Timeline
Booked Intel to package 3 million TPUs
Google booked Intel to package 3 million TPUs by 2028
Google booked Intel to package 3M TPUs by 2028
Google is developing Frozen v2 chip, targeting deployment as early as 2028.
Google booked Intel to package 3 million TPUs by 2028
Committed $75B capex to AI infrastructure
ByteDance released SwanTale, a unified speech-audio model, on Hugging Face
ByteDance launched Seedance 2.5 model
ByteDance Seed releases SpectraReward, a training-free reward model for text-to-image generation.
Released iLLaDA, an 8B diffusion language model trained on 12T tokens.
Ecosystem
ByteDance
Evidence (5 articles)
AgiBot WITA-Omni Scores 85.21 on DailyOmni, Beats Gemini
Jul 30, 2026ByteDance iLLaDA: 8B Diffusion LM Matches Qwen2.5 Base, Lags on Instruct
Jun 27, 2026MiniMax H3 Video Model Beats Seedance 2.0, Opens Weights
Jul 31, 2026Time's First AI A-List: Alibaba, ByteDance, Zhipu AI Make Cut
Apr 29, 2026ByteDance Builds In-House AI CPUs for TikTok-Scale Agent Inference
May 31, 2026