LIVE · MON, JUN 29, 2026 --:--:-- ET
Issue Nº 69 COST TOTAL $14614.96 ARTICLES TODAY 5 TOKENS TOTAL 9.24B
aiexpert
Running the wire
Chips Cerebras AI chip IPO raises $5.5B on 68% first-day pop; WSE-3 wafer-scale competes with Nvidia Chips Palantir deploys NVIDIA Nemotron in air-gapped US govt AI stack Funding Q2 2026 records $1B+ startup exits: SpaceX $75B IPO, Cerebras $5.55B, Quantinuum $1.68B Funding Anthropic files confidentially for IPO at $965B valuation after $65B Series H; leapfrogs OpenAI Market SpaceX raises $25B in debt after IPO; poses diversification challenge for investors holding both equity and bonds Chips MLPerf Training v6.0: NVIDIA Blackwell sweeps, AMD within 5-6% on dense LLM training Funding Menlo Ventures raises $3B, largest in 50-year history, all-in on AI startups across stack Funding British Business Bank deploys £400m per year into UK scaleups via 10 first-time VC microfunds Funding SatVu closes £30M (€34M) thermal satellite round led by NATO Innovation Fund; HotSat constellation Breaking AWS FinOps Agent enters public preview; AI-powered cost anomaly investigation for cloud ops Chips YOFC hollow-core fiber hits 51.3 Tb/s over 128 miles unrepeatered; AI-era backbone milestone Funding Reed Semiconductor raises $100M for AI power delivery; oversubscribed round signals infrastructure demand Chips Samsung ships industry-first HBM4E samples at 16Gbps, 48GB per stack; 20%+ speed gain over HBM4 Market Micron guides $50B Q4 revenue, 86% margins; signs 16 strategic customer agreements worth ~$100B Chips d-Matrix Corsair inference accelerator enters full production; claims 10x faster decode than GPU-only with 5x less energy Market SoftBank commits €75 billion to build 5 GW of AI data center capacity across France through 2031 Funding UK government backs £400 million venture capital initiative for diverse fund managers Chips NVIDIA Blackwell platform arrives; B200/B300 GPUs ship with 4x H100 inference speed, 25x lower cost/energy Breaking HP deploys OpenAI Frontier across enterprise operations; joins six inaugural platform adopters Market 79% of global AI data center capacity faces elevated climate hazard risk; operators shift to rural, extreme-weather zones Chips Cerebras AI chip IPO raises $5.5B on 68% first-day pop; WSE-3 wafer-scale competes with Nvidia Chips Palantir deploys NVIDIA Nemotron in air-gapped US govt AI stack Funding Q2 2026 records $1B+ startup exits: SpaceX $75B IPO, Cerebras $5.55B, Quantinuum $1.68B Funding Anthropic files confidentially for IPO at $965B valuation after $65B Series H; leapfrogs OpenAI Market SpaceX raises $25B in debt after IPO; poses diversification challenge for investors holding both equity and bonds Chips MLPerf Training v6.0: NVIDIA Blackwell sweeps, AMD within 5-6% on dense LLM training Funding Menlo Ventures raises $3B, largest in 50-year history, all-in on AI startups across stack Funding British Business Bank deploys £400m per year into UK scaleups via 10 first-time VC microfunds Funding SatVu closes £30M (€34M) thermal satellite round led by NATO Innovation Fund; HotSat constellation Breaking AWS FinOps Agent enters public preview; AI-powered cost anomaly investigation for cloud ops Chips YOFC hollow-core fiber hits 51.3 Tb/s over 128 miles unrepeatered; AI-era backbone milestone Funding Reed Semiconductor raises $100M for AI power delivery; oversubscribed round signals infrastructure demand Chips Samsung ships industry-first HBM4E samples at 16Gbps, 48GB per stack; 20%+ speed gain over HBM4 Market Micron guides $50B Q4 revenue, 86% margins; signs 16 strategic customer agreements worth ~$100B Chips d-Matrix Corsair inference accelerator enters full production; claims 10x faster decode than GPU-only with 5x less energy Market SoftBank commits €75 billion to build 5 GW of AI data center capacity across France through 2031 Funding UK government backs £400 million venture capital initiative for diverse fund managers Chips NVIDIA Blackwell platform arrives; B200/B300 GPUs ship with 4x H100 inference speed, 25x lower cost/energy Breaking HP deploys OpenAI Frontier across enterprise operations; joins six inaugural platform adopters Market 79% of global AI data center capacity faces elevated climate hazard risk; operators shift to rural, extreme-weather zones
Chips

MLPerf Training v6.0: NVIDIA Blackwell sweeps, AMD within 5-6% on dense LLM training

The MLPerf Training v6.0 benchmark suite, released by MLCommons on June 16, 2026, shows NVIDIA Blackwell achieving the fastest time-to-train across every workload tested, with the company submitting results on all seven benchmarks—the only vendor to do so. NVIDIA's GB300 NVL72 (Blackwell Ultra) system achieved leading per-accelerator and full-scale performance on both legacy dense LLM workloads and the new 671-billion-parameter mixture-of-experts (MoE) models added this round: DeepSeek-V3 and GPT-OSS-20B. CoreWeave, running cloud infrastructure, achieved the fastest DeepSeek-V3 time on 8,192 GPUs: 2.02 minutes.

AMD's MI355X came within 5% on Llama 2-70B fine-tuning and 6% on Llama 3.1-8B pre-training versus NVIDIA B200 using comparable FP4 precision recipes (MXFP4 vs. NVFP4). However, AMD did not submit results on the new MoE benchmarks; all entries for DeepSeek-V3 were NVIDIA-only, leaving the competitive picture incomplete on sparse-model training at scale. Microsoft Azure scaled Llama 3.1 405B (dense, 405B parameters) to 8,192 Blackwell GPUs in 7.07 minutes, a record-scale training job.

For practitioners, the headline spans two layers: hardware and software. At hardware level, NVIDIA's full-stack sweep and only-vendor-on-all-tests status signals platform maturity for production large-scale training. At software level, NVIDIA reports GB300 delivered 1.3x throughput gains on DeepSeek-V3 versus GB200 in six months driven purely by software optimization (CUDA graphs, kernel fusions, MoE router improvements)—no hardware change. This indicates that enterprises with current NVIDIA GPUs can expect performance gains between hardware-generation cycles. Cloud submissions doubled versus the prior round (v5.1), signaling a structural shift toward training-as-a-service rather than on-premises GPU procurement. For chip procurement teams and inference-provider planning, AMD's 5-6% parity on dense models makes it a node-level alternative, but lack of MoE results leaves uncertainty about competitiveness on the sparse-architecture workloads becoming industry-standard.

Sources