
Preface
On 2020.10.8 (local time), AMD released its new generation of Zen 3 processors, which sparked a lot of discussion. AMD had been suppressed by Intel for a long time, but in recent years it doubled its market share thanks to its “Yes” factor. This time, AMD officially claimed a 19% IPC uplift with Zen 3, which really feels like rubbing Intel’s face into the ground. In the past, AMD usually competed by stacking more cores and going for the “big bowl, great value” style, but this time its single-core performance even surpassed the i9-10900K for the first time, beating Intel by 16% in Cinebench R20. Although the price went up a bit, the once dragon-slaying boy has now become the new dragon. Meanwhile, rumors said Intel’s 11th-gen desktop processors would still use the 14nm+++ process. Even so, it still makes people want to shout: “AMD Yes!”
With AMD’s processor launch as an excuse, I figured I’d summarize the current categories of consumer CPUs. The two processor makers most commonly used in desktop and mobile platforms keep releasing new products every year, and their product lines can get a little confusing for ordinary consumers. So I put together this summary to make it easier for everyone to shop and learn.

First, let me quote Baidu Baike’s introduction to the CPU:
The Central Processing Unit (CPU) is a super-large-scale integrated circuit and serves as both the computing core (Core) and control core (Control Unit) of a computer. Its main functions are to interpret computer instructions and process data in computer software. A CPU mainly includes the arithmetic unit (Arithmetic Logic Unit, ALU), cache memory (Cache), and the buses (Bus) for data, control, and status that connect them. Together with memory and input/output (I/O) devices, it forms one of the three core components of an electronic computer.
When it comes to consumer PC CPUs, people basically choose directly between AMD and Intel. A chip is really like a giant maze, with tens of billions of branching paths inside but only a handful of exits. From an engineering perspective, that’s basically like building a microscopic version of Wuhan, while also making sure the city’s “infrastructure” runs in an orderly way: what should go in goes in, what should go out goes out, and internal “public security” has to be fully guaranteed. Every “road” and every “building” has to survive trillions of passes without any issues. A chip that meets spec has no concept of “repair,” because its lifespan doesn’t depend on how frequently you use it, but on the limits of fundamental physics and quantum mechanics. In normal consumer understanding, it’s basically something that “can’t be worn out.” The reason there are so few top-tier processor manufacturers is that every single chip you buy has to satisfy all of the above conditions. That takes extremely deep technical accumulation and massive capital investment.

Intel’s Blue Team
Intel is currently the processor manufacturer with the highest market share in consumer desktops and mobile devices (for now). As a world-leading CPU maker, Intel’s blue logo always gives off a premium vibe. Intel is also nicknamed the “toothpaste factory,” because it dominated the field early on and basically had no real competition for a long time, which led to very slow technological iteration. It kept squeezing the same old process for years, and in some respects there hasn’t been a leap in performance from 4th-gen processors to today’s 10th-gen ones.

Common Series
Pentium
Pentium was once Intel’s glory days. I still remember Intel’s old ad jingle — “deng deng deng deng deng” — “give your computer a Pentium heart.” Pure nostalgia. Although advances in process technology and rising consumer spending led Intel to launch the Core series in 2006, pushing Pentium off its pedestal, the “Pentium” brand was never abandoned. As a familiar consumer brand, it gradually shifted toward budget products. But these days, even when people are squeezing their PC budget as hard as possible, they’ll usually grit their teeth and go for a Core instead of choosing Pentium.
Celeron
A processor positioned slightly lower than Pentium. It also has an OEM market and is generally used in mini PCs and set-top boxes. On Xianyu, the “junk-picking” crowd often uses it for soft routers…. But ordinary consumers actually use it pretty rarely. You might use one for building a NAS or a soft router, but for normal office work this series is not commonly chosen.
Core
This is Intel’s current flagship consumer series, and also the one most people choose. Core is the familiar i-series, divided into i3, i5, i7, and i9, with the X series going further in cores, clock speed, and performance. It’s split into desktop and mobile platforms. At the moment, the desktop platform is up to the 10th generation, while the mobile platform has reached the 11th generation.
Atom
This was the series Intel originally launched to grab a share of the smartphone chip market. What left the deepest impression on me was Lenovo’s K900, released in 2013 with the Atom Z2580. It had a silver, sharp-edged design, dual-core four-thread CPU, 2GHz clock speed, 2GB RAM, and a 13MP camera — pretty stunning at the time. They even got Kobe Bryant to endorse it.
Later, the mobile chip market became too brutal, and Atom insisted on using x86 architecture, which made software adaptation harder for developers. The market response was mediocre. At the end of April 2016, two new versions of the Atom mobile processor line that had originally been planned for release in 2016 were canceled, and Intel exited the smartphone chip market. It’s now discontinued.
Xeon
This is Intel’s processor line for servers. Wait, wasn’t this supposed to be about consumer-grade CPUs? Why are we talking about servers now? Actually, it’s because early Xeons had ridiculously good price/performance, and they used the same LGA 1150 socket. That meant even the lowest-end E3 models performed on par with Core i7 chips of the same era. The legendary “god-tier CPU” was the E3-1231 v3: Core i5 price, Core i7 performance — insanely good value. But today’s Xeons are truly Xeons: there are obvious differences from the Core series in price, performance, and stability, and the lineup has become more vertically segmented into Xeon E, Xeon W, Xeon, and so on. Still, those early E-series chips are still circulating in the market, because Intel kept squeezing toothpaste….
Model Numbers


xx xxx xx
The first one or two digits indicate the generation (because we’ve already reached the 11th generation, it’s now two digits)
The bigger the middle number, the more powerful it is
The last two may be absent; if present, they are usually a letter suffix
Suffix Buffs
| Suffix | Meaning |
|---|---|
| B | Desktop soldered standard-voltage version |
| F | No integrated graphics |
| G | Integrated graphics from AMD |
| H | Laptop soldered standard-voltage version |
| S | Special edition |
| K | Unlocked multiplier, can be overclocked |
| Q | Means quad, four cores |
| M | Dual-core mobile low-voltage, commonly used before 4th gen |
| Y | Mobile ultra-low voltage; back when ultrabooks were popular, they were basically all Y-series |
| T | Lower-power desktop version, slightly below standard voltage |
| U | Low-voltage mobile version; most business ultrabooks today are U-series |
| X/XE | HEDT (high-end desktop) processors, usually with very high core counts and support for quad-channel memory; all specs are far beyond ordinary PCs |
| G1-G7 | Intel’s new naming method for the 11th generation; G emphasizes graphics performance, and the bigger the number, the stronger it is |
AMD’s Red Team
In its early days, AMD helped Intel redesign product pricing and once fought Intel blow for blow. But later, because its process technology lagged too far behind Intel, Intel had already entered the 22nm era while AMD was still polishing 32nm. For many years AMD was pinned under Intel’s foot, and because of its low prices, it became the go-to “scam recommendation” for shady PC shop dealers targeting beginners. But AMD has always fought on two fronts, with both CPUs and GPUs, going up against Intel and NVIDIA respectively. Now AMD’s CPUs are manufactured by TSMC, letting it jump straight into the 7nm process era, and its RX series graphics cards can still put up a fight against NVIDIA. I’m really looking forward to the future of the consumer electronics market.
Series
Athlon
The early flagship of Team Red, mainly known for being cheap. K8 was once the main force before being overshadowed by other series, but because of its low price it won the favor of many consumers.
“Farmer Enterprise” Series
This is actually a nickname for some of the architectures AMD used during its development, such as Piledriver, Steamroller, and Excavator. Because the names were so distinctive and the prices were so low, people loved calling AMD a “rural enterprise.” It was also through AMD’s continuous polishing of this “farmer enterprise” line that the later Zen-based Ryzen came to be. During that period, the attractive pricing won over wave after wave of loyal fans.
Ryzen
This is AMD at its peak right now, going head-to-head with Intel’s popular consumer Core series. The most noteworthy thing is that AMD’s share in the mobile consumer market doubled almost entirely thanks to Ryzen. Even the proud ThinkPad released 3500U versions, which for a while became the go-to “entry-level but actually awesome” machines. Then this year the 4000 series sparked another wave of hype, with the 4800U pushing mobile processors all the way to 8 cores and 16 threads. Keep in mind, the famous toothpaste factory had been using dual-core mobile processors for more than a decade, and it wasn’t until 2018’s 8th-gen Core that Intel finally “generously” moved up to quad-core. The Ryzen lineup uses the R prefix, similar to Intel’s i-series, with R3, R5, R7, and R9, increasing in cores, clock speed, and performance. One especially notable chip is the sold-out 4800H. With its 45W standard-voltage 8 cores and 16 threads plus Vega integrated graphics, it was cheaper than Intel laptops using 10th-gen Core i5, while its performance absolutely rubbed 10th-gen Core i7 into the ground. At the time, even the strongest 10th-gen Core i7 only had up to six cores, which forced Intel to launch the 8-core 10875H in response — and then that got hammered too. The 4800H’s monster specs won over a huge number of fans, which led to supply shortages, and then it turned into a price-hiked dragon 😬
Threadripper
This is the HEDT platform processor AMD launched in recent years, competing against Intel’s X/XE suffix processors. The latest at the moment is the 3rd-gen 3000 series released in 2019, with 64 cores and 128 threads. Hearing that gives you a kind of “wow, that’s insane, but I’ll never use it” feeling. In reality, for ordinary PCs — gaming or office work, for example — there’s absolutely no need to consider the HEDT platform unless you just won 10 million and have no idea how to spend it. But HEDT platforms are extremely suitable for productivity workloads that heavily depend on threads and cores, such as video editing and 3D modeling/rendering. For ordinary consumers, AMD’s Threadripper and Intel’s X/XE basically serve as giant performance monsters used for brand endorsement: look, we can make CPUs this powerful, so come buy our chips.
EPYC
This is AMD’s latest server processor line, competing against Intel’s Xeon. Performance is very strong, and power consumption is also terrifying at 200W+. I can’t afford it and don’t need it, so I can only admire it from afar and hear that its compute power is amazing. But personally, I think aside from supercomputers, servers are still mainly about stability. Many standards were defined by Intel, and Intel also has a lot of technical acceleration patents. So if AMD wants to truly break into the server space, maybe it still has a long road ahead.
Model Numbers
xxxxxx The first four digits are numbers, followed by two characters that may or may not be present; if present, they are suffix letters
The first digit indicates the Zen architecture generation. Desktop and mobile originally differed by 1, but starting from the 5th generation AMD unified them for consistency. The latest currently starts with 5.
Starting from the second digit, bigger means more powerful, with 9 being the highest.
The third digit is generally 0, except for Threadripper and the Ryzen 3950X.
The fourth digit is currently 0 for all models.
Suffix Buffs
| Suffix | Meaning |
|---|---|
| E | Low power, roughly equivalent to Intel’s T |
| X | Supports XFR dynamic frequency scaling, roughly equivalent to Intel’s K (though now AMD’s whole lineup is multiplier-unlocked) |
| U | Laptop low-voltage version, same as Intel |
| H | Laptop standard-voltage version, same as Intel |
| G | Means it has integrated graphics; by default, models without G do not have integrated graphics, the opposite of Intel’s F |
| T | Means Turbo support, similar to Intel’s Turbo Boost |
Ending
This is just an introductory article, simply covering the product lines of Intel and AMD. If you’re a complete beginner, learning to build your own PC and not planning to spend too much time on it, then my suggestion is simple: within the price range you can afford, buy the most expensive one. After all, even if you don’t know the hardware, doesn’t money know what’s good?

If you’re a newcomer and want to become a PC-building enthusiast, or just want to understand the specs well enough to avoid getting fooled by shady sellers, I’ll continue later with a more advanced summary, including CPU parameters such as frequency, cores, threads, cache size, architecture, process node, power consumption, and interfaces.
References
- Here’s a desktop CPU performance tier chart, click here
- Intel® Processor Names and Numbers
- AMD Ryzen Threadripper 3rd Generation release date, news, and rumors
- AMD Ryzen Mobile Processors with Radeon Graphics
All articles in this blog, unless otherwise stated, are licensed under @Oreoft . Please indicate the source when reprinting!



