During AMD’s opening keynote at Computex 2024, firm CEO Dr. Lisa Su revealed AMD’s newest AI PC-focused chip lineup for the cellular market, the Ryzen AI 300 collection. Based on AMD’s new Zen 5 CPU microarchitecture, the Ryzen AI 300 collection – codenamed Strix Point – is meant to supply an across-the-board enchancment in cellular SoC efficiency, with AMD proclaiming that the Ryzen AI 300 collection will provide the quickest AI inference efficiency throughout the compact and transportable PC market.
Under the hood, the brand new cellular SoC from AMD incorporates not solely their new Zen 5 CPU structure, but in addition their new RDNA 3.5-based built-in graphics, and the third technology XDNA2-based NPU, the latter of which is rated to ship 50 TOPS of efficiency for AI-based workloads. As a end result, the Ryzen AI 300 collection represents a major improve in AMD’s cellular chip lineup, with all the main elements of the platform receiving a significant improve versus their Zen 4-era Phoenix/Hawk Point SoCs. The one factor the brand new platform will not get, nonetheless, is a course of node enchancment; AMD is constructing Strix Point on a 4nm node, similar to Phoenix/Hawk Point earlier than it.
For this morning’s announcement, AMD has unveiled the primary two Ryzen AI 300 SKUs designed for notebooks. The first of those is the Ryzen AI 9 HX 370, which options 12 Zen 5 cores with a most increase frequency of as much as 5.1 GHz, and comes geared up with 36 MB cache (12 MB L2 + 24 MB L3). The different chip to be introduced is the Ryzen AI 9 365, which has two fewer Zen 5 cores (10 cores) and operates with a 5,zero GHz increase frequency and a 10 MB L2 + 24 MB L3 cache allocation.
AMD Ryzen AI 300 Series: Zen 5 CPU, RDNA 3.5 GPU & New NPU
As AMD gears up its cellular portfolio for the latter a part of 2024, it is clear that the corporate is seeking to align itself with Microsoft’s Copilot+ AI PC program, and the efficiency necessities that include it. In explicit, Copilot+ requires an NPU with not less than 40 TOPS of efficiency, of which there are none in the marketplace at this second, and the primary will not arrive till Qualcomm’s Snapdragon X gadgets hit retailer cabinets later this month. Meanwhile, AMD has aimed a bit past this with its third technology, XDNA2-based NPU, which is rated to attain as much as 50 TOPS of INT8 efficiency.
With the brand new concentrate on AI efficiency additionally comes one other new cellular processor naming scheme from AMD, changing the earlier system AMD introduced two years in the past. The acquainted Ryzen branding is retained, in fact, however AMD is now figuring out it is NPU-equipped chips with the “AI” moniker as properly. And maybe most significantly, the numbering system has been reset. AMD is now beginning at 300 slightly than what would have been the Ryzen Mobile 9000 collection.
The Ryzen AI 300 naming system is comparatively straightforward to decipher, with the Ryzen AI half denoting it has an NPU included throughout the silicon. In distinction, the 9/HX a part of the naming notes the model degree or the stature of the place the half sits within the product stack. The final half, the 370 half, is extra attention-grabbing, as the three denotes the collection, with the three that means from AMD’s branding that it has the third technology of NPU, whereas the final two numbers point out the half quantity and SKU.
Of course, we would be remiss to not level out that this comes half a 12 months after Intel reset its personal cellular product naming system, giving us the Core Ultra 100 collection. And with what’s presumably the Core Ultra 200 collection on the way in which primarily based on Intel’s forthcoming Lunar Lake cellular CPU, AMD appears to have achieved the one wise factor and restarted their cellular processor numbering system in order that it is much like, however greater than, Intel’s numbering system.
| AMD Ryzen AI 300 Series Mobile Processors (Zen 5/Strix Point) |
