The Linux Foundation has announced that it is taking over TRACE, an open-source standard used for AI runtime attestation, according to a Techub News brief citing Crypto Briefing. The move is aimed at improving accountability in AI systems and supporting trust and compliance across different computing environments. The item did not disclose additional technical details about TRACE’s governance structure, implementation roadmap, or participating organizations. Based on the source text, the announcement centers on the Linux Foundation’s role in assuming stewardship of the standard and the intended goal of making AI systems more accountable at runtime. The report also frames the handover as a step tied to broader trust and compliance needs in cross-environment computing scenarios.

Techub News said Nvidia has released the ACES framework, a proposal aimed at changing how AI skills are assessed. The framework focuses on evaluating AI based on how it performs in real-world settings rather than relying on static checks. No further details were disclosed in the report, but the idea could have important implications for AI development.

Nvidia used this week’s Gamescom event in Cologne, Germany, to lay out new support for its upcoming RTX Spark platform, which the company said is scheduled to launch this fall. According to the announcement, the platform has backing from several game publishers, including Electronic Arts, Embark Studios and Ubisoft, and multiple AAA titles are set to arrive on it. Nvidia said RTX Spark will natively integrate Electronic Arts’ Javelin anti-cheat technology and support RTX features including DLSS 4.5. The company also outlined a broader set of gaming technology updates tied to the same announcement. Those include DLSS 4.5 ray reconstruction, which is already available, path tracing support for titles including CONTROL Resonant, and a plan to bring NVIDIA ACE technology to games including Aniimo in early 2027. The company described RTX Spark as a platform built for next-generation Windows PCs, combining personal AI agents, advanced content creation and high-performance gaming in a single environment.

Binance said it will add three new spot trading pairs — USDC/ARS, ACE/USDC and PUMP/U — at 16:00 Beijing time on Aug. 26, according to an exchange announcement cited by BlockBeats. The exchange also said spot algo order trading bot services will go live for the same pairs at the same time. Binance noted that ARS refers to the Argentine peso, a fiat currency, and is not a digital asset ticker. The announcement also said users in some countries or regions will not be able to trade these spot pairs, and that account verification is required before participation. No further rollout details were disclosed in the notice provided by BlockBeats.

Researchers Jianhao Ma of Tsinghua University and Yuxin Chen of the University of Pennsylvania’s Wharton School have released a paper addressing a roughly 40-year question in optimization theory: how far standard gradient descent can go if its structure is left unchanged and only the step-size schedule is tuned. Their result states that for any pre-specified nonnegative step-size sequence, gradient descent has a lower-bound convergence rate of Ω(T^-1.9319), ruling out the possibility that step-size design alone can match the O(1/T²) rate achieved by Nesterov’s accelerated method. The paper, as described in the report, is notable not only for the theorem itself but also for how the proof was produced. The core argument was generated through iterative work with GPT-5.6 Sol Pro, with the researchers supplying the problem target and a high-level “resisting oracle” strategy, then correcting gaps as they appeared. The result was later translated into Lean 4 code with Codex and formally checked line by line. According to the report, the final formalization used zero “sorry” and zero “admit,” meaning no proof steps were skipped. The code has been made public on GitHub alongside a traceability file linking the paper’s theorems to the Lean implementation.

Humanoid robots can already handle staged tasks such as walking, dancing, and boxing, but reliable work in messy real-world settings remains a major hurdle. ACE Robotics Chairman Wang Xiaogang told Reuters that embodied intelligence could reach its "ChatGPT moment" by the end of next year, with progress driven by world models and better environmental data capture. Founded in July 2025, the Chinese startup builds AI models for humanoid robots and is backed by Ant Group and SenseTime. It raised more than $100 million in the first half of 2026 and, according to Reuters, aims to seek an IPO as soon as regulations allow. Wang said the industry has accumulated only about 100,000 hours of data over the past few years, far too little to train embodied foundation models. Other groups are pursuing related work, including HumanoidExo, a wearable exoskeleton introduced in October to collect human motion data. Boston Dynamics and Alibaba have also rolled out new robot-focused AI systems this year.

On Aug. 22, HYPE broke above 77 USDT and set another record, while HOOD climbed 11.69% to $106.22. BTC, ETH and SOL all rose in the CEX top-volume list, and coins such as GALA, STX, ENS and BCH posted strong 24-hour gains on OKX. GMGN’s top meme list was led by UP, WWW and Hachi. The update also noted $3 billion in stablecoins minted by Circle and Tether over the past two days, a record $2.739 billion in short liquidations on Aug. 19, and fresh comments from Vitalik Buterin, Brian Armstrong, Tom Lee and CryptoQuant founder Ki Young Ju.

Crypto markets and adjacent tech news were both active on Aug. 21. Among the most-traded tokens on centralized exchanges, BTC, ETH, SOL and BNB all posted gains over the past 24 hours, while HYPE briefly rose above $73.4, according to OKX data, with a 24-hour increase of 26%. Separate ranking data from OKX and msx.com showed strong moves in spot tokens and crypto-linked equities, including a 60.15% jump in BTCT.M. On the corporate side, Bloomberg reported that Anthropic is preparing for an initial public offering that could match or exceed the record level set by SpaceX, with a public filing possible as soon as the end of August. The report said Anthropic raised capital in May at a $96.5 billion valuation, brought in $6.5 billion in financing, and had reached a $6.5 billion revenue run rate by the end of July. Bloomberg also said the company is working with Morgan Stanley, Goldman Sachs and JPMorgan and is weighing a super-voting share structure for CEO Dario Amodei. Other developments included ICE signaling openness to joining a new Polymarket funding round, the U.S. SEC proposing two fundraising exemption tracks for crypto projects, Riot Platforms signing AI data center leasing agreements worth about $9.8 billion in long-term revenue, and CZ saying a compliant U.S. entry by Hyperliquid would be positive for the broader industry.