NVIDIA RTX Spark: 128GB Unified Memory for On-Device AI
NVIDIA RTX Spark brings ~1 PFLOP FP4 and up to 128GB unified memory to laptops and desktops — enough to run frontier-sized open models on-device. Shipping fall 2026.
NVIDIA RTX Spark brings ~1 PFLOP FP4 and up to 128GB unified memory to laptops and desktops — enough to run frontier-sized open models on-device. Shipping fall 2026.
Microsoft's Discovery AI designed a quantum chip material 1,000x more reliable than previous qubits — the company now targets a scalable quantum computer by 2029, a halved timeline.
Arm doubles guidance for its AGI CPU data center chip to $2B in FY2027-FY2028 combined, positioning data center as its largest business segment.
OpenAI moves from foundation models to silicon with a MediaTek/Qualcomm/Luxshare chip partnership targeting 2028 mass production.
Kuo: OpenAI is co-designing custom smartphone chips with MediaTek, Qualcomm, and Luxshare — targeting 2028 mass production as the company moves toward hardware-native AI.
Huawei confirms its Ascend 950-based supernode will fully support DeepSeek V4, completing China's domestic frontier AI hardware-software stack.
MIT's new artificial muscle fibers use ionic pumps instead of motors—silent, scalable, and strong enough to bend robotic arms and shake hands naturally.
Curated AI insights — sent when there's something worth your inbox.