
MDLog Labs
155 posts



0G APAC 黑客松已圆满落幕,评审工作正在紧锣密鼓地进行中。 The 0G APAC Hackathon has successfully concluded, and the judging process is now in full swing. 在过去的两个月里,我们共同见证了: • 亚太地区 295 份技术项目提交 • 线上 8 周的 Builder Hour 与 Workshop AMA 交流 • 线下横跨亚太 5 大城市的 Builders Open AI Day Over the past two months, we have witnessed: • 295 technical project submissions from the APAC region • 8 weeks of online Builder Hours and Workshop AMA sessions • Offline Builders Open AI Days spanning across 5 major APAC cities 这一切始于一个核心发问:要构建真正可验证、兼顾隐私且原生上链的 AI Agent,还有哪些底层痛点尚待解决? It all started with one core question: To build AI Agents that are truly verifiable, private, and native on-chain, what underlying pain points remain to be solved? 从智能体基础设施、可验证交易,到智能体经济、Web 4.0 与主权隐私 —— 亚太开发者用行动诠释了 0G 为何是“为 AI Agent 而生的区块链”。致敬亚太开发者的硬核交付,感谢大家一路同行。 From agent infrastructure and verifiable transactions, to agent economies, Web 4.0, and sovereign privacy — APAC developers have demonstrated through action why 0G is "the blockchain built for AI Agents." Salute to the hard-core delivery of our APAC developers, and thank you all for walking this journey with us. 项目: Projects: @Agentra69 @AIsphere0G @AlphaJournal_ @AetherCredit @TrustFolio_0G @LokiLabsAi @MDLOG_labs @coal_payments @MemoriaDA_ @SealedMind_0G @SunflowerLau27 @synapsemesh @blindmarkt @Ivaronix @VAMVault_ @herald_protocol @Ghast_AI @stealth_pay @Wraith__DA @NeoSoulAI @ayushsaklani976 @Lineage_0G @yieldgeko @zer0Gig @0gclawforge @Proof__Pay @0xgents @zerovulntech @0arena_labs @GhostProver 贡献者: Contributors: @zax_raider @emmanuel_haanks @czar_skah @alhadji_ACE @rajkaria_ @GroverInnovate @Vigothecreator @encrypt_wizard @czar_skah @bapeetttt @Armanibanks100 @exluminated @kurokurohitam Build For 0G APAC Hackathon! Build on 0G!


What if your vault could think — and prove it? Aegis Vault: AI strategies, on-chain policy guards, TEE-sealed execution. Every trade verifiable. Built on @0G_labs for #HackQuest 👇 Aegis Vault hackquest.io/projects/Aegis… #0GHackathon #BuildOn0G @0G_labs @0g_CN @0g_Eco @HackQuest_












Lately, the term “parallelized optimistic execution” has been thrown around a lot of people have asked me what “parallelized optimistic execution” actually means. Keep reading below to find out 👇 In Proof-of-Stake consensus, validators receive a block from a single proposer at that height. Usually, validators wait for +2/3 of the network’s voting power to approve that block before a validator commits and executes the transactions inside of the block. Optimistic execution refers to validators executing blocks before the block has been committed. Instead of waiting for the entire consensus round to finish (which could take multiple seconds), the validators "jump the gun" and execute block transactions in a cloned version of the state, before the block is approved. In the majority case, proposed blocks are approved in the first round of consensus, so when the validators receive +2/3 votes from the rest of the network, they only need to commit the cloned version of the state that they were working on, instead of re-executing all of the block’s transactions, which reduces the time to start validating the next block. In the minority case where the block fails the consensus round, the validators discard the results on the cloned version of the state and restart consensus with a new block proposer. No transactions can be committed without +2/3 of the voting power behind it. Parallel execution refers to two or more transactions being executed at the same time. It’s only possible if the transactions are logically proven to be mutually exclusive from each other, i.e. they update different parts of state. For example, if Alice send a transaction to interact with smart contract 1, and Bob sends a transaction to interact with smart contract 2, and SC1 and SC2 don’t interact with each other, then it’s possible to execute those two transactions in parallel on a multi-core validator. For example, SC1 and SC2 can be different CW-20 token contracts. However, if Alice sends a transaction to transfer tokens to Bob, and Bob sends a transaction to transfer tokens to Charlie, then these two transactions are dependent, so it’s not possible to execute them in parallel and they must be executed serially. In high-user environments, the majority of transactions are mutually exclusive, so executing them in parallel increases throughput and lowers block times even more. Stay tuned for Nibiru’s parallelized optimistic execution implementation!


