you get the idea.
Why is this important
The bigger the cache size, the more ‘needed data’ a CPU can store and the faster it can operate, at least theoretically. I say theoretically because, past a certain threshold of cache size, the law of diminishing returns takes effect. The good news is most CPUs aboard laptops and PCs today carry large enough and approaching-this-limit caches, but to maintain objectivity, the ideal CPU has a cache that’s at least 14MB.
5. Power rating
All things equal, the power a CPU uses in watts is a measure of its throughput performance. That’s to say, high-performance CPUs use more power than low-performance CPUs, and that follows simple logic. If a CPU is doing more work, it’ll require much more power and vice versa.
There’s a lot more nuance to CPU and power usage ratings, but most of that is outside the scope of this discussion.
Why is this important
Manufacturers typically have special naming conventions to denote the power rating of a CPU.
Both Intel and AMD high performance – high power rated CPUs for laptops carry the H suffix to their name. So when you see a Core i7 12700 H CPU, you know you’re dealing with a high performance – high power CPU, specifically CPUs with a power rating of 45W.
Intel CPUs with a P suffix to their name are 28W CPUs. AMD variants with a HS suffix are 35W CPUs. So a bit lower than the current 45W H suffixed CPUs but still very much high-performance type.
My current laptop has a CPU with the U suffix, and U denotes ultra-low power on both Intel and AMD setups. Actual power ratings vary, it could be 15W – 9W for Intel U CPUs or as low as 10W for AMD CPUs.
Power ratings are an important consideration to make when CPU shopping. You don’t want a U marked CPU if you plan on using your PC/Laptop for some performance heavy lifting – serious graphics design, gaming or video editing all included. H and X marked CPUs are top of the bunch, and usually they’re the CPU type to support overclocking. So if top tier performance is what you’re gunning for, then this is what you want.
Mid-tier P or HS (AMD) are also good performance-wise. Provided they thick the other check boxes for performance we’ve covered – clock speed, cores – they should comfortably handle high-performance tasks like graphics designing, video editing and high-end gaming (at reasonable settings).
Note that some other CPUs don’t carry the power specific X, H, HS, or U rating suffix. They carry other naming suffixes, which we will talk about shortly. Feel free to regard these CPUs as mid-tier, with everything I’ve said concerning mid-tier CPUs applicable to them as well.
Basics aside, let’s address the other consideration you need to make when choosing a CPU for your laptop.
AMD vs Intel – which is the best option
If you’re new to this, you are probably not a fan of either, and that’s a good place to start. The PC enthusiast sphere is littered with fans on both sides of the divide – many will swear by Intel CPUs, and many others are convinced AMD is the way to go if you want bang for buck.
Setting the facts straight
AMD started as an underdog in CPU manufacturing. Intel was the more established manufacturer with decades worth of experience and a lot of proprietary patented tech. In trying to gain market share, AMD focused on churning out very budget-friendly but still capable, to an extent, CPUs. Of course, some corners were cut, and in those yesteryears of CPU manufacturing, AMD CPUs were known to come with a lot of issues. Overheating was one of them; progressive drops in performance with age was another.
But a lot has changed since that time. With its recent Zen 3 Rembrandt CPUs, AMD has shown that it is way past the time when its CPUs were riddled with quality control issues. Recent gen AMD CPUs match the Intel competition for quality and reliability, and some even go as far as thumping Intel CPUs on the performance front. This is certainly true for the Zen 3 (Ryzen 5 – Ryzen 9 series) processors. Does this mean the next CPU in your laptop/PC should be an AMD setup? Nah, this is just to say, much like Intel, AMD CPUs have come of age.
So which should you choose?
The best CPU for your peculiar use case. If that’s an AMD CPU, fine; if it’s an Intel setup, all good as well. Determining the best CPU for your need basically comes down to two things:
- Your budget
- Your needs
Budget
CPUs are one of the most expensive components in a laptop or PC build. It’s simple, higher-end CPUs will cost more low specced CPUs with fewer cores, lower clock speed and less optimized processor architecture.
Your needs
What do you plan on doing with your laptop or PC? Just writing, watching movies, surfing the web and some other lightweight tasks? There’s no need to burn cash on a top of the line CPU.
If you’re going to be doing some performance heavy lifting, anything from gaming to video editing and graphics designing, then it pays to invest in a better mid-specced CPU.
If you’re doing some high-end gaming (max settings everywhere), heavy-duty video editing or animated content production, you want a top of the line CPU.
Obviously, the goal of every PC/laptop shopper is to find the middle ground between performance and affordability, something that does what you want in and around the budget you can afford. Here’s a template to guide your decision.
Understanding naming conventions
You’re going to come across a lot of cryptic CPU naming schemes as you shop for your next laptop or PC build. Understanding these naming conventions is important to making the right selection. Here’s an overview to get you started.
Intel naming
Processor categories
Intel processors can be grouped into three categories
- Intel Core
- Intel Pentium and Intel Celeron
- Intel Xeon
Intel Core processors are the most common consumer-grade processor category. You’ll find them widespread in laptop and PC setups. This category combines consumer-grade performance and affordability.
Intel Pentium and Intel Celeron CPUs are lower-tier processors that are generally way cheaper than Core or Xeon setups.
Intel Xeon processors are top of the line processors with myriad special features oriented for high-performance setups like servers and high-end workstations.
Processor subcategories
The popular Intel Core category is further subdivided into subcategories Core i3, Core i5, Core i7 and Core i9. The higher the number across the same processor generation (we’ll cover this shortly), the better the performance. So a Core i9 will have better performance and better processor enhancements compared to a Core i7 processor of the same generation.
Intel Celeron and Intel Pentium processors don’t have these subcategories.
Generation number
Processor technology evolves with time, and usually, manufacturers update the processor architecture to integrate newer improvements, features, and performance boosts yearly. Each update is a generation, and Intel’s naming convention makes it such that the first number (or two numbers in the case of 10th gen and upwards processors) after the category or subcategory naming denotes the CPU generation.
So a CPU with the name Intel Core i5 8600H is an 8th gen CPU. Again higher numbers are better – a 10th gen CPU will have better performance, feature set, and enhancement improvements than a 9th gen processor.
SKU numbers
Every digit after the gen number represents the CPU SKU which is basically a chronological production number assigned to CPUs as they roll off the production line. Generally speaking, higher numbers compared to lower numbers in the same processor category and for the same gen indicate a more recent CPU, usually with more features and enhancement.
Letter modifiers
The final piece in Intel’s naming convention for CPUs is the letter modifier – the last character you’ll find in the CPUs name. Each letter suffix tells you what you get with any one particular Intel CPU, so more like a general overview.
- E for CPUs with embedded graphics processors.
NB. Graphic processors are special processors optimized for processing and rendering visuals on your machine. Graphics processors can be integrated (embedded), discrete (separate from the main CPU but still bundled in with the CPU) or dedicated (existing as a separate processing block).
- F for CPUs which require a separate dedicated graphics processor card to function (Laptops rarely, if ever, ship with this kind of processor)
- G for CPUs with their own discrete (separate but still included in the package) graphics processor.
- H for CPUs rated high performance. H CPUs are usually top of the line models used in high-end gaming and workstation setups.
NB that H CPUs are high end ‘mobile’ CPUs. Mobile here means they’re made for laptops and not desktop setups.
- K for unlocked CPUs. Unlocked means the CPU can be ‘overclocked’, and overclocking is simply redlining your CPU to get more than usual performance juice out of it. Most unlocked CPUs are usually high-performance CPUs, so you’re more likely to see the K suffix used in conjunction with an H suffix like so – HK.
- Q for CPUs with four cores – ‘quad-core. Again most quad-core CPUs for laptops are usually high performance, so you’ll see this modifier used alongside the H modifier like so HQ.
- T for CPUs designed to be power efficient.
- U for ‘ultra-low’ power CPUs which trade performance for energy savings.
- Y for extremely ultra low power CPUs which use even smaller amounts of energy, making them extremely suited for battery savings.
- X for super high-performance CPUs. X rated CPUs are the absolute top of the line CPUs made by Intel.
AMD Naming
There’s a lot less variation in the naming convention but surprisingly more confusion around AMD’s naming structure.
Processor category
AMD CPUs can be broadly classified into five categories:
- Ryzen 3 CPUs
- Ryzen 5 CPUs
- Ryzen 7 CPUs
- Ryzen 9 CPUs
- Threadripper CPUs
Across each generation (we’ll address AMD CPU gens next), you’ll find the full spectrum of these CPU categories. What differentiates one from the next? Basically, performance but more specifically, the number of cores.
- Ryzen 3 CPUs have up to four cores
- Ryzen 5 CPUs have up to six cores
- Ryzen 7 CPUs have up to 8 cores
- Ryzen 9 CPUs have up to 16 cores
- The range-topping Threadripper CPU class has up to 64 cores
Again, the higher the number, the better, but remember that this only holds through across same-gen CPUs.
Generation number
As was the case with Intel CPU, AMD refreshes its processor architecture periodically (usually yearly). Each refresh is theoretically better, with more features and better optimization. You can tell what generation a CPU is by looking at the first number after the processor category name. So for a Ryzen 7 1700X processor, the generation number is 1, meaning this is a first-gen Ryzen 7 (8, octa-core) processor.
Currently, AMD has four generations of Ryzen CPUs. Annoyingly there’s a chronologically stagger, so the numbers don’t really match up with the generation. Here’s what I mean:
- Gen one Ryzen CPUs are tagged Ryzen * 1***
- Gen two CPUs have the Ryzen * 2**** tag
- Gen three CPUs have the Ryzen * 3**** tag
- Then Gen four CPUs have the Ryzen * 5 **** tag as opposed to Ryzen * 4 **** tag
It gets weirder when you move over to APUs. You know how Intel has CPUs with integrated graphics, those with the E or H label, well, AMD call its CPUs with integrated graphics APUs and they get their own special gen naming deviation.
- Gen one Ryzen APUs are tagged Ryzen * 2***
- Gen two APUs have the Ryzen * 3**** tag
- Gen three APUs have the Ryzen * 4**** tag
- Then Gen four APUs have the Ryzen * 5 **** tag as opposed to Ryzen * 4 **** tag
Performance modifier
The next digit after the generation number is the performance modifier. It indicates the performance level of that specific CPU within its gen and category class.
4,5,6 indicate moderate to high performance, going up 7, 8 indicates top-of-the-line enthusiast-level performance.
SKU numbers
SKU number naming for AMD CPUs is similar to that for Intel CPUs. For AMD CPUs, the SKU number comes after the performance modifier, which we just talked about, and basically, higher numbers are better since they indicate a ‘bump’ in speed according to AMD.
Letter modifier
Again like Intel, AMD CPUs also bear additional letter suffixes to give you a general overview of what the CPU is like.
- G marked AMD CPUs are those with integrated graphics
- H marked AMD CPUs high-performance, laptop type CPUs
- M marked are extremely ultra-low power AMD CPUs built for extreme energy savings with a cutback on performance, of course
- S marked AMD CPUs
- T marked AMD CPUs are ultra-low-power Desktop-type CPUs
- U marked AMD CPUs are ultra-low-power Laptop type CPUs
- X marked AMD CPUs are the range-topping option for Desktop setups.
NB. Unless stated otherwise, ‘Laptop-type’ or, in AMD parlance, ‘mobile’ CPUs are APUs. That is, they come preloaded with integrated graphics. AMD calls its integrated graphics modules’ Radeon Graphics.’
A note on Integrated graphics and dedicated graphics
I already introduced graphics processors earlier on, and basically, they’re specialized processors dedicated to handling the visual components of computing.
Laptops and Desktop PCs can come with either an integrated graphics processor or a dedicated graphics processor.
Integrated graphics processor
If the graphics processor is integrated, it means it’s embedded into the original CPU architecture. In more technical terms, it is part of the original CPU ‘die.’ Most, if not all, laptop CPUs will ship with an integrated graphics processor. Integrated graphics processors are kinda limited on the performance front, so while they’ll work well for everyday tasks like surfing the web, using your computer, running mid-performance graphic design apps, they’d struggle with the performance-intensive stuff.
Dedicated graphics processor
Dedicated graphics processors or GPUs (graphic processing units) or graphics cards are standalone processing units that exist outside of the original CPU setup. GPUs pack more power than integrated graphics processors and, as such, are able to handle even the most performance-intensive applications.
Any laptop meant for high-performance graphics design, 3D design or animation, video editing or animation, or high-end gaming needs a GPU to work fine. But GPUs are expensive, and you’ll typically find them in high-end, expensive laptop setups.
Which should you go for?
Well, that’s entirely dependent on your needs.
- If you plan on using your PC for any serious graphics design, video editing or animation tasks, then a dedicated graphics card is a needs must.
- If you need a laptop for everyday PC stuff, then a GPU is overkill; you’re good with the integrated graphics processor most laptops ship with.
To conclude
That about sums up all you need to know about CPUs. Remember, the right CPU for you is the CPU that best satisfies your needs. Not the Intel CPU used by your favorite Youtuber or the AMD CPU making the rounds for best budget CPU. To make the right selection identify your needs, then use this guide to select the best possible option based on your newfound understanding of CPUs.
