Marvell said on Aug. 4 that it has launched the Bravera SC6 SSD Controller, model MV-SF1410, aimed at the PCIe Gen6 NVMe SSD market for AI data centers, hyperscale cloud operators, and enterprise database deployments. The company said the chip is built for a shift in AI system design toward tiered memory architectures, where NVMe SSDs act as a high-capacity memory layer rather than serving only as conventional storage. According to Marvell, that setup can provide tens to hundreds of terabytes of capacity at a lower cost than GPU HBM and DRAM, helping keep GPUs utilized while reducing infrastructure costs. Bravera SC6 supports PCIe Gen6, remains backward compatible with PCIe Gen1 through Gen5, and complies with the NVMe 2.2 specification. It also features 16 NAND channels, transfer speeds of up to 3600 MT/s, support for ONFI and Toggle NAND, and compatibility with SLC, MLC, TLC, and QLC flash. Marvell said samples are expected to become available in the fourth quarter of 2026.

A public challenge from BitGo CEO Mike Belshe has become a flashpoint in the debate over whether artificial intelligence can actually break the cryptographic foundations behind Bitcoin. On Aug. 1, Belshe transferred 100 BTC, worth about $6.5 million at early-August prices, into a BitGo wallet address and openly challenged Anthropic to take it, after a wave of discussion around Claude’s security research and the Coldcard weak-randomness incident. When GoPlusZH checked the chain while writing, the funds were still sitting untouched. The article argues that AI is changing cybersecurity fast, but mostly by improving vulnerability discovery, exploit development, and analysis of weaker or insufficiently reviewed designs. It points to Anthropic’s incident report on Claude’s cybersecurity evaluations, as well as research showing progress against the post-quantum candidate HAWK and reduced-round AES. At the same time, it draws a sharp distinction between those results and the mature cryptographic stack used by Bitcoin, including SHA-256 and secp256k1. GoPlusZH’s main point is that real-world crypto losses rarely come from “breaking encryption” in the abstract. They usually come from implementation flaws, poor key management, phishing, social engineering, or supply-chain compromise. In that framing, AI looks less like a force that overturns mathematics and more like a powerful amplifier of existing operational and security weaknesses.

NVIDIA has released storage benchmark results for its Vera CPU, saying the NVIDIA BlueField-4 STX storage processor delivered higher performance than a comparison x86 CPU across several storage-heavy tasks tied to AI-native infrastructure. According to the company, the strongest gain came in CRC32C integrity checking, where performance reached up to 3.67x that of the x86 comparison system. Reed-Solomon data recovery improved by up to 3.26x, compression by up to 3.29x, decompression by up to 1.72x, and a combined compression-plus-encryption storage pipeline by up to 3.21x. NVIDIA also reported gains of up to 1.43x for AES-128 encryption and 1.29x for AES-128 decryption. The company said the benchmarks reflect a shift in AI infrastructure as AI agents handle knowledge retrieval, tool use, long-term memory management, and larger context windows. In that setting, storage is no longer limited to basic reads and writes, but becomes part of the inference path itself, with encryption, compression, validation, and recovery tasks adding pressure to CPUs. NVIDIA said Vera is designed to raise throughput per unit of CPU resource so storage systems can support more AI agent workloads without materially increasing infrastructure costs.

NVIDIA has released new benchmark results for its Vera BlueField-4 STX storage processor, saying the company’s Vera CPU-based AI-native storage platform posted clear gains over traditional x86 CPUs across a range of storage tasks. According to NVIDIA, the tests covered encryption, compression, data integrity checks, and recovery workloads that are becoming more demanding as AI agent applications scale and storage systems handle enterprise knowledge bases, long-term memory, KV cache, tool-call data, and model outputs. In the benchmark figures disclosed by the company, AES-128 encryption throughput improved by as much as 1.43x and decryption throughput by up to 1.29x. Reed-Solomon recovery reached up to 3.26x, CRC32C data integrity checks rose by as much as 3.67x, compression throughput increased by up to 3.29x, and decompression performance improved by as much as 1.72x. NVIDIA also said overall throughput in a combined compression-and-encryption storage pipeline improved by as much as 3.21x. The company said Vera CPU uses its in-house Olympus core architecture with 88 Armv9.2-compatible CPU cores, 176 threads, Scalable Coherency Fabric, and a SOCAMM2 LPDDR5X memory system, with up to 3.4 TB/s of interconnect bandwidth and up to 1.2 TB/s of memory bandwidth.

Anthropic’s Frontier Red Team said its internal model, Claude Mythos Preview, identified two cryptographic weaknesses in largely autonomous research runs, but neither result breaks encryption used in practice today. One finding targeted HAWK, a post-quantum digital signature candidate under review at the U.S. National Institute of Standards and Technology, where the model found a previously unknown symmetry in the lattice structure and cut the attack cost for HAWK-256 from 2^64 to 2^38. The work took about 60 hours and roughly $100,000 in API usage. The second result focused on a reduced 7-round version of AES-128 rather than the full 10-round AES. Anthropic said the model developed a fingerprinting method called Möbius Bridge that removed one enumeration step and ran 200 to 800 times faster than earlier meet-in-the-middle attacks. That effort took about three days and billions of tokens, with humans providing only limited guidance. Anthropic framed the results as evidence that AI can conduct cryptanalysis research, not as proof that Claude has broken modern deployed encryption. The targets were an unadopted candidate scheme and a deliberately weakened AES variant, leaving real-world systems unaffected.

A broad set of crypto and AI developments landed over the past day, spanning regulation, market structure, fundraising, protocol upgrades and security incidents. Kenya cut the paid-in capital requirement for stablecoin issuers by 40% to about $2.32 million while keeping strict reserve and redemption rules in place. Russia’s central bank published its first draft framework for organized trading in digital assets, and Myanmar passed a cybercrime law that allows life sentences for crypto-related fraud. In the U.S., Senate Republicans are still trying to move the Clarity Act before the August recess, though ethics provisions and bank lobbying remain major obstacles. On the corporate side, PayPal posted better-than-expected second-quarter results and did not address a previously reported buyout approach. Luno and Visa both outlined layoffs tied to restructuring and capital allocation, while Morgan Stanley Investment Management rolled out exchange-traded products tied to Ethereum and Solana. Zcash activated its Ironwood NU6.3 upgrade, Layer 2 TVL on Ethereum fell to its lowest level since 2023, and Bitcoin briefly dropped below $63,000 as AI and semiconductor weakness spilled into crypto. Security reports also stayed in focus, with Blockaid saying crypto losses from hacks topped $1 billion in the first half of 2026 and several fresh token incidents reported across the market.

Anthropic said an unreleased version of its Claude Mythos model uncovered two previously unknown cryptographic attacks, including one against HAWK, a post-quantum digital signature scheme now in the third round of the U.S. National Institute of Standards and Technology’s competition. According to the company, Claude identified a symmetry in HAWK’s structure that cuts the cost of recovering a secret key in the smallest configuration from 2^64 operations to 2^38, a roughly 67 million-fold reduction. Anthropic said fixing the issue would require roughly doubling HAWK’s key size, which would undercut the compactness and speed that make the scheme attractive for blockchains and other systems sensitive to signature size. The company also described a separate result on a 7-round research version of AES-128, where Claude was asked to avoid all five established families of AES cryptanalysis and invent a new approach. Anthropic said the model eventually produced a technique called “Möbius Bridge,” reducing one of the nine key bytes an attacker previously needed to guess. It also reported that Claude broke 13 rounds of LEA, though deployed LEA uses 24 rounds. Anthropic said it disclosed the findings to the algorithms’ authors and relevant U.S. government and industry partners before publication, and coordinated the HAWK finding with NIST.

Taiwan lawyer Jerry created Donna, an AI case management assistant that reads entire case files, answers queries with page citations, and auto-files documents. Data stays on lawyers' own OneDrive, Google Drive, or NAS. Monthly cost starts at ~$60.