
He practiced his birthday entrance routine for weeks and it shows 🥹
Blangs
18.5K posts

@TezosGaming
Hosting Tuesday Tezday • Tezos Gaming Night • TezTalks | Gaming on Tezos, building nodes/testnets, AI tooling & Etherlink updates

He practiced his birthday entrance routine for weeks and it shows 🥹








📣 Tezos Intents are coming in 2026. Move assets across chains from one smart account - while keeping full custody. → Cross-chain swaps from your Etherlink account → Zero gas to declare intent → Solver access for RFQ-based execution → No scattered wallets or bridge UX Sign up ↓

One of the most important llama.cpp PRs is still waiting for mainline: github.com/ggml-org/llama… MTP support means models that ship with MTP heads can potentially run significantly faster. Early Qwen3.6 27B MTP-enabled GGUFs are already running around 2x baseline speed. Qwen3.6 27B re-uploads are likely on the horizon, and once MTP lands in llama.cpp, 2x-speed GGUFs could become the new normal. What a time to be alive.



BREAKING: Nvidia, $NVDA, and PulteGroup are partnering with Span to install in-home mini data centers. Each packs 16 Blackwell GPUs, 4 AMD EPYC CPUs, and 3TB RAM, powered by unused household electricity for AI inference.

