6 Takeaways from AMD’s CES consumer chip announcements

AMD CEO Lisa Su started CESA keynote will discuss AMD chipset announcements. Ryzen 7000A series of mobile CPUs that range from the RX 7000 to the high-performance X3D classes of desktop GPUs. The main topics included the increasing role of AI (both in everyday computing and in traditional roles) as well as the impact (and requirements) of hybrid work.

Su has promoted some commercial aspects of her partner’s products, including detailed discussions about robotics (and Magic Leap’s AR for healthcare) Certificate 60601). An astronaut has been talking to us about research and adaptive computing in AMD. He also demonstrated simulations of (wo?) manned missions to Moon. AMD isn’t limited to your console. It’s also available in space.

She concluded her speech with a discussion about sustainability in its components, technologies, and how this translates into large-scale computing and high-performance computing. AMD also launched Instinct MI300, a combination GPU/CPU that is designed for data center-level AI training and services. It will be available in second half of the year.

You should note that some of this information is derived from other briefing materials as well as the keyword. More CES news can be found on our website. live blogplus more details about partner commercials.

On-Chip Adaptive AI is AMD’s most interesting mobile development

AMD has revealed its XDNA architecture for accelerating AI inference, which can scale up or down depending on how many CPUs a hardware developer throws at it. (It comes from the company’s acquisition of Xilinx). Hence the trademark “Adaptive AI”. Because it can pass data from array to array without having to communicate with external memory, the CPU, or the GPU, AMD says it reduces latency and makes it a flexible architecture.

All of this is a prequel to the debut of the company’s new generation of HS mobile CPUs, which include a small, integrated AI accelerator that can handle up to four concurrent AI streams, either for separate tasks or to perform a single task using parallel threads. It’s optimized for energy efficiency, and because it frees up CPU and GPU cycles that would otherwise take to process the AI, AMD claims it results in less performance drop and better battery life.

AI algorithms are pervasive in computing these days, and mobile devices are increasingly relying on them for normal operations, such as cleaning video during video conferencing and optimizing battery-saving behavior. AMD claims it is up to 50% more efficient than the Apple M2 Neural Engine. Intel uses a separate chip to accelerate its AI, dubbed Movidius, which means its presence in a laptop will depend on individual manufacturers.

Su also previewed the Alveo V70 accelerator, which is a small, low-power (75W) AI inference acceleration card. Pre-orders will start now and ship in the spring.

HX is faster, but HS is more modern

This is a cleavage that I’m not really happy with. Both Intel and AMD use “HX” to designate CPUs intended for high-performance gaming or creativity-focused laptops, which tend to be used by people who are looking for the latest technology (and who are usually willing to pay for it). But the aforementioned Adaptive AI Accelerator is only found in the Ryzen 9 7940HS, Ryzen 7 7840HS, and Ryzen 5 7640HS, which are CPUs designed for thin but powerful models, and they also have the advantages of using a 4nm manufacturing process, while the HX’s are still on 5nm. (This is partly how they can safely reach higher clock speeds.)

A smaller process tends to be more energy efficient and faster. HS needs that to hit the 35-45W power target for that class of laptops, but given the amount of AI that goes into creative apps and games, it was nice to be able to get some extra power on the smaller die. .

HS chips also use the newer RDNA3 architecture for their integrated graphics. It is not an uncommon practice. The rationale is that the HX class of laptops will use discrete GPUs. But most laptops with discrete GPUs still rely on the integrated version in order to reduce power consumption, for example, when the speed of the discrete GPU is not necessary.

On the other hand, the HX chips are the first AMD mobile processors to use Chiplets. And the top Ryzen 9 has 16 cores / 32 threads.

Gaming laptops It was announced at the show that takes advantage of all that AMD has to offer Alienware M16 and M18And the Asus TUF Gaming A16 Advantage Edition and the new Lenovo Legion Pro series. The first Ryzen 7040 series laptops will begin shipping in March. HP CEO Enrique Lloris came on stage to promote the new Dragonfly Pro laptops.

There is not enough support for USB 4

It’s bad enough that it’s another year without Thunderbolt support on AMD laptops, but even USB 4 isn’t supported on the lower end of the CPU spectrum where “mainstream thin and light” chips live. And AMD’s information seems to indicate that even the HX models may not have it because a USB 4 controller may be optional.

AMD intentionally underestimates 65W desktop CPUs

And that’s a good thing. The 65W class of desktop processors are for low-power, compact, or inexpensive computers, and they tend to be popular with developers. For this generation, AMD has designed the processors to run at 65W but is able to pull nearly twice that through overclocking, which the company claims can deliver a 30% performance increase. Since so much of your daily computer use is for mundane tasks — at least if you’re looking at this class of processor — defaulting to a low-power state makes sense.

Ryzen 9 joins the X3D generation

AMD’s intergenerational workforce uses the company’s proprietary 3D V-Cache architecture to pack more cache on the die to deliver better performance from essentially the same CPU. This year, AMD is adding the Ryzen 9 7950X3D to the mix — there are also a couple of Ryzen 7s as before, the 7900X3D and 7800X3D. AMD will be compiled Star Wars Jedi: Survivora highly anticipated game, with wizards.

Mobile discrete GPUs don’t sound very exciting

Maybe it’s just me, but nothing really seems to be noticeable about the RX 7600M, RX 7600M XT, RX 7700S, and RX 7600S. It seems like a good step up from last year’s offerings, but it doesn’t look nearly as worthy of the HX-class as the RTX 4090 or RTX 4080. And there are plenty of new technologies aimed at boosting the performance of AMD GPUs in conjunction with CPUs, such as the SmartShift Radeon Super upgrade. Resolution, FidelityFX Super Resolution 3 (better class of upgrade) and Hypr-RX mode (maximize frame rate with one click and reduce latency using all available settings) aren’t quite here yet. In the first half of this year, we were told at some point.