Tech News
← Home  ·  All topics

Ai Memory

8 GoKawiil briefs on this topic

Apple warns memory costs and Siri AI spending will squeeze holiday margins

Apple reported Q4 2026 gross margin of 47.9%, down from 50.1% last quarter, with product gross margin falling to 37.2% from 40.1%, despite record September-quarter revenue of $114.5 billion. CFO Kevan Parekh said that excluding tariff refunds, the sequential margin decline would have been about one percentage point, driven mainly by rising memory costs, and warned that pressure on margins will continue, with added costs expected from Siri AI investment in the holiday quarter.

China's CXMT advances HBM memory output despite reported 70% chip failure rate

China's largest DRAM manufacturer, CXMT, is reportedly set to begin limited production of advanced high-bandwidth memory chips within the next two months, despite yields suffering from a failure rate as high as 70%. The push comes amid China's broader effort to build domestic capacity for AI-critical memory chips.

SK Hynix Q3 operating profit jumps 632% on AI memory demand

SK Hynix reported third-quarter revenue of 104.62 trillion won and operating profit of 83.33 trillion won, both well above analyst estimates compiled by LSEG. Revenue grew 327.9% and operating profit 632.1% from a year earlier, driven largely by sales of high-bandwidth memory chips used in AI systems. The results follow an August announcement of a 40 trillion won share buyback and cancellation plan alongside a higher shareholder return target.

SK Hynix halts Solidigm's planned US IPO review

SK Hynix said its subsidiary Solidigm will pause its review of pre-IPO and IPO options, citing recent capital market conditions and share price trends. The chipmaker's shares initially fell 3% before paring losses to trade 0.3% lower on the news. SK Hynix said it will keep exploring other ways to fund Solidigm's planned U.S. semiconductor fabrication plant.

Nvidia hikes Shield TV Pro price 50% to $299, cites memory costs

Nvidia has raised the price of its Shield TV Pro streaming device from $199.99 to $299.99, a $100 increase despite no hardware changes. An Nvidia spokesperson told Ars Technica the hike, effective October 2, stems from rising component costs, particularly memory, across the industry. The company has also discontinued the standard Shield TV, leaving the Pro as its only Shield-branded device.

Enterprises Push AI Systems Toward Continuous Learning, Not Just Memory

A commentary argues that most enterprise AI deployments today only store and retrieve information rather than genuinely learn from experience over time. It draws a comparison to an employee who interacts with customers and attends meetings for a year but retains none of the skills or judgment gained, remaining as capable as on their first day. The piece calls for AI systems that actually improve and adapt with use, rather than simply accumulating data.

Developer releases open-source AI memory tool that stores full conversation logs, no summarization

A new project called Lossless-memory offers a long-term memory system for personal AI assistants that stores complete conversation logs rather than summarizing or embedding them for similarity search. Every record carries a timestamp, forming what the creator calls a 'Temporal Backbone,' enabling time-first queries like 'yesterday evening, about the budget.' The system has reportedly been running daily since July 2026 for a single user as the memory layer of an AI assistant.

OpenAI's ChatGPT Now Auto-Updates Memory, Offers New Controls

ChatGPT's memory feature has evolved from requiring users to explicitly say 'remember this' into an automatic system that continuously updates and revises what it retains about you, a process OpenAI calls 'dreaming.' This means the chatbot now decides on its own what details from your conversations to keep, edit, or discard, shaping how it responds to future prompts.