AI
ZAYA1-8B Proves AMD MI300 Viable for AI Training
Zyphra releases ZAYA1-8B, an efficient 8B-parameter model trained on AMD MI300 GPUs that matches larger models…
Tag: transformer architecture
NVIDIA
ZAYA1-8B Achieves GPT-5 Performance with 760M Parameters
Zyphra's ZAYA1-8B achieves GPT-5 performance with just 760M active parameters, while Subquadratic claims 1,000x efficiency gains…
AI
AI Architecture Advances: Subquadratic Claims 1,000x
Miami startup Subquadratic claims 1,000x efficiency gains through subquadratic architecture, while Google releases speculative decoding for…
Microsoft
AI Architecture Breakthroughs Drive Local Inference Revolution
Microsoft and NVIDIA have released new AI models emphasizing efficiency and local deployment capabilities. Microsoft's MAI-Image-2-Efficient…
NVIDIA
Transformer Architecture Advances Drive AI Efficiency Breakthroughs
AI architecture advances in 2024 focus on hybrid model designs, local inference capabilities, and enterprise deployment…