▷ Amd ryzen

AMD Ryzen are the most fashionable processors today, and it is not for less for the good work that AMD has done with these chips. Among its most important characteristics we find: a very well optimized manufacturing process, a very good engineering design, brutal performance in both simultaneous tasks, consumption and great temperatures.

We have prepared this post to explain everything you need to know about AMD Ryzen and its Zen microarchitecture. Do you want to stay up to date with this generation of processors that have marked a before and after?

Index of contents

Before starting we leave you the AMD area that we have designed on our website:

What is AMD Ryzen and Zen architecture?

AMD Ryzen is the trade name for all the processors released to the market by AMD since last year 2017. This name refers to AMD's next-generation microarchitecture, " Zen, " and to AMD's resurgence thanks to these new processors. AMD Ryzen came to the market after AMD will go more than five long years without being able to compete with Intel, because its previous processors, the AMD FX, did not turn out to be competitive either in performance or in energy efficiency, causing the company to lose almost all its market share.

Key features of Zen microarchitecture

AMD understood the failure of the Bulldozer architecture that brought AMD FX to life, thereby taking a 180 degree turn with the design of its new Zen architecture. To return to the path of success, AMD hired Jim Keller, a prestigious architect of CPU that had led AMD's golden age in the market with Athlon 64 processors and its K8 architecture. Keller and AMD had a daunting task ahead of them, as AMD had lagged far behind in performance and energy efficiency compared to Intel, rightly losing users' trust in their processors.

Zen's design is based on two fundamental keys:

• 14nm FinFET Manufacturing: AMD FX processors were manufactured using a 32nm lithographic process, putting them at a distinct disadvantage compared to Intel's 14nm designs. AMD understood that it needed to use the most advanced technologies to be able to close the gap with its great rival. That's where Gobal Foundries and its advanced 14nm FinFET process comes into play. The jump from 32nm to 14nm represents a huge improvement in energy efficiency, and the ability to put more transistors in a processor of equal size, more transistors equals higher performance. Design focused on improving the IPC: The IPC was the second Achilles heel of AMD FX processors. This concept represents the performance of a processor for each core and for each frequency MHZ. The Bulldozer architecture is characterized by having a very low IPC, so it was the second key point to solve with Zen. The Zen architecture duplicates many of the internal elements of the core, making them much more powerful than the Bulldozers. AMD has managed to improve the IPC by 52% compared to the Bulldozer architecture, a huge advance that has not been seen in more than ten years.

Zen architecture has been developed for more than three years within AMD, a long process of meditation on what your future processors should be. The name Zen is due to a Buddhist philosophy originating in China in the Vila century which preaches meditation in order to achieve enlightenment that reveals the truth. It seems like the perfect, tailor-made name for the company's new architecture.

SenseMI technology is a key element of the Zen architecture. Actually, this name encompasses four main characteristics that make these processors work really well:

• Pure Power: AMD Zen seeks maximum energy efficiency, the company wants a single core for all its products, so it must be highly adaptable to very different use situations, from large servers to the most compact laptops. This technology is responsible for optimizing the use of energy based on the working temperature of the processor. Zen-based processors include hundreds of sensors spread across its entire surface, making it possible to know very precisely the operating temperature of each part of the processor, and spread the workload without sacrificing performance or energy efficiency. Precision Boost: Once the temperature of the processor is known precisely, and if it is within the allowed, it is necessary to increase the frequencies to achieve the best possible performance. This is done by Precision Boost, a technology that increases voltage and clock speed very precisely in 25 Mhz steps. Precision Boost and Pure Power come together to enable Zen-based processors to achieve the highest possible clock frequencies. XFR (eXtended Frequency Range): There are situations where not all cores in a processor are used, causing power consumption and temperature to drop, leaving room for a further increase in clock frequency. That's where XFR comes in, taking the performance of Ryzen processors to a new level. Neural Net Prediction and Smart Prefetch: These are two technologies that rely on artificial intelligence techniques to optimize workflow and cache management with preloading of intelligent information data, optimizing access to RAM and processor caches. Artificial intelligence is the order of the day, and AMD also includes it in its best processors.

We recommend reading our best PC hardware and component guides:

• AMD history, processors and graphics cards of the green giantGuide to the best graphics cards How to clean a graphics card step by step

Zen internal design

If we focus on the internal design of Ryzen processors, the Zen architecture is made up of quad-core units , these units are what are called CCX complexes. Each CCX consists of four Zen cores along with 16 MB of shared L3 cache between all of them. This means that one kernel can access a greater amount of cache than it would be if it is shared fairly, whenever it needs it and another kernel needs less.

Within each CCX, the cores and the cache communicate with each other through the Infinity Fabric bus. It is a bus designed by AMD that is highly versatile, this bus serves to communicate with each other all the internal elements of a processor, and can even be used to communicate with each other different processors mounted on the same motherboard. Infinity Fabric is a highly versatile bus, which can cover a large number of needs. But not everything is pink, being able to do many things usually involves some inconvenience and this time is no exception. Infinity Fabric has considerably higher latency than the bus used by Intel in its processors, this higher latency is the main cause of Ryzen's lower performance in video games.

Almost all AMD Ryzen processors are made up of dies or silicon tablets that contain two CCX complexes, these two CCXs also communicate with each other through the Infinity Fabric bus. This means that all AMD Ryzen processors physically have eight cores, the company deactivates several of these cores in order to offer a wide range of processors from four to eight cores.

A final important feature of Zen is SMT technology, short for simultaneous multithreading. It is a technology that allows each core to manage two threads of execution, which allows doubling the number of logical cores of a processor. Thanks to SMT, Ryzen processors offer four to sixteen processing threads.

First generation Ryzen processors

The first Zen-based processors were the Ryzen 7 1700, 1700X, and 1800X, all released in early March 2017 for the AM4 platform. All of them demonstrated great performance from the start, being exceptionally good at workloads that make use of a large number of cores. The Zen architecture upgrade has been so great that these processors are able to quadruple the performance of the AMD FX-8370, AMD's previous top-of-the-range processor. These processors quickly caught the attention of image professionals, as they enabled the rendering of very high resolution videos at high speed. To all this are added very competitive prices, AMD offered its eight-core processor for the same price that Intel sold you a four-core processor.

Despite this great improvement, these processors were even inferior to Intel in a sector of the market that nine big money, video games. Intel was still the king of video games, although it must be said that the distance with AMD had been reduced alarmingly for Intel, for the first time in many years, AMD had processors capable of putting Intel in trouble even in its most important field. favorable. The great price-performance ratio of the AMD Ryzen attracted players very quickly.

A little later, in the spring and summer of 2017, Ryzen 5 1600, 1600X, 1500X, 1400, 1300X and 1300 processors arrived, offering between four and six cores, completing the entire range of first-generation AMD Ryzen processors. They are all manufactured using the Global Foundries 14nm FinFET process, the code name for their die is Summit Ridge.

AMD Ryzen 7 1700, 1700X, and 1800X

They are all eight core processors and sixteen processing threads, the only difference between them is the operating frequency. All of them support overclocking, which is why many users bought the Ryzen 7 1700, the cheapest of the three, and overclocked it to the frequencies of the Ryzen 7 1800X, achieving the best performance while spending less money. All of them have a 16 MB L3 cache and a 4 MB L2 cache. The following table summarizes all its characteristics.

Processor	Cores / threads	Base frequency (GHz)	Turbo frequency (GHz)	Cache L3 (MB)	L2 cache (MB)	Memory	TDP (W)
AMD Ryzen 7 1800X	8/16	3.6	4.1	16	4	DDR4-2666 dual-channel	95
AMD Ryzen 7 1700X	8/16	3.4	3.9	16	4	DDR4-2666 dual-channel	95
AMD Ryzen 7 1700	8/16	3	3.7	16	4	DDR4-2666 dual-channel	65

AMD Ryzen 5 1600, 1600X

Both are physical six-core and twelve-thread processors, they came to offer a much better balance between price and performance, especially in video games. They maintain a 16MB L3 cache and a 3MB L2 cache. The Ryzen 5 1600X is capable of a maximum frequency of 4 GHz, while its little brother settles for 3.6 GHz.

Processor	Cores / threads	Base frequency (GHz)	Turbo frequency (GHz)	Cache L3 (MB)	L2 cache (MB)	Memory	TDP (W)
AMD Ryzen 5 1600X	6/12	3.6	4.0	16	3	DDR4-2666 dual-channel	95
AMD Ryzen 5 1600	6/12	3.2	3.6	16	3	DDR4-2666 dual-channel	65

AMD Ryzen 5 1500X and 1400

They are the first-generation AMD Ryzen quad-core, eight-thread processors, still maintaining their 16MB L3 cache and a 2MB L2 cache. These processors start from 3.5 GHz and 3.2 GHz and are capable of reaching 3.7 GHz and 3.4 GHz.

Processor	Cores / threads	Base frequency (GHz)	Turbo frequency (GHz)	Cache L3 (MB)	L2 cache (MB)	Memory	TDP (W)
AMD Ryzen 5 1500X	4/8	3.5	3.7	16	two	DDR4-2666 dual-channel	65
AMD Ryzen 5 1400	4/8	3.2	3.4	8	two	DDR4-2666 dual-channel	65

Ryzen 3 1300X and 1200

All of them are quad- core and four-thread processors, in both cases they have an 8 MB L3 cache and a 2 MB L2 cache. They are the entry-level models to the first generation of Ryzen. Its base frequencies are 3.5 GHz and 3.1 GHz respectively, and turbo frequencies of 3.7 GHz and 3.4 GHz.

We recommend reading our post on Intel Core i3 8100 vs i3 8350K vs AMD Ryzen 3 1200 vs AMD Ryzen 1300X (Comparative)

Processor	Cores / threads	Base frequency (GHz)	Turbo frequency (GHz)	Cache L3 (MB)	L2 cache (MB)	Memory	TDP (W)
AMD Ryzen 3 1300X	4/4	3.5	3.7	8	two	DDR4-2666 dual-channel	65
AMD Ryzen 3 1200	4/4	3.1	3.4	8	two	DDR4-2666 dual-channel	65

Second-generation AMD Ryzen processors

In April of this year 2018, the second generation AMD Ryzen processors were launched, manufactured at 12 nm FinFET and with a Zen + architecture that includes some improvements focused on increasing the operating frequency and reducing the latency of its internal elements. A MD assures that the managed to reduce the latency of the L1 cache by 13%, the latency of the L2 cache by 24% and the latency of the L3 cache by 16%, this means that the IPC of these processors has increased approximately 3% compared to the first generation. These improvements help achieve better processor performance, albeit primarily in video games, which are very sensitive to latencies. All of them are made using the Global Foundries 12nm FinFET process, the code name for their die is Pinnacle Ridge.

AMD Ryzen 7 2700X and 2700

They are the successors to the Ryzen 7 1700, 1700X, and 1800X. This time AMD has decided that the intermediate model does not make sense, so it has only released two processors. Its fundamental characteristics are the same as those of the first generation, although they enjoy higher clock speeds and improved latencies.

We recommend reading our post about AMD Ryzen 7 2700X vs Core i7 8700K at equal frequency

Processor	Cores / threads	Base frequency (GHz)	Turbo frequency (GHz)	Cache L3 (MB)	L2 cache (MB)	Memory	TDP (W)
AMD Ryzen 7 2700X	8/16	3.7	4.3	16	4	DDR4-2933 dual-channel	105
AMD Ryzen 7 2700	8/16	3.2	4.1	16	4	DDR4-2933 dual-channel	95

AMD Ryzen 5 2600X and 2600

They have arrived to succeed the Ryzen 1600X and 1600. They also maintain the same fundamental characteristics, although with higher clock frequencies and somewhat lower latencies. They are considered the current processors with the best balance between price and performance on the market, and ideal for gamers.

We recommend reading our post about AMD Ryzen 5 2600X vs Core i7 8700K in games and applications

Processor	Cores / threads	Base frequency (GHz)	Turbo frequency (GHz)	Cache L3 (MB)	L2 cache (MB)	Memory	TDP (W)
AMD Ryzen 5 2600X	6/12	3.6	4.1	16	3	DDR4-2933 dual-channel	65
AMD Ryzen 5 2600	6/12	3.4	3.8	16	3	DDR4-2933 dual-channel	65

3rd generation AMD Ryzen

The third generation AMD Ryzen processors will arrive next year 2019 if everything goes as planned. For now little is known about them, apart from the fact that they will use the 7 nm manufacturing process of Global Foundries and that they will be based on the Zen 2 architecture.

Zen 2 is rumored to make the leap to six- or eight-core CCX complexes, making it possible to manufacture single die processors with a maximum of 16 or 12 cores. Zen 2 is also expected to result in an improvement in the CPI of the processors , the main objective of AMD would be to reduce the latencies of communication between the internal elements of the processor, something that will be especially beneficial in video games.

AMD Ryzen 5 2400G and Ryzen 3 2200G, the union of Zen and Vega graphics

Without a doubt, the AMD Raven Ridge APUs have been one of the most interesting launches of the company for this year 2018. It is the eighth generation of APUs of the company, and the most important to date to include inside the architecture Zen. Previous AMD Bristol Ridge APUs were based on the Excavator architecture, the latest evolution of the Bulldozer that was unable to compete in performance with Intel processors. The move to Zen cores means that Raven Ridge offers you a very powerful processor, and capable of accompanying a mid-range graphics card without problems, something that in previous generations of APUs was not possible.

These processors are based on a design made up of a complex CCX, which means that they both offer four physical cores. The difference is that the Ryzen 5 2400G has SMT technology, while the Ryzen 3 2200G lacks it. AMD has streamlined some of the CCX parts to reduce manufacturing costs, so they offer just 4MB of L3 cache and only 8 PCI Express lanes. This cut in the PCI Express lanes limits the bandwidth of the graphics cards, although with mid-range models such as the Radeon RX 580 or the GeForce GTX 1060 there should be no performance problem.

Another drawback of Raven Ridge is that the IHS is not soldered to the die of the processor, but instead uses thermal paste to make the joint. This lowers the manufacturing cost, but has the consequence that the heat dissipates worse, so the processors tend to heat up more than the soldiers.

We recommend reading our post on Comparison AMD Ryzen 5 2400G and Ryzen 3 2200G vs Coffee Lake + GT1030

Processor	Cores / threads	Base / turbo frequency	L2 cache	L3 cache	Graphic core	Shaders	Graphics Frequency	TDP	RAM
Ryzen 5 2400G	4/8	3.6 / 3.9 GHz	2 MB	4 MB	Vega 11	768	1250 MHz	65W	DDR4 2667
Ryzen 3 2200G	4/4	3.5 / 3.7 GHz	2 MB	4MB	Vega 8	512	1100 MHz	65W	DDR4 2667

The CCX is accompanied by a graphics core based on the Vega architecture, AMD's latest graphic design . The AMD Ryzen 3 2200G has a graphic core consisting of 8 Compute Units, that is, 512 Stream Processors that operate at a maximum frequency of 1100 MHz. As for the Ryzen 5 2400G, it has 11 Compute Units, which translate to 720 Stream Processor s at a clock frequency of 1250 MHz.

AMD has included in these processors its most advanced memory controller, capable of offering native support for DDR4 at 2933 Mhz in dual channel configuration. The integrated graphics are very sensitive to the speed of memory so the faster it works the better the games will go.

These two processors have been very competent in current video games , although you will have to settle for 720p resolution in the most demanding if you want to enjoy a good experience. The dependence on DDR4 memory partially limits its performance in video games, the revolution will come when AMD decides to include a dedicated memory in this type of processors, although this would have the drawback of raising its price significantly.

This ends our interesting post about AMD Ryzen, remember that you can share it with your friends on social networks, with this you help us spread it so that it can help more users who need it. Need help? You can go to our hardware forum with free registration and we will be delighted to help you.