The Open-Source AI Stack
RSS
All hardware

Hardware · AI PC

Qualcomm Snapdragon X Elite

Qualcomm

AI-PC class part: a capable NPU for small on-device models, but the 135 GB/s memory bus caps decode speed for anything larger.

CPU · GPU · NPU 135 GB/s memory bus 64 GB Unified memory
AI PC. Bus width tracks bandwidth (135 GB/s, sets decode speed); the box tracks capacity (64 GB, sets what fits). Unified memory is one pool shared by CPU and accelerator.

Specs

Memory
64 GB LPDDR5X
Bandwidth
135 GB/s
NPU
45 INT8 TOPS
Power
80 W
Form factor
soc
Interconnect
none
Released
2024-06

What it runs (single unit, Q4_K_M, 4K context)

Model Fits? Decode ceiling
Llama 3.1 8B Instruct yes ~27 tok/s
Llama 3.3 70B Instruct yes ~3.3 tok/s
Qwen 2.5 72B Instruct yes ~3.2 tok/s
DeepSeek-V3 no —

Ceiling is the theoretical rooflineruntimeA performance model that bounds throughput by either compute or memory bandwidth, whichever is the limiting resource for an operation's arithmetic intensity. Open full entry ; open the explorer to set quant, context, and runtime and see the realistic range.

Sources