🚀 #MOVA# Weekly Report 🔗 Ecosystem & Operations • WLFI Supernode Campaign activated — backed by a 20M WLFI incentive pool, unlocking nodes, early access & ecosystem rewards via MovaLiquid. • MOVA × WLFI Subscription Campaign expands access — 20 rounds, 1M WLFI each, enabling seamless participation with USD1 / MOVA. 🎨 Community & Events • Labour Day ⚙️ — recognizing the infrastructure powering payments, capital flows, and the next era of on-chain finance. 🧩 Technical • Next Testnet — Jupiter: EVM Parallel Execution Breakthrough – Optimistic parallel execution (Block-STM) → 3.5×–5× faster processing – Multi-version state cache → eliminates lock contention – Sustains 10,000+ TPS in high-frequency scenarios A major step toward multi-core, high-throughput PayFi infrastructure. 📘 Full report on Medium 👇 #MovaChain# #PayFi# #Web3# #RWA# #BlockchainInfrastructure# #JupiterTestnet# #ParallelEVM#

🇭🇰 The Scaling Summit HK 2026 | Session Recap Host: @499_DAO × City University of Hong Kong Co-host: @0G_labs × OpenSchool x IDM of CityU Special Partner: @BAI_AGI x @hetu_protocol 📌 Topic: High-Performance Execution Layer & Parallel EVM Programmable Intelligence demands extreme performance. ⚡ Top system architects dissected the high-performance execution layer. From parallel EVM to decentralized agentic routing, we explored what it truly takes to process millions of concurrent, autonomous transactions. 🎙️ Speakers & Mod: ▪️ @SerenaSeek, Founder of @BlockCentral_ai (Mod) ▪️ @jinglingcookies, AI Lead at @monad ▪️ @kinnnnnnn_____, CEO of @letsburnlab ▪️@0xLaughing, APAC Lead of @GoKiteAI #TheScalingSummit# #Infrastructure# #ParallelEVM#

ICYMI @MessariCrypto published a full breakdown of @GravityChain TL;DR - World computer for the agent era - Gravity Skill for AI coding agents - 12,000+ TPS powered by Gravity’s parallel EVM - Native oracle for real-world data Read the full report.

#FairyTaleTuesday# The Queen of Elphame (The Elfin Queen of Faerie) whisking away Thomas The Rhymer in one of Scotland's greatest folkloric ballads. Her seductive role is paralleled in 'Tam Lin', who escapes with the help of a mortal woman, unlike Thomas who returns to his otherworldly amour. (image H.M Brock's Book of Ballads)

SpaceX has almost finished writing V1.0 of an in-house AI training stack in C that exact-maps to 220k GB300s with 800G NICs, making heavy use of pipeline parallelism and getting as close to bare metal as possible. The potential speed improvement vs JAX for large training runs is over an order of magnitude.

Twinkle is now at v0.4.0! 🔥 The fully open-sourced solution for multi-tenant Training-as-a-Service, with Tinker API compatibility. Now packed with broader model coverage, more training algorithm support, and an improved backend built to scale. Here’s what’s cooking: 🐳 DeepSeek V4 Support: Flash FSDP2 + Expert Parallelism (EP) training, plus native tool-call parsing and cleanup. 🤖 Qwen3.5 Evolution: Maximize efficiency with padding-free / packed-sequence support and MoE GatedDeltaNet sequence parallelism. 🔮 Gemma 4: Full multimodal training support is officially here, complete with a fresh 12B cookbook! 🧬 LoRA Level-up: Added rsLoRA for Multi-LoRA, FSDP2 for Multi-LoRA SFT, and EP LoRA SFT examples for DeepSeek V4 and Qwen3.5 MoE. ⚡ NPU Acceleration: Huge stability and speed gains with fused operators (RMSNorm, RoPE, SwiGLU, SDPA) and FLA patches. Time to supercharge your cluster and squeeze out every ounce of compute. 🏎️💨 👉 Check out the full release notes at and drop us a ⭐ on GitHub: ❤️

Understand entire hour-long videos and wield tools and search — an efficient multimodal model with 30B total params but only 3B active at inference 🎬 Title: Kwai Keye-VL-2.0 Technical Report URL: 🎬 Overview An open-source multimodal foundation model from Kuaishou, built for long-video understanding and agentic intelligence. It's a Mixture-of-Experts (MoE) model with 30B total parameters but only 3B activated at inference. ❓ Challenges Solved Processing hour-level videos demands enormous compute. ・Many frames make long-range temporal dependencies hard to capture ・The challenge was addressing that compute constraint while keeping strong performance across diverse tasks 💡 Methodology & Proposed Approach ・Long-context: adapts DeepSeek Sparse Attention (DSA) to GQA-based architectures for lossless 256K context processing, capturing key frames and long-range temporal dependencies ・Infrastructure: scalable video I/O, heterogeneous ViT-LM parallelism, custom DSA kernels ・Training: Cross-Modal Multi-Teacher On-Policy Distillation (MOPD) with Context-RL and Video-RL to address catastrophic forgetting during multi-task alignment 📊 Experimental Results ・State-of-the-art among models of similar scale ・Especially strong on fine-grained temporal localization (TimeLens) ・Excels at long-video comprehension on Video-MME-v2 and LongVideoBench ・Also capable at multimodal agent collaboration across Code, Tool, and Search, with self-correction 🌍 Use Cases It fits long-video understanding, search, and moderation, plus backbones for video-handling autonomous agents. As the first application of sparse attention to multimodal at this scale, its big strength is making hour-level video processing cost-realistic. #VideoUnderstanding# #Multimodal#

New course: Efficient Inference with SGLang: Text and Image Generation, built in partnership with LMSys @lmsysorg and RadixArk @radixark, and taught by Richard Chen @richardczl, a Member of Technical Staff at RadixArk. Running LLMs in production is expensive, and much of that cost comes from redundant computation. This short course teaches you to eliminate that waste using SGLang, an open-source inference framework that caches computation already done and reuses it across future requests. When ten users share the same system prompt, SGLang processes it once, not ten times. The speedups compound quickly, especially when there's a lot of shared context across requests. Skills you'll gain: - Implement a KV cache from scratch to eliminate redundant computation within a single request - Scale caching across users and requests with RadixAttention, so shared context is only processed once - Accelerate image generation with diffusion models using SGLang's caching and multi-GPU parallelism Join and learn to make LLM inference faster and more cost-efficient at scale!

We’re excited to welcome Mooncake to the PyTorch Ecosystem! Mooncake is designed to solve the “memory wall” in LLM serving. By integrating Mooncake’s high performance KVCache transfer and storage capabilities with PyTorch native inference engines like SGLang, vLLM, and TensorRT-LLM, it unlocks new levels of throughput and scalability for large language model deployments. Mooncake enables prefill decode disaggregation, global KVCache reuse, elastic expert parallelism, and serves as a fault tolerant PyTorch distributed backend. 🔗 #PyTorch# #OpenSourceAI# #LLM# #AIInfrastructure#