A new interactive dataset tracking historical memory and storage prices from 1960 to 2026 shows the lowest retail $/GB for DRAM, NAND flash, and HBM, with breakdowns by generation and accelerator cost…
The US grid cannot add capacity fast enough to meet surging datacenter power demand from AI labs and hyperscalers, driving behind-the-meter (BTM) solutions to power over half of new US datacenters by …
Micron Technology is set to report fiscal Q3 earnings after the close, with analysts expecting sales of $35.6 billion and adjusted EPS of $20.49, exceeding management's guidance. The memory chip maker…
The era of subsidized AI API usage is ending as platform providers like OpenAI and Anthropic face unsustainable losses, with OpenAI reporting a $38.53 billion net loss in 2025. Hardware costs, particu…
SGLang achieved a 5x throughput improvement for serving DeepSeek-V4 on NVIDIA GB300 hardware since its Day-0 launch, reaching ~11,200 tok/s/GPU at 50 tok/s/user through kernel and runtime optimization…
OpenAI reported $13.07 billion in revenue against $34 billion in costs and expenses in 2025, resulting in a net loss of $38.53 billion, according to financials seen by Ed Zitron. The losses, nearly ei…
A viral claim that half of 2026 US datacenter capacity is canceled is based on flawed data and uninformed forecasts, according to SemiAnalysis. The firm's own tracking shows only ~1% movement in hyper…
Rising AI token costs are forcing companies like Pylon to set usage caps, reversing the era of unlimited spending. Pylon CEO Marty Kausas faced a $1.4 million bill and now limits tokens for non-techni…
OpenAI's net loss soared from $5 billion in 2024 to $39 billion in 2025, according to audited financial figures. The company is considering shifting from flat-rate subscriptions to token-based billing…
Coinbase launched an MCP-based tool enabling AI agents to trade, pay for research, and purchase compute via the x402 protocol, initially for crypto with plans to expand to stocks and prediction market…
SemiAnalysis purchased every OpenAI and Anthropic subscription tier and maxed weekly usage with long-horizon coding and agent tasks, finding that a fully-used $200 ChatGPT Pro plan could cost OpenAI a…
Rio de Janeiro's municipal IT company, IplanRIO, has released Rio 3.5 Open 397B, a large open-weight AI model post-trained from Alibaba's Qwen 3.5 397B base model. The release was spotted by SemiAnaly…
A proposal for an AML-like system to monitor AI infrastructure using open-source intelligence aims to detect hidden training runs and rogue data centers, potentially serving as a verification mechanis…
Nvidia's Blackwell architecture achieves 20 times more AI agents per megawatt than its Hopper generation, according to the new AgentPerf benchmark. The efficiency leap, driven by FP4 precision and adv…
Anthropic released Claude Fable 5 on June 9 with built-in classifiers that silently downgrade users attempting to develop rival AI models, targeting Chinese AI labs. The company faced immediate backla…
CoolIT Systems announced a 15kW single-phase direct liquid cooling coldplate on June 1, 2026, achieving nearly four times the capacity of its prior 4kW design and over ten times the cooling required f…
AMD's MI300X accelerator, with 192GB of HBM3 memory and roughly half the list price of NVIDIA's H100, remains underutilized due to software incompatibilities. As of early May 2026, running vLLM with D…
A compiler engineer at SemiAnalysis spent over $10,000 in a single afternoon running AI agents to fuzz NVIDIA's closed-source ptxas compiler, uncovering approximately 80 miscompilation bugs in three d…