Just six months ago, AMD (and Intel) was touting its “AI PC” laptop processors with the company’s second-gen neural processing unit for low-power AI acceleration — and already last year. Now it’s announcing a new mobile CPU with a theoretically better third-generation NPU (XDNA 2, up to 50 TOPS), an updated CPU using the new generation Zen 5 process and an upgraded integrated GPU (RDNA 3.5). Launching the rebranded Ryzen AI, AMD’s Copilot Plus-ready 300 series of parts.
Formerly code-named Strix Point, the Ryzen AI 300 series is the successor to the Ryzen 8000 generation; Because the older chips don’t have the NPU bandwidth of 300 — that’s too limited — they’ll keep their current name.
The laptop launch partners for the new processor are the same Copilot Plus PCs like Nvidia had its RTX AI for PCs back in the day, mainly from Asus.
What’s in the name?
Arriving in laptops starting in July, Ryzen AI chips keep the same essential naming conventions, with AI shoes worn and modifiers like HX now used to indicate market tier rather than power-consumption class.
AMD is launching with two flagship parts, the Ryzen AI 9 HX 370 and the Ryzen AI 9 365, which differ by the number of CPU and GPU cores.
The new Zen 5 CPUs actually appear to be split into 4nm Zen 5 cores and 3nm Zen 5c cores; AMD won’t provide any details on the architecture, so it’s not clear why the CPU mixes the two, though that’s not unprecedented for AMD (it also mixes on its graphics cards).
It could be a way to match the performance/efficiency cores of Intel and Apple’s chips for additional power savings or to manage thermals. Since the new chips are intended to serve a wide range of laptop types, with power draws ranging from 15 watts to 54 watts (small ultralight to large thin and light), they can be configured to give manufacturers more granularity in specification. Overall power draw of the system. As you’d expect, there are constraints on the number of cores, frequencies and cache sizes, which can result in at least modest generation-over-generation performance improvements.
Ryzen AI 300 Series Chip Specs
| Ryzen AI 9 HX 370 | Ryzen AI 9 365 | |
| Zen 5c CPU cores (3nm) | 8 | 6 |
| Zen 5 CPU cores (4nm) | 4 | 4 |
| Total threads | 24 | 20 |
| Single core burst frequency | 5.1 | 5 |
| CPU cache | 36MB | 34MB |
| iGPU | Radeon 890M | Radeon 880M |
| GPU compute units | 16 | 12 |
| NPU | XDNA 2 up to 50 TOPS | XDNA 2 up to 50 TOPS |
| Power envelope | 15-54 W | 15-54 W |
The Radeon 890M and 880M integrated GPUs aren’t a big leap from the last generation, but the 890M represents a new maximum for the number of compute units. The previous top-of-the-class was the 780M with 12 CPUs, and now there’s an 890M that increases the maximum number of GPU CUs to 16.
See this: See everything announced from the AMD keynote at Computex 2024
NPU leapfrogging
If you’re wondering why Qualcomm and Microsoft launched their Copilot Plus two weeks before Computex, my theory is that it would otherwise have been overshadowed by AMD (and maybe Intel?) — because, among other reductionist talking points, There are more than 50 TOPS. 45 tops. TOPS, a measure of integer math performance in NPUs, is only really important to Microsoft, as it is its primary criterion for what counts as a Copilot Plus PC. Copilot uses only integer math to perform its tasks, such as summarizing, drafting, quick and dirty background removal in video conferences, tracking everything you do, and more.
But integer math is very limited if it needs to compute with very large or very small numbers, as is required for image- and video-output generative AI. It uses floating point calculations (which specify how numbers are manipulated in computer registers). Most existing applications that implement AI tools or features — creative, gaming, and more — already use 16-bit floating point (called “single precision”) and, to a lesser extent, 32-bit FP (double precision). doing.
All that to say is that there is another aspect of consumer CPU’s NPU where Qualcomm has been lapped by AMD: Block floating point. It’s essentially a way to compress 16-bit floating point values ​​so that they can be manipulated potentially as fast as the 8-bit integers NPUs use, in theory, without losing 16-bit precision.
Ryzen 9000 series
| Ryzen 9 9950X | Ryzen 9 9900X | Ryzen 7 9700X | Ryzen 5 9600X | |
| Zen 5 CPU cores (4nm) | 16 | 12 | 8 | 6 |
| Total threads | 32 | 24 | 16 | 12 |
| Single core maximum burst frequency | 5.7 | 5.6 | 5.5 | 5.4 |
| CPU cache | 80MB | 76MB | 40MB | 38MB |
| Power | 170 w | 120 w | 65 w | 65 w |
| socket | AM5 | AM5 | AM5 | AM5 |
| is available | July 2024 | July 2024 | July 2024 | July 2024 |
AMD also introduced new Ryzen 9000 X series desktop processors using the Zen 5 architecture. They still use the AM5 socket, so you can upgrade older systems with them, but if you replace your motherboard with one of the newer models featuring the X870/X870 chipset, you’ll get USB 4.0, PCIe 5 and higher speeds for AMD’s EXPO. will get (Profiles used for its one-click overclocking).
