Cerebras First-Earnings Drop: Why AI Chip Margins Became the New Nvidia Comparison
Q1 2026 revenue of $193.4M and 47% core margin put Cerebras under a Nvidia-sized microscope as shares slip 7.8% on guidance. What the math implies.
MarktechPost·
UC San Diego's DFlash replaces autoregressive drafting with a lightweight block diffusion model for speculative decoding. It drafts whole token blocks in a single forward pass and conditions on target hidden features through KV injection. The paper reports up to 6.08x lossless speedup on Qwen3-8B, while NVIDIA reports up to 15x throughput on Blackwell at fixed interactivity. DFlash ships 20 checkpoints and supports SGLang, vLLM, and TensorRT-LLM. The post DFlash Speculative Decoding Drafts Whole Token Blocks in Parallel for Up to 15x Higher Throughput on NVIDIA Blackwell appeared first on MarkTechPost.
Read full articleQ1 2026 revenue of $193.4M and 47% core margin put Cerebras under a Nvidia-sized microscope as shares slip 7.8% on guidance. What the math implies.
The black market surge highlights geopolitical tensions, impacting global tech supply chains and prompting China to boost domestic chip production. The post Nvidia’s banned AI chips double in price on China’s black market appeared first on Crypto Briefing.
Nvidia's dominance in AI infrastructure reshapes market dynamics, influencing prediction markets and challenging competitors' growth prospects. The post Nvidia becomes world’s largest company with $4.8T market cap in June 2026 appeared first on Crypto Briefing.
The black market surge highlights the ineffectiveness of export controls and underscores the persistent global demand for advanced AI technology. The post Nvidia’s AI chips surge in price on China’s black market, with B200 racks fetching 50% premiums appeared first on Crypto Briefing.
SAN DIEGO, June 23, 2026 — NVIDIA today announced NVIDIA BioNeMo Agent Toolkit, which provides domain-specific tools and skills for the agentic life sciences era. Including more than a decade’s […] The post NVIDIA Announces BioNeMo Agent Toolkit — Tools for Agents to Accelerate Scientific Discovery appeared first on AIwire.
US crackdown on illicit exports has made it riskier, harder and more expensive to buy tech giant’s processors
Building AI systems at scale is demanding, requiring low-latency inference, fast vector search, strong GPU price-performance and infrastructure that can grow without multiplying operational complexity. NVIDIA’s latest work with Amazon Web Services (AWS) addresses each of those constraints. Across Amazon OpenSearch and Amazon EC2, NVIDIA AI infrastructure is giving enterprises more practical paths to deploy […]
The new open reasoning model delivers 30B-class intelligence in a 16B-parameter footprint, with 3.1B active parameters, validated independently on NVIDIA accelerated computing infrastructure. DONOSTIA, Spain, June 23, 2026 — Multiverse […] The post Multiverse Computing Launches Pulsar 16B in Collaboration with NVIDIA appeared first on AIwire.