Inference Chips Are Becoming Loan Collateral in a $400M Bet
Upper90 will lend General Compute up to $400 million secured by SambaNova inference chips, not Nvidia GPUs. Why inference silicon as loan collateral is an untested bet.
Tag
#AI Inference
7 articles
TECHi reporting and analysis covering AI Inference. 7 articles, newest first.
Fireworks raised $1.5B on a number that reframes open weights: 95%
Fireworks AI raised $1.505B at a $17.5B valuation. The number that matters is 95%: nearly all its tokens come from customer-specialized models, not raw open weights.
AMD and Cerebras Split AI Inference in Two — Read the Footnote
AMD and Cerebras will split AI inference across Helios racks and the Wafer-Scale Engine. The 5x per-watt claim is modelled against Cerebras-only hardware.
Qwen3.5 on Ironwood: What Google’s 4.7x Gain Proves
Google says Ironwood tuning made Qwen3.5 up to 4.7x faster. The result proves software co-design—not that TPUs beat Nvidia GPUs in matched tests.
AWS Puts SageMaker GPU Decisions Behind a Guided UI
AWS added a guided SageMaker UI for AI inference recommendations. TECHi explains why workload assumptions, not the button, still decide cost and speed.
Nvidia's $81.6B AI Factory Quarter: The Honest Read
Nvidia reported $81.6B revenue, $75.2B data center sales and a $91B Q2 guide. Here is the honest read on the AI factory boom.
Google's 3.2Q Token Shock: Inference Demand Goes Vertical
Google says monthly tokens across its surfaces jumped from 9.7T in May 2024 to 3.2Q+ in May 2026. The real story is inference demand.