LIM Center, Aleje Jerozolimskie 65/79, 00-697 Warsaw, Poland
+48 (22) 364 58 00
ts@ts2.pl

Category: AI

Nvidia’s New China-Only AI Superchip Outguns H20 Amid U.S.-China Tech Showdown

Nvidia is developing a China-only AI chip named B30A, based on the Blackwell architecture, designed to outpace the current H20 while complying with U.S. export controls. The B30A will be a single-die design and is expected to deliver roughly half the raw computing power of Nvidia’s dual-die B300 accelerator. The B30A will include high-bandwidth memory…
Read more

Sam Altman’s AI Bubble Warning: Is the Tech Frenzy Nearing a Dangerous Peak?

OpenAI CEO Sam Altman told tech reporters in San Francisco that the AI industry is in a frenzy and warned of a bubble. OpenAI’s private funding rounds have pushed its valuation to around $300 billion or more, with talk of a potential secondary sale at about $500 billion. Mira Murati’s new AI venture reportedly raised…
Read more

Pixel 10 Pro: Google’s AI-Powered Flagship Takes on iPhone 17 Pro

Pixel 10 Pro ships with Android 16 at launch and Google guarantees 7 years of OS updates, security patches, and feature drops. Pixel 10 Pro preserves the Pixel camera bar design, is slightly thicker than the Pixel 9 Pro, and moves the SIM tray to the top-left edge (with some regions moving to eSIM only).…
Read more

Billion-Dollar Bids, Breakthroughs & Backlash – AI News Roundup (Aug 18–19, 2025)

Arm Holdings hired Amazon AI chip engineer Rami Sinno to lead its plan to design in-house AI chips and chiplets. SoftBank took a $2 billion stake in Intel and also backed a $30 billion OpenAI investment, plus funded the Stargate AI data center project in Ohio with Foxconn. OpenAI launched ChatGPT Go in India at…
Read more

AI Stock Frenzy: Nvidia Near $4T, Tesla Scraps ‘Dojo’, Palantir Shines Amid Global AI Shake-Up

Nvidia reached a $4 trillion market capitalization, becoming the first public company to do so, with its shares closing above the milestone in early July. Nvidia remains near record highs as AI demand powers its chips, while the Nasdaq and S&P 500 posted fresh record closes in mid‑August driven by AI mega‑caps. AMD’s data-center chip…
Read more

AI Breakthroughs, Billion-Dollar Bids & Backlash – Inside the Explosive 48 Hours of Aug 17–18, 2025

OpenAI released GPT-5 during Aug 17–18, 2025 rollout, billed as a major upgrade with strong coding, math, and science capabilities, while Sam Altman acknowledged it still lacks self-learning. In China, Pan Jianwei’s team used AI to position over 2,000 neutral-atom qubits in 3D in 1/60,000 of a second, a 10× larger array than prior work,…
Read more

Hurricane Erin Intensifies as High-Tech Tools – from AI to Drones – Track Its Every Move

Erin formed in the Atlantic in mid-August 2025 and is the first hurricane of the season. As of August 15, 2025, Erin is a Category 1 hurricane with sustained winds near 75 mph and a central pressure around 993 mb, located hundreds of miles east of the northern Leeward Islands and moving west-northwest at about…
Read more

AI Breakthroughs, Billion-Dollar Bets & Backlash – Global AI News Roundup (Aug 14–15, 2025)

OpenAI released GPT-5 in August 2025, described as multiples more capable than the previous generation by Kunal Kothari of Aviva Investors. Anthropic unveiled Claude for Financial Services in mid-July 2025, setting context for the GPT-5 release. Apple plans a desk assistant robot by 2027 and an LLM-powered Siri on iPhones possibly in 2026, including a…
Read more

3D Generative AI Showdown: OpenAI Shap-E 2 vs Google’s AI Shopping View vs NVIDIA Picasso Edify

OpenAI’s Shap-E uses a two-stage pipeline where an encoder compresses 3D assets into latent implicit-function parameters and a conditional diffusion model generates new shapes from prompts. The original Shap-E, released in 2023, could turn text prompts or reference images into 3D objects in seconds and outperformed Point-E due to NeRF-based rendering and CLIP guidance. Shap-E…
Read more

Next-Gen Laptop Chip Showdown: Intel’s Lunar Lake vs. AMD’s Ryzen AI vs. Qualcomm’s X Elite – Who Wins the AI PC Crown?

Intel Lunar Lake Core Ultra 200V uses a disaggregated 3 nm TSMC N3B compute tile with 4 Lion Cove P-cores and 4 Skymont E-cores (8 cores total), SMT disabled to save ~15% die area and boost perf/W by ~5%. Intel also boosts single-thread IPC by ~14% versus Redwood Cove and widens caches to 2.5 MB…
Read more