Subsequent gen AMD laptops are going to be bringing an upgraded GPU structure
Up to date: Mar 21, 2024 2:27 pm
WePC is reader-supported. If you purchase via hyperlinks on our website, we could earn an affiliate fee. Costs topic to alter. Learn more
AMD just isn’t fairly but saying something about its next-generation f graphics playing cards, but it surely has confirmed that RDNA 3+ GPU architecture is coming to Strix Point APUs in 2024. It can additionally deploy the next-gen neural processing as it’ll characteristic the XDNA 2 NPU on board as properly.
Not fairly the subsequent main launch of RDNA 4, the replace brings optimization to what’s already on provide with the likes of the best AMD GPUs within the RX 7000 collection. Strix Level appears to be aiming to enhance upon what the current 8000G processors launch offered as that introduced the mainstream graphics to onboard graphics.
Strix Level APU data
AMD has not clarified what the structure has to supply in being built-in into the APUs, however that RNDA 3.5 will probably be built-in into them. They do nonetheless counsel that the Ryzen 8060 APUs will provide a 3x enchancment in generative AI capabilities with the most recent XDNA 2 NPUs and fuller Ryzen AI software program suite.
The processors are going to be integrating the Zen 5 core structure for the processing energy behind it. Plus enhancing upon the Hawk Level number of 16 TOPs of AI energy, it might come to round 48 TOPs.
These are anticipated to be launched within the second half of 2024, as they arrive via OEM channels. On the identical time, we may very well be seeing Intel’s Arrow Lake.
Ryzen 8050 anticipated options | Strix Level Mono | Strix Level Halo |
---|---|---|
Course of Node | Zen 5 (4nm) monolithic | Zen 5 chiplet |
Cores | As much as 12 (Hybrid Zen 5 + 5C) | As much as 16 |
Cache | 32MB shared L3 | 64MB shared L3 |
Compute Models | 16 RDNA 3+ | 40 RDNA 3+ |
Reminiscence controller | 128-bit LPDDR5X | 256-bit LPDDR5X |
AI engine | XDNA 2 ~25 TOPs | XDNA 2 ~50 TOPs |
Anticipated efficiency | On par RTX 3050 Max-Q | On par RTX 4070 Max-Q (90W) |
Discussion about this post