Jensen delivers a GPU to SpaceX
Mighty tiny, and packing lots of memory, the DGX Spark is made for AI
Is Meta’s acquisition of Rivos a YANK?
How many AIPs does it need to quit Nvidia?
Imagination’s open source driver grows up
Imagination adds Vulkan 1.2 and more GPUs to its open stack
Intel NPU5 in Panther Lake
Panther Lake’s NPU 5: area-lean, FP8-savvy, and built to be everywhere
Huawei’s Ascend 950 AI processor
Second in a line AI processors.
Loongson 9A1000 GPU development
A TOPS number but no NPU.
Dimensity 9500 goes big on cores, big on AI
MediaTek leans into ‘all big cores,’ bigger GPU, and agentic on-device AI.
Nvidia Rubin CPX and disaggregated long-context inference
Massive context and the inference dichotomy.
Intel’s Battlematrix delivers Llama 8B cost‑efficiency payday
Arc Pro B60 outpaces Nvidia in Llama 8B per-dollar performance.
Arm’s CPU-first AI push gets teeth with Lumex
…And a new Mali at last.
Q2’25 PC GPU shipments increased by 8.4% from last quarter, suggesting pre-tariff panic buying
Data center GPUs were up an average of 5% from last quarter.