Several Chinese Banks Tighten Leveraged Trading on Gold and Silver, Margin Requirement for Gold-Silver Leveraged Trading Soars to 140%

NFT trading platform Rarible has announced its official launch on Solana, with Claynosaurz serving as the platform’s first featured NFT collection. Over the coming days and weeks, Rarible will continue to onboard additional Solana ecosystem NFT projects. The company said its team has spent the past several months on development, testing, and launch preparations, while engaging with multiple NFT communities within the Solana ecosystem to learn about different projects’ stories, cultures, and community needs—feedback that has shaped several of the platform’s current product decisions. It clarified that Gacha Station, previously rolled out on Solana, was not a standalone feature but a preliminary trial ahead of this launch. Rarible’s goal is not only to list NFT collections, but also to collaborate with project teams to design market experiences tailored to each project’s unique traits. The platform noted that today’s launch is only a basic version, with plans to add more projects, features, market experience optimizations, content, and community activities in the coming period, while advancing its platform development in partnership with the Solana ecosystem.

Prediction market platform predict.fun announced that all crypto asset up/down prediction markets have launched a "one-click buy" feature, allowing users to directly purchase relevant positions.

According to market data from BIT (bit.com), tokenized SpaceX (SPCX) shares rose more than 5%, currently trading at $113.8, hitting a new intraday high.

August 6 Update: #Bitcoin ETFs: 1D NetFlow: +3,781 $BTC(+$243.49M)?? 7D NetFlow: +9,034 $BTC(+$581.79M)?? #Ethereum ETFs: 1D NetFlow: +27,749 $ETH(+$52.91M)?? 7D NetFlow: +40,578 $ETH(+$77.37M)??

Western Digital Chief Product Officer Ahmed Shihab published an article stating that the core of AI storage competition lies not in simply pursuing the fastest storage medium, but in the ability to sustainably scale capacity at an affordable cost. He noted that many architectures function well in early stages, but when data volumes expand from several petabytes to hundreds of petabytes or even exabytes, they may face issues due to runaway costs. Flash storage is ideal for high-performance, low-latency use cases such as model weights, GPU spillover, KV caches, and session contexts; HDDs, meanwhile, are better suited for large-scale, long-term, cost-sensitive data including training corpora, logs, checkpoints, compliance records, synthetic data, and inference outputs. He concluded: "Flash handles the present, HDD handles the entire lifecycle." At AI scale, storage costs themselves become an architectural challenge. Storing large batches of data on high-performance media long-term will crowd out budgets for computing, networking, power, and personnel. For many bulk storage workloads, the question is not whether flash can store data, but whether customers can afford the full cost of using flash over the long term. Future AI storage will not be dominated by a single technology; instead, it should be designed in layers based on performance, cost, power consumption, density, reliability, and data lifecycle. He stressed that this is not a competition between flash and HDDs, but rather matching the appropriate medium for different workloads from the outset—otherwise, the architecture could undermine business sustainability as it scales.

Tesla said TeraFab will be built in Grimes County, Texas, adding that the number of chips required by Tesla and SpaceX will far exceed current and future global production capacity. (Jinshi)