Topic: ai model optimization
-
Compact AI Models Drive Life-Saving Advances
The RxScanner, an AI-powered handheld spectrometer, initially failed in South Africa due to slow data transfer to a server in the US, prompting the developer to create a smaller, offline version that runs on an Android phone. "Small AI" models, which run on low-power devices without needing const...
Read More » -
Apple pursues AI chip deals to boost its servers
Apple's internal AI chips, including the M2 Ultra, are insufficient for the computational demands of its new Siri features, forcing it to rely on Nvidia chips in Google's cloud. The company's own server chip, Baltra, has been delayed, with a rival to Nvidia not expected until 2029, prompting Appl...
Read More » -
Google’s Gemma 4 optimized for Pixel 10’s TPU
Google launched Gemma 4 E2B, an AI model fine-tuned for Tensor Processing Units (TPUs) that runs directly on the Pixel 10’s custom TPU for faster, more power-efficient on-device inference. This on-device execution reduces reliance on cloud servers, improving latency and privacy while enabling tas...
Read More » -
Satellites on Fire Raises $2.7M for Climate Tech
A climate tech startup founded by Argentine teenagers has secured $2.7 million in seed funding to expand its AI-powered wildfire detection platform into the United States market. The company's software platform aggregates data from multiple satellites and uses AI to analyze heat signatures, detec...
Read More »