The ZKP project uses zero-knowledge proofs to create a privacy-first network addressing AI's data bottleneck. Its Proof-of-Intelligence mechanism trains AI on encrypted data without exposure. Presale targets $1.7B, mainnet in 2027.
AMD CEO Lisa Su said on July 24 that the artificial intelligence accelerator market is expected to reach $1.4 trillion by 2030. AI accelerators are specialized chips built for AI workloads, including matrix operations used in deep learning, and are designed to handle large-scale parallel tasks with much higher efficiency and energy performance than traditional CPUs. Mainstream products in the segment include Nvidia GPUs and custom ASICs made by companies such as Broadcom. These chips form the core compute layer behind the training and inference of large language models and other advanced AI systems. The comments point to the scale AMD expects for demand in specialized AI hardware over the rest of the decade.
CryptoComLearn examines Litecoin's design: faster block times (2.5 min), Scrypt mining, and 84M max supply. Launched in 2011 by Charlie Lee, LTC positions itself as a low-cost payment network, yet remains Bitcoin's 'little sibling.'
Laszlo Hanyecz’s GPU mining breakthrough pushed Bitcoin’s hash rate up by 130,000%, boosting network security while challenging the early ideal of equal participation.
A TrendForce research note says the memory hierarchy behind AI systems is being redrawn as the industry shifts from training to inference. High Bandwidth Memory, or HBM, dominated the training era because it solved a bandwidth problem. In inference, the pressure point has changed. Key-value cache growth during prefill and decode is pushing capacity to the forefront, creating room for three different approaches: HBF for larger-capacity memory between HBM and SSD, SSD POD for offloading cache into lower-cost storage tiers, and SRAM for ultra-fast on-chip access in decode-heavy workloads. The report walks through how each technology fits into the stack rather than framing them as direct substitutes. HBF, being developed by SanDisk and SK hynix, is aimed at much larger capacity than HBM but has not reached mass production. NVIDIA’s SSD POD concept, tied to its Dynamo framework and NIXL transport library, is designed to move KV cache from limited GPU memory into CPU RAM, local SSDs, and remote storage without stopping inference. SRAM, used aggressively by companies such as Groq and Cerebras, trades capacity for speed by keeping memory on chip. The piece also points to recent product moves from NVIDIA, Google, SambaNova, Cerebras and Etched as evidence that inference-focused hardware design is accelerating.
Tether has launched a cross-platform BitNet LoRA framework through QVAC Fabric, allowing AI model training and fine-tuning on smartphones, GPUs, and other consumer hardware with lower memory use.
Nvidia closed at a record $208.27 with market value above $5.09 trillion. After Intel posted a strong first quarter, Nvidia's $5 billion investment in Intel rose to about $17.7 billion, implying an unrealized gain of roughly $12.7 billion.
Pi Network mandates node upgrade by March 12, followed by Pi Day 2026 on March 14. Pi coin jumped 9.35% in 24h to $0.216, with 29% weekly gain. AI compatibility tests and a new token framework proposal also emerge.
