LIVE · FRI, JUN 26, 2026 --:--:-- ET
Issue Nº 66 COST TOTAL $14555.55 ARTICLES TODAY 8 TOKENS TOTAL 9.15B
Podcast
aiexpert
Subscribe · $8/mo
Running the wire
Funding Onsemi acquires Synaptics for $7B to consolidate Edge AI and Physical AI stack Chips Apple skips M6 Pro/Max, fast-tracks AI-focused M7 to late 2027 as bandwidth race heats up Market OpenAI leans toward delaying IPO to 2027 over market volatility, holding firm on $1 trillion valuation Chips Solidigm ships 122TB SSD with unlimited 5-year random-write endurance; 84% less NAS power than HDD+TLC Breaking Google launches Gemini 3.5 Flash: outperforms Pro tier on coding, 40% cheaper, 4x faster Funding SE3 Labs emerges from stealth with €5.5M for autonomous defense AI, backs German sovereignty Market Anthropic, OpenAI face enterprise cost-cutting as customers switch to cheaper models Research OpenAI Codex agents now primary tool across all departments; 80% of users complete 30+ minute tasks Chips IBM's 0.7nm Nanostack breaks sub-1nm barrier with 100B transistors on fingernail die Chips Nvidia triple-qualifies HBM4 suppliers; SK Hynix, Samsung, Micron all production-ready for Vera Rubin Q3 ship Market Micron Q3 earnings blowout: $41.5B revenue quad, $22B customer contracts lock HBM supply through 2027 Breaking Hugging Face Ships vLLM on HF Jobs: spin OpenAI-compatible LLM endpoint in one command Policy White House EO mandates federal PQC migration by 2030-2031; quantum-safe silicon demand accelerates Market NVIDIA GeForce NOW doubles down on cloud gaming with Steam Summer Sale discounts Market SK Hynix targets $29B Nasdaq ADR listing July 10; memory chip giant aims for US re-rating Funding SpaceX formalizes $60B Cursor acquisition, largest startup deal ever; xAI coding tools consolidation Funding Anthropic signs 1GW+ data center leases with Google financial backing, pivots from cloud rentals Research Sakana Fugu Ultra: multi-agent orchestrator scores 95.5 GPQA, 73.7 SWE-Bench Pro, routes around export controls Market Micron shatters records: FY Q3 $41.5B revenue, 84.6% gross margin, $50B Q4 guide at 86% Market Micron Q3 blowout: 84.9% gross margin, HBM4 ramp locks in pricing power through 2027 Funding Onsemi acquires Synaptics for $7B to consolidate Edge AI and Physical AI stack Chips Apple skips M6 Pro/Max, fast-tracks AI-focused M7 to late 2027 as bandwidth race heats up Market OpenAI leans toward delaying IPO to 2027 over market volatility, holding firm on $1 trillion valuation Chips Solidigm ships 122TB SSD with unlimited 5-year random-write endurance; 84% less NAS power than HDD+TLC Breaking Google launches Gemini 3.5 Flash: outperforms Pro tier on coding, 40% cheaper, 4x faster Funding SE3 Labs emerges from stealth with €5.5M for autonomous defense AI, backs German sovereignty Market Anthropic, OpenAI face enterprise cost-cutting as customers switch to cheaper models Research OpenAI Codex agents now primary tool across all departments; 80% of users complete 30+ minute tasks Chips IBM's 0.7nm Nanostack breaks sub-1nm barrier with 100B transistors on fingernail die Chips Nvidia triple-qualifies HBM4 suppliers; SK Hynix, Samsung, Micron all production-ready for Vera Rubin Q3 ship Market Micron Q3 earnings blowout: $41.5B revenue quad, $22B customer contracts lock HBM supply through 2027 Breaking Hugging Face Ships vLLM on HF Jobs: spin OpenAI-compatible LLM endpoint in one command Policy White House EO mandates federal PQC migration by 2030-2031; quantum-safe silicon demand accelerates Market NVIDIA GeForce NOW doubles down on cloud gaming with Steam Summer Sale discounts Market SK Hynix targets $29B Nasdaq ADR listing July 10; memory chip giant aims for US re-rating Funding SpaceX formalizes $60B Cursor acquisition, largest startup deal ever; xAI coding tools consolidation Funding Anthropic signs 1GW+ data center leases with Google financial backing, pivots from cloud rentals Research Sakana Fugu Ultra: multi-agent orchestrator scores 95.5 GPQA, 73.7 SWE-Bench Pro, routes around export controls Market Micron shatters records: FY Q3 $41.5B revenue, 84.6% gross margin, $50B Q4 guide at 86% Market Micron Q3 blowout: 84.9% gross margin, HBM4 ramp locks in pricing power through 2027
Chips

Solidigm ships 122TB SSD with unlimited 5-year random-write endurance; 84% less NAS power than HDD+TLC

Solidigm announced it is now shipping the D5-P5336, the world's highest-capacity PCIe SSD at 122TB (actual 122.88TB), and the first SSD with unlimited random-write endurance over a 5-year period. The drive, sampling to customers since announcement and now in production, doubles Solidigm's earlier 61.44TB model. SK Hynix subsidiary Solidigm uses 192-layer QLC (quad-level cell) NAND and maintains the same controller across its 7TB-to-122TB product line, simplifying customer qualification and deployment at scale.

The drive consumes up to 84% less storage power in NAS deployments versus legacy HDD+TLC solutions, delivers 3.4x more terabytes per watt than 30TB TLC drives, and enables up to 4 petabytes of storage per rack unit with a 4:1 footprint reduction versus hybrid HDD+TLC. Power draw is 24 watts active and 5 watts idle—competitive with traditional PCIe 4.0 SSDs despite 2x the capacity. Sequential reads hit 7.4 GB/s (saturating PCIe Gen4), while endurance is rated at 0.6 DWPD (drive writes per day), or 134.3 petabytes written over the warranty period.

The 122TB drive targets AI data pipelines, object storage tiers, and hyperscale deployments where space and power constraints are acute. Micron is shipping a competing 122TB 6600 ION PCIe 5.0 SSD, and Solidigm has public roadmaps to ~245TB drives in 2026 as SK Hynix ramps 321-layer QLC. The move signals broader industry shift from HDD+SSD hybrid architectures toward all-flash data centers driven by AI workload density and energy-per-terabyte math. Dell has already integrated the drive into PowerScale systems.

For architects: 122TB-class SSDs change the storage-vs.-compute trade-off for AI inference and training pipelines. Teams building greenfield data center stacks should evaluate all-SSD architectures where density + power efficiency beat per-GB cost of HDDs. The unlimited random-write endurance claim is contingent on 32KB I/O patterns; 4K random write endurance sits lower (12% remaining after 5 years per spec). Customers should validate random I/O profiles in their workloads before committing capacity. NeoCloud providers and AI infrastructure operators (CoreWeave, Lambda, Crusoe) are early adopters.

Sources

Get the signal before the noise.