Did you know, AMD’s latest chip development effort named “Strix Halo” has been confirmed and will be branded as the Ryzen AI Max 300 series. This is a Chiplet-based mobile processor with one or two “Zen 5” CCDs — similar to those found in “Granite Ridge” Desktop PC processors — a chip that combines a large SoC die with a powerful integrated GPU (iGPU). This combination allows AMD to equip the processor with up to 16 full-sized “Zen 5” CPU cores, along with an iGPU that can have up to 40 RDNA 3.5 compute units (2,560 stream processors), and a 256-bit LPDDR5/x memory interface for UMA.
“Strix Halo” is designed for ultra-portable gaming laptops or mobile workstations, where small PCB size is critical and a discrete GPU is not a viable option. For enthusiast gaming laptops with discrete GPUs, AMD is developing the “Fire Range” processor, which is essentially the mobile BGA version of “Granite Ridge” and the successor to the Ryzen 7045 “Dragon Range” line. The Ryzen AI Max series features three models based on the number of CPU cores and iGPU compute units — Ryzen AI Max 390+ (16 cores/32 threads with 40 CU), Ryzen AI Max 390 (12 cores/24 threads with 40 CU), and Ryzen AI Max 385 (8 cores/16 threads, 32 CU). A test model with the technical name “Ryzen AI Max 390” has appeared on the Geekbench AI online benchmark database.
The entry in the online database for this Geekbench AI benchmark submission mentions a processor self-identified as “AMD Eng Sample: 100-000001421-50_Y,” corresponding to the core and thread count of the Ryzen AI Max 390 (12 cores/24 threads, 40 CU). The processor has a base CPU clock speed of 3.20 GHz and a Max Boost clock of up to 5.00 GHz, at least for this test model (retail processors may vary). This processor is powering a test workstation sample of the HP ZBook Ultra 14 G1a (Mobile Workstation) and is paired with 64 GB of memory.
The processor recorded a Geekbench AI single-precision score of 4733, a half-precision score of 4944, and a quantized score of 13944. HotHardware notes that this is a significant 60% difference compared to the Ryzen 9 9900X Desktop PC processor. There could be many reasons behind this figure. The screenshot shows the test workstation laptop running in “Power: Balanced” mode, and the test used 256-bit AVX2 SIMD instructions (not the latest AVX512). The “Zen 5″ cores on “Strix Halo” are ported from “Granite Ridge” and EPYC “Turin” because they share the same 8-core CCDs and feature full 512-bit FP data paths. This differs from the “Zen 5″ cores on the “Strix Point” monolithic silicon, which are limited to 256-bit FP data paths, even when executing AVX512 or VNNI instructions. Therefore, AI tests using AVX512/VNNI may yield different results. Additionally, this is an engineering sample, and AMD may intentionally throttle the performance of test models.
According to TechpowerUP
