Industry News

Tracks company news, strategic changes, funding activities, and personnel adjustments across the blockchain and crypto industries, delivering a full-spectrum industry overview for our users.

Memory Card Prices Double in Four Months: How Long Will the Surge Last?

NAND flash memory prices have entered a rapid upward cycle, with consumer-grade storage products like microSD cards seeing significant retail price increases. For example, a SanDisk Extreme 128GB microSD card rose from $17 in October 2025 to nearly $40 by February 2026—a 130% surge in under four months. This price surge is driven by structural shifts in the NAND market, primarily due to soaring demand from AI data centers. These large-scale buyers are securing the majority of NAND wafer supply through long-term contracts, leaving limited inventory for the consumer market. According to TrendForce, NAND contract prices rose 55–60% in Q1 2026, with enterprise SSD prices climbing 53–58%. Retail prices rose even more sharply due to constrained supply in the distribution channel. Unlike the 2016–2017 price cycle caused by production transitions, the current spike is demand-led. AI data centers are consuming NAND capacity at an unprecedented rate, with 2026 demand growth estimated at 20–22% against supply growth of only 15–17%. Manufacturers are prioritizing high-margin enterprise products over consumer-grade storage, further tightening retail availability. New production capacity from major suppliers like Samsung, Micron, and Kioxia is not expected until late 2027 or 2028. Until then, consumer storage prices are likely to remain high, with no significant price relief anticipated in the near term.

marsbit04/16 03:13

Memory Card Prices Double in Four Months: How Long Will the Surge Last?

marsbit04/16 03:13

The Allbirds, the Internet-Famous Shoes That Took Silicon Valley by Storm, Are Now All in on AI

Allbirds, the once-popular sustainable shoe brand favored by Silicon Valley elites and celebrities, has announced a drastic pivot from footwear manufacturing to AI infrastructure. On April 15, 2026, the company revealed plans to abandon its shoe business entirely, rebrand as "NewBird AI," and focus on GPU-as-a-service and AI cloud solutions. The move caused its stock to surge over 800% in a single day. The brand, known for its wool-based eco-friendly shoes, had struggled financially in recent years. Revenue fell from a peak of $298 million in 2022 to $152 million in 2025, with cumulative losses of $419 million over five years. In March 2026, Allbirds sold its intellectual property and footwear assets for just $39 million—a fraction of its former $4.1 billion valuation. The company secured up to $50 million in convertible notes to fund the acquisition of GPU hardware for AI compute leasing. However, the announcement lacked details about technical capacity, clients, or infrastructure plans. Critics highlight the high execution risks in the competitive AI infrastructure market, dominated by major cloud providers. The shift reflects a broader trend of companies rebranding around AI to attract investor interest, despite uncertain fundamentals. Allbirds also removed its "public benefit" corporate mission, signaling a departure from its original sustainability ethos. The move underscores the power of AI narrative in today’s capital markets, where storytelling often precedes substance.

marsbit04/16 02:13

The Allbirds, the Internet-Famous Shoes That Took Silicon Valley by Storm, Are Now All in on AI

marsbit04/16 02:13

Agents Have Entered the Harness-Driven Era

The article discusses the significance of the leaked Claude Code from Anthropic, highlighting its revelation of advanced Agent engineering practices centered on "Harness" design. Rather than relying solely on model capabilities, modern AI systems now depend on a structured engineering framework—the Harness—to maximize performance. This framework includes six core components: multi-layered System Prompts, Tool Schema, Tool Call Loop (with Plan and Execute modes), Context Manager, Sub-Agent coordination, and Verification Hooks. The Harness enables tighter integration between training and inference, supports long-chain tool execution, and improves reliability through objective verification. It also drives six key training directions: behavior alignment via System Prompt, end-to-end tool-use training, integrated plan-execute training, memory compression, sub-agent orchestration, and multi-objective reinforcement learning. The shift to Harness-driven development reduces the emphasis on pure prompt engineering, favoring instead multidisciplinary talent with skills in AI, backend engineering, and infrastructure. The market is evolving toward more secure, private, and vertically integrated Agent deployments, with "model shell" companies needing either strong infrastructure or deep domain expertise to compete. Claude Code’s leak underscores that future AI advancements will be shaped by engineering architecture as much as by algorithmic innovation.

marsbit04/15 10:11

Agents Have Entered the Harness-Driven Era

marsbit04/15 10:11

活动图片