Viewing archive Sun, Jul 19

Archive

Chips tomshardware.com

NVIDIA RTX 50 Super GPUs are reportedly ready, but stuck in limbo due to excessive GDDR7 pricing — 3GB GDDR7 module costs triple the price of 2GB

The 3GB GDDR7 chips that the RTX 50 Super GPUs will use reportedly cost twice to thrice as much as the 2GB chips found on vanilla RTX 50-series graphics cards. This would likely push the retail price of these GPUs way beyond Nvidia's target MSRP.

Chips cxodigitalpulse.com

General Compute Secures Up to $400 Million Debt to Scale the World’s Fastest Inference Neocloud

Built on specialized SambaNova chips, the company’s air-cooled infrastructure deploys in weeks and runs AI inference up to 16x faster than standard GPU clouds SAN FRANCISCO, CA — General Compute, the world’s fastest inference neocloud, today announced a committed debt facility of up to $400 million from Upper90...

Big Tech cxodigitalpulse.com

Intel and Google Cloud Announce Collaboration to Accelerate Intel’s AI-Enabled Enterprise Transformation

SANTA CLARA, Calif., and SUNNYVALE, Calif – Intel and Google Cloud today announced an expansion of their previously announced multi-year strategic collaboration, which would accelerate Intel’s enterprise-wide digital evolution through the deployment of Gemini Enterprise and Google Cloud.

Models techcrunch.com

No product? No problem. This Disrupt 2026 session shows how to get pre-seed funding with conviction, storytelling

It’s not just you: AI startups are taking in a huge amount of seed funding, and in the process making things harder for anyone looking for funding even at a pre-seed stage. We’ve covered the trend in detail, and at this year’s TechCrunch Disrupt event, we want to help pre-seed founders now being held to seed-stage […]

Quick takes

Frontier labs are packaging free educator Claude and Gemini tools as classroom distribution across U.S. K-12 and Indian robotics programs.
Models
Workplace AI is shifting from chat to multi-hour agents that act across apps and files while buyers measure cost per successful task.
Models
Voice interfaces are productizing, with new ChatGPT voice models shipping and enterprise agents already handling high-volume customer conversations.
Models
Open embedding models are being framed as infrastructure for enterprise RAG, agent memory, code search, and agentic retrieval.
Models
Agent builders are elevating skills, sandboxed hosting, evaluation, and open or synthetic data as the stack beyond model weights.
Research
Inference economics are attracting capital, from a four-hundred-million-dollar chip-backed loan to Fireworks' seventeen-point-five-billion valuation on cheaper model serving.
Models
Frontier labs are packaging free, standards-aligned educator tools for K-12 and robotics classrooms, treating schools as a durable distribution surface for premium model access.
AI
Chat products are shifting from conversational assistants toward multi-hour work agents that act across apps and files, with enterprises already deploying voice and chat agents at operational scale.
Models

Market Pulse

AI Pulse
44/100
bearish

AI-linked equities are under pressure, with Meta Platforms -2.79%, Taiwan Semiconductor Manufacturing -2.77%, NVIDIA -2.21% driving the tracked basket lower.

META -2.79%
Meta Platforms labs
TSM -2.77%
Taiwan Semiconductor Manufacturing chips
NVDA -2.21%
NVIDIA chips
GOOGL -2.17%
Alphabet. labs

Recurring Movers

META 12 hits · -2.79%
TSM 12 hits · -2.77%
NVDA 12 hits · -2.21%
GOOGL 12 hits · -2.17%