JUST IN: Apple is relying on Nvidia after its M2 Ultra chips fell short for advanced AI workloads Apple tried running Google's Gemini models on its own servers for the revamped Siri but the chips couldn't handle it, forcing it onto Nvidia GPUs hosted in Google Cloud. Apple's next AI server chip codenamed "Baltra" has been delayed and an M7 Ultra that could rival Nvidia's Blackwell won't be ready until 2029.(Coin Bureau)