τTaosisTaoScale

Bittensor subnet news

Subnet teams' own channels, X posts, press, newsletters and events read from the chain — GitHub lives in its own tab

OfficialX10SN102d ago

Excited to be one of the new verified providers on the Liquid Inference network by @Architect_Fi Inference is rapidly becoming commoditized and financialized, and Liquid Inference is well positioned to bring real-time price discovery and continuous competition to every token.

Quoting @BrettHarrison

The Liquid Inference router network is growing quickly. We now have 28 verified companies serving >1000 models, offering capacity dynamically and updating quotes in real-time. Average time for providers to onboard and integrate models is 3 hours, full verification within a day.

Full post on TaosisOpen on X ↗

On-chainChild keys10SN103d ago

Validator TAO.app (5CoZ…bifj) scheduled a child-key change on 125 subnets

Validator hotkey 5CoZxgtfhcJKX2HmkwnsN18KbaT9aih9eF3b6qVPTgAUbifj announced new child keys on SN1, SN2, SN3, SN4, SN5, SN6, SN7, SN8, SN9, SN10, SN11, SN12 and 113 more, applying at block 9,281,605 (~24h). Stake delegated to it there routes through its children after that.

ValidatorTAO.appSubnetsSN1, SN2, SN3, SN4, SN5, SN6, SN7, SN8, SN9, SN10, SN11, SN12, … (125 subnets)

Between blocks 9,274,365 and 9,274,433

On-chainChild keys10SN104d ago

Validator tao.bot (5E2L…eZ5u) re-scheduled a child-key change on 3 subnets

Validator hotkey 5E2LP6EnZ54m3wS8s1yPvD5c3xo71kQroBw7aUVK32TKeZ5u announced new child keys on SN5, SN8, SN10, applying at block 9,273,357 (~24h). Stake delegated to it there routes through its children after that. It replaces the changes that were pending on SN5 (block 9,270,481), SN8 (block 9,270,839) and SN10 (block 9,270,837), which will not apply then.

Validatortao.botSubnetsSN5, SN8, SN10

Between blocks 9,266,106 and 9,266,170

OfficialX10SN104d ago

Cost per token keeps falling, and every year the total bill goes up anyway. Reasoning models think longer. Agents run in loops. Context windows keep growing. Demand for inference is growing faster than our efficiency at serving it. The gap between those two curves is your margin. Pareton is here to close this gap.

Quoting @a16z

Token prices keep falling and GPU rental prices keep rising Intelligence gets cheaper, usage goes up, and a model ends up in every piece of software twitter.com/64844802/status/2108603721840275576

Full post on TaosisOpen on X ↗

On-chainChild keys10SN104d ago

Validator tao.bot (5E2L…eZ5u) re-scheduled a child-key change on 2 subnets

Validator hotkey 5E2LP6EnZ54m3wS8s1yPvD5c3xo71kQroBw7aUVK32TKeZ5u announced new child keys on SN8, SN10, applying at block 9,270,837 (~24h). Stake delegated to it there routes through its children after that. It replaces the changes that were pending on SN8 (block 9,264,359) and SN10 (block 9,265,797), which will not apply then.

Validatortao.botSubnetsSN8, SN10

Between blocks 9,263,594 and 9,263,656

On-chainChild keys10SN105d ago

Validator Tensorplex Labs (5E4z…ULde) scheduled a child-key change on 21 subnets

Validator hotkey 5E4z3h9yVhmQyCFWNbY9BPpwhx4xFiPwq3eeqmBgVF6KULde announced new child keys on SN1, SN5, SN6, SN7, SN8, SN9, SN10, SN11, SN12, SN13, SN14, SN84 and 9 more, applying at block 9,265,380 (~24h). Stake delegated to it there routes through its children after that.

ValidatorTensorplex LabsSubnetsSN1, SN5, SN6, SN7, SN8, SN9, SN10, SN11, SN12, SN13, SN14, SN84, … (21 subnets)

Between blocks 9,258,178 and 9,258,244

OfficialX10SN106d ago
Openτensor Foundaτion@opentensor

.@Pareton_ai miners achieved up to 54% faster AI inference on @engyai’s setup, using the same GPUs. On the latest Novelty Search, Pareton explains how Bittensor’s Subnet 10 improves the software that runs AI models, with each winning optimization becoming the next baseline to beat. The team also got a miner-developed optimization merged into @vllm_project, one of the leading open-source engines for AI inference. Hosted by @const_reborn Full episode in the first comment

Full post on TaosisOpen on X ↗

OfficialX10SN107d ago

Specialist vs. Generalist in an AI World. Specialists win once models commoditize. One model, one GPU class, and the margin shows up. Generalists win on the problems that sit between fields. Pure depth misses the bridge. We sit in the middle. Narrow where the search has to compound. Broad where the work is connecting hardware, workload, and the buyer. Full episode with @VenturaLabs.

Quoted post

Ep. 103 - Xavier Lyu Xavier @xavi3rlu is building Pareton @Pareton_ai Timestamps 1:10 - Takeaways From the Exploit Summit 3:00 - Pareton's Target Hardware Today 3:26 - Chasing Underutilized Non-CUDA GPUs 4:27 - How You Become Hardware Agnostic 5:35 - Campaigns: One Model, One GPU Class 6:09 - Optimizing Eighty 5090s vs. One 8:01 - Can One Campaign Span a Mixed Cluster? 8:46 - Concurrent Streams & What Clients Want 10:49 - Competitors and the Pricing Problem 13:04 - Baseten, $13B, and Vertical Integration 13:55 - What Owning the Full Stack Means 15:18 - Where the Gold Is If the Shovels Are Built 18:13 - On-Prem Deployment & Outsourcing R&D 19:20 - Why Not Just Run an LLM on a Loop? 21:47 - If the Search Space Gets Exhausted 22:28 - Models Commoditize, Competition is Margin 23:30 - Exploring GPUs Outside NVIDIA 24:23 - Designing dTAO, Then Living Under It 26:45 - The Non-Technical Side of a Subnet 27:25 - Distribution as the Real Bottleneck 28:11 - Benchmarks That Never Became Revenue

Full post on TaosisOpen on X ↗

OfficialX10SN108d ago

Join us now on Discord discord.com/channels/799672011265015819/1220143…

Quoting @opentensor

This Thursday on Novelty Search :: Subnet 10 :: @Pareton_ai Pareton uses Bittensor miners to make AI inference faster and cheaper by improving the software that runs models in production. SN10 continuously tests miner-submitted optimizations on the same workloads and hardware, keeping only the changes that improve performance without sacrificing output quality. Live on Thursday :: 9PM UTC :: via Bittensor Discord Hosted by @const_reborn

Linkdiscord.comdiscord.com/channels/799672011265015819/1220143188774227988

Full post on TaosisOpen on X ↗