9 results on this page
Samsung reportedly increased prices for new orders using its 4nm, 5nm, and 8nm foundry processes in July. Price increases reportedly reach 15% for customers in China as AI demand fills available 4nm capacity.
Samsung Foundry is reportedly raising prices for multiple process nodes as customer orders increase. In July, SF4 4nm wafer prices rose 10%–15% for US and Chinese customers, while Taiwan customers saw increases of about 5%–10%; SF5 5nm prices also increased 10%–15%.
A court reportedly heard that former Samsung engineers stole process recipes for the company’s 18nm-class DRAM technology for sale to CXMT. One former Samsung engineer who moved to the Chinese memory maker is now reportedly in custody.
Samsung’s semiconductor manufacturing roadmap spans four campuses across South Korea and the United States, including Pyeongtaek, Hwaseong, Giheung, and Taylor. The analysis examines how fab execution and yield problems could affect major customer commitments, including a reported $16.5 billion Tesla deal.
OpenAI’s Private Safety Processing is designed to detect abuse across multiple interactions without requiring customers to hand over their data. The approach preserves Zero Data Retention and contrasts with Anthropic’s stated 30-day retention requirement.
AWS presents three serverless patterns for invoking Amazon Bedrock AgentCore agents asynchronously from AWS Step Functions. Task-token callbacks, direct service integrations, and durable functions help eliminate idle compute costs while agents process requests.
OpenAI reaffirmed Zero Data Retention for eligible API customers using frontier models. It also previewed Private Safety Processing, designed to support advanced AI safety work without compromising customer data privacy.
OpenAI said it will enhance safety processes for paying users who access its most advanced AI models. The move responds to customers using AI for increasingly complex tasks and handling more sensitive information.
Cerebras launched the CS-4, its first system to combine three wafer-scale processors in a single rack. The inference-focused machine is shipping this quarter, with Cerebras claiming up to 30 times the speed of GPU-based systems for frontier models.