AMD also has unique and exciting technologies that enable their products to perform better and sell better. In this resource, you will learn about the main technologies of AMD, which can help you choose products from this company.
This article will not cover processors (both CPUs and GPUs) that were released earlier than 2010. Also, we will not consider server processors and special equipment, since very highly specialized technologies can be used there, which ordinary users do not care about. All AMD technologies, but almost without a technical description, you can read on their website.
AMD Turbo Core
This technology allows multiple GPUs (up to 4) to be combined for parallel 3D operation, resulting in high performance.
To build such a system you need:
- video cards with CrossFireX support
- special bridges for connecting video cards (you can do without them, but this option is the most optimal)
- powerful power supply, because a lot of energy will go
- motherboard with connectors for multiple video cards and CrossfireX support.
You can choose cards from the same series (there are exceptions), but different models. You can read about the series of AMD video cards and processors by following the links. In this case, the system operates at the speed and frequency of the weaker chip. Video memory is not summed up and also takes the smallest value. You can see about a similar technology from NVIDIA SLI here.
This system differs in that it serves to facilitate the graphics processor. In this case, one graphics card will handle physics and the other graphics.
This is such an asymmetric CrossFireX, in which one card is the graphics core in the APU processor (hybrid), and the other card is another video adapter in the system. Thus, it is possible to increase system performance even if the cards have different capacities.
But this barrel of honey also has a fly in the ointment. This system does not work on OpenGL and DirectX up to version 10. Also, a prerequisite is the superiority of one in performance no more than 2 times. Also, the performance of this system depends on the optimization of the application.
This feature automatically switches the running graphics based on the current task. For weak tasks, you can save battery power and use the graphics built into the processor. In demanding applications, more powerful discrete graphics are included. This allows you to suffer no less when choosing a video card in a laptop.
In case of inactivity, the video card or video core may consume almost no power. Also, this technology is responsible for disabling the currently unnecessary video cards in the Crossfire system.
Provides direct control over energy consumption. Thus, it can set the maximum frequencies for high performance, while the card does not go beyond the TDP. In applications that squeeze all the juices out of the card, for example stability tests, the frequencies will be lower, and in games, on the contrary, higher.
This technology is similar to the previous one, but its goal is to find a balance between performance and power consumption. It controls the power and frequency of the graphics chip and changes them depending on the load on the GPU. This technology is able to completely turn off some unused blocks in idle mode.
The computer quickly wakes up from sleep mode. Because of this, its inclusion becomes less critical and more frequent.
This is a power management feature for the processor that is made to conserve power. It lowers the clock speed when the load is light. Thus, the processor consumes less power, generates less heat, and its cooling can slow down its speed. The name also tells us about it: “Cold and quiet.
This technology reduces power consumption and heat dissipation by shutting down unused processor blocks.
This is a program that allows you to monitor system performance, performance settings in real time, adjust fan speed, and conduct system stability tests. With it, you can overclock processors.
AMD TressFX Hair
This technology is designed to create realistic physics of hair and fur behavior. It takes into account a large number of different parameters for each curl: lighting, wind, gravity, collision of the strands with each other and with the character’s body.
For calculations, the GPGPU interface is used, which allows this technology to work on all video cards. But still, Radeon cards have an advantage, because they work faster in this interface than NVIDIA video cards.
AMD TressFX Hair competes with similar technology from NVIDIA NVIDIA Hairworks.
AMD Quick Stream
It is a technology for managing your traffic. It changes the bandwidth of data transmission over the Internet. For applications that are sensitive to a drop in speed (streaming video, games), a large bandwidth is given, while for other applications it is maintained at the minimum comfortable level.
It is also possible to change priorities and create different profiles.
This is an API that allows you to use the processing power of GPUs (not to be confused with a line of video cards with this name). OpenCL is used for programming. This technology can also be used for hardware acceleration of physics in games if the game engine supports OpenCL.
It is known that the GPU can handle mathematical calculations much faster than the CPU. And this technology makes it possible to channel this potential for economic calculations, weather forecasts, physics in games, scientific computing, and the like. You can read about other interesting facts about video cards in this article.
This technology is designed to eliminate image tearing and improve picture smoothness. It does this by synchronizing the graphics card (or graphics core in the APU) and the monitor. Thus, they operate at the same frequency.
This technology requires support from both the hardware in the computer and the monitor to work. And this is a significant drawback, because in the absence of a good monitor with support for this technology, you will not be able to try it out.
Low Frame Rate Compensation (LFC)
This technology became part of FreeSync with the arrival of a new driver version. It constantly monitors the rendering speed of the video card and, if the rendering frequency drops below the minimum frequency of the monitor with FreeSync technology, it inserts additional frames. This is controlled by a special algorithm. Also, this technology can be paired with vertical sync (VSync on) when the FPS is below the minimum display frequency. And before, when there was no LFC, FreeSync in this case did not work at all, which caused breaks. In this case, it is better to turn on VSync, because with it there will be less gaps.
This is a purely software technology, as evidenced by its access in drivers.
Designed to display images on multiple monitors from one computer. It is worth noting that this technology does not require you to have identical monitors.
This is a low-level API from AMD that aims to improve gaming performance through better optimization. In particular, the load on the processor is reduced. Indeed, in games that support this technology, the performance with it is higher than with DirectX. Mantle is designed to compete with OpenGL and DirectX, but so far only works with graphics cards of the GCN architecture. You can read more about the main innovations here.
This technology is designed to connect an external video card to your laptop to enjoy good graphics, because, as a rule, desktop video cards are more powerful than built-in discrete ones. To do this, you need a graphics card certified for this technology and a suitable laptop.
Supported graphics cards:
- Radeon R9 Fury
- Radeon R9 Nano
- Radeon R9 300 family
- Radeon R9 290X
- Radeon R9 290
- Radeon R9 280
- mobile versions of upstream GPUs
- future graphics cards based on Polaris architecture
- Radeon Software Crimson Edition 16.2.2 or 16.3
- 1x Thunderbolt 3
- Thunderbolt 3 cable
- Windows 10 Build 10586 or later
- BIOS with eGFX support
- Thunderbolt Firmware v16
- Pass Thunderbolt Certification
Virtual Super Resolution is a downsampling technology built into the driver itself. The picture is rendered at a higher resolution than the screen resolution and then downsampled. The increase in quality is due to the increase in sampling points in higher resolution.
The feature is activated in Catalyst drivers. It does not work on all cards:
- R9 285 and up
- R9 380 and up
- R9 Fury / Nano
It is an API that allows you to enjoy 3D images on your PC. This requires a video card or APU that supports this technology and a special certified monitor. Without a special monitor, you can only use the anaglyph method (red-blue image with appropriate glasses).
This technology allows designers to create more expansive and accurate acoustic scenes. This is achieved through a dedicated graphics processor (DSP) to create the audio environment. Also, this technology allows you to create new sound effects, expands the number of sound objects and channels. AMD TrueAudio is also responsible for creating surround sound in headphones for gamers, which improves player positioning.
A technology built to develop VR on AMD hardware and enhance the comfort of VR applications.
It includes plug-and-play compatibility with headsets, reducing delays in the transfer of data about head movement to the GPU and thereby increasing smoothness of rotation, support for two GPUs for rendering separately for each eye, technology for direct connection of a VR headset to AMD Radeon graphics cards …
These are proprietary memory extensions. Thus, memory sticks certified for this technology will have different work profiles that will serve different purposes. These can be both overclocking profiles for enthusiasts and energy-saving profiles.
AMD was the first to introduce the new video memory standard in its products. These were video cards based on the Fiji chip, which can be found here. This memory is like a sandwich. In it, the blocks lie on top of each other.
The advantages of this memory include:
- throughput that is significantly higher than that of GDDR5
- takes up little space and can be placed on the substrate together with the chip, thereby increasing the speed of data exchange between them
The main disadvantage of the first version of this memory is the low volume of 4 GB, but this problem will be solved in the new version. The features of the new HBM2 were told at the presentation of the Vega architecture at CES 2017.
Many other technologies from AMD were not mentioned here due to their insignificance or specificity. If you find a mistake, first see the instructions in the Introduction section. If we did not take into account something, then write in the comments. We will definitely answer you. Also, if you want to know about some other technology from AMD, then write in the comments and we will soon add it to the list if it seems interesting to us.