Section
Infrastructure
The latest Infrastructure news, analysed daily by the i10X newsroom.
FilterGemini
-
Live Avatars: Real-Time AI Faces and Infrastructure Demands
Live Avatars move from async video to live streaming endpoints via Gemini 3.8 and 14B diffusion models. Discover the latency, concurrency, and TCO challenges for enterprise deployment.
By
-
Google Accelerates Gemini 4 Release Amid AI Competition
Google plans to launch Gemini 4 in late 2024 ahead of schedule, fueled by strong early results and…
-
Real-Time Voice AI Stress-Tests Mobile Edge Computing
OpenAI Advanced Voice Mode and Google Gemini Live are pushing mobile networks and edge infrastructure to their limits.…
-
Gemini Nano 4: Hardware-Gated AI Arrives on Android
Google mandates Gemini Nano 3 for core AI features while rolling out Nano 4 on Pixel 11. Discover…
-
Gemini 3.5 Transcribe: Enterprise Speech-to-Text from Google
Explore Gemini 3.5 Transcribe, Google’s new speech-to-text model with 2.6% WER, streaming support, and Vertex AI security for…
More stories
-
Google Gemini: Tiered Models for Cloud & Edge AI
Discover how Google’s Gemini 1.5 models optimize inference across cloud, enterprise, and edge devices. Reduce costs and latency with Pro, Flash, and Nano. Explore the strategic…
-
Gemini Omni Now Available to All Google AI Paid Subscribers
Google has rolled out Gemini Omni, its natively multimodal AI model, to Plus, Pro, and Ultra subscribers. Explore the infrastructure implications, latency benefits, and what it…
-
Google I/O 2026: Gemini’s OS-Level Pivot in Android 17
Explore Google I/O 2026’s strategic shift as Gemini integrates at the OS level in Android 17 and XR. Learn about edge AI, infrastructure impacts, and stakeholder…
-
Google Gemini Shifts to On-Device Edge AI and XR Glasses
At Google I/O 2026, Gemini moves to on-device models in Android and XR smart glasses for continuous perception. This cuts latency and server costs while changing…
-
Google Quietly Adds Gemini Nano to Desktop Chrome
Google is silently installing the Gemini Nano AI model in desktop Chrome browsers, enabling local AI via Prompt API. Explore impacts on enterprises, developers, users, bandwidth,…
-
Google Gemini Omni: Real-Time Video AI Explained
Google’s Gemini Omni is a native video AI model enabling real-time understanding and generation. Bypass fragmented pipelines for continuous streaming I/O. Discover impacts on developers, infrastructure,…
-
Google Gemini 3 Deep: AI Model Launch and Infrastructure Funding
Explore Google’s Gemini 3 Deep rollout, a multimodal AI model advancing reasoning and tool use, timed with Alphabet’s record bond sale to fuel AI data centers…
-
Gemini vs OpenAI: Google’s Infrastructure Strategy
Explore how Google’s Gemini challenges OpenAI through custom TPUs, ecosystem integration, and cost efficiency. Dive into the AI infrastructure battle shaping the future for developers and…
-
Gemini 3 Pro Limits: Navigating Access Quotas
Discover the fragmented rate limits and quotas for Gemini 3 Pro across APIs, consumer apps, and free tiers. Learn why these restrictions exist, their impact on…
-
Gemini 3 Tops FrontierMath: AI Math Record & Costs
Google’s Gemini 3 achieves 40%+ accuracy on the FrontierMath benchmark, surpassing GPT-5 in AI mathematical reasoning. Explore the trade-offs of its compute-intensive Deep Think mode and…
-
Google’s Private AI Compute: Secure Enterprise AI
Explore Google’s Private AI Compute, a confidential computing solution for Gemini models that processes sensitive data securely in the cloud using TEEs and remote attestation—keeping it…