AMYAS #THETA ^$^

[Theta Network @Theta_Network]

AI has never been more focused on inference. Training happens once but inference is what runs for the lifetime of a model.

That ongoing demand is what Theta was built for. 🧵

[Theta Network @Theta_Network]

GPU demand across the AI industry is at record levels, and Theta EdgeCloud is no exception. Demand on our network is at an all-time high. 📈

Rent a GPU and get building: https://t.co/Ttsx2DWMMg

[Theta Network @Theta_Network]

Why are leading research labs choosing EdgeCloud? May gives a few answers. New partners across the UK and Middle East, published breakthroughs running on AWS Trainium, smarter agents and faster model serving. The work speaks for itself.

Read more:
https://t.co/lg6SxSzHqV

[Theta Network @Theta_Network]

@OfficialXYO Proud to be the first & to be partnering with another DePIN OG!

You've been a pleasure to work with. Looking forward to what we (& your tech) can do.

[Theta Network @Theta_Network]

🟢 Our new GPU node browser is live on https://t.co/jTqfxpOIii.

It's now even easier to browse AI compute and filter by GPU count, memory and reliability, then see exactly what each node offers before you rent.

Discover for yourself: https://t.co/Ttsx2DWMMg

[AMYAS #THETA ^$^ @Fahed40953316]

@Theta_Network the global shortage of AI chips (which Google and Blackstone are trying to solve physically) is exactly what Theta is tackling via blockchain by letting users rent out idle GPU/CPU power across its global peer-to-peer network for AI processing and video streaming.

[Theta Network @Theta_Network]

The main result was that response start times stayed stable as prompts got longer. ✅

At 1,000, 2,000, and 4,000 words, response start times stayed around 0.8 seconds. Heavy prompts did not disrupt generation for other users. ✅

[Theta Network @Theta_Network]

We ran an experiment at Theta EdgeCloud to test a more efficient way to serve LLMs in production.

Long prompts can slow down everyone sharing the same GPUs. We tested an architecture designed to keep latency/throughput stable as prompt length and traffic increases. 🧵

[Theta Network @Theta_Network]

We're pleased to welcome City St George's, University of London to Theta EdgeCloud's academic network, our 34th university partner globally & our 2nd in the UK. 🧵

[Theta Network @Theta_Network]

Theta EdgeCloud is the first decentralised compute platform approved by @AWS to run their custom AI silicon. We did it in collaboration with @yonsei_u, one of Asia's top AI research institutions. 🧵

[Theta Network @Theta_Network]

NEW: We just shipped PageIndex for Theta EdgeCloud AI agents. It fixes one of the most common reasons AI agents give bad answers, and the benchmark numbers are worth a look.

It's also already available for your Theta-powered AI agents. 🧵

[Theta Network @Theta_Network]

April was a big one. @Twitch integration for AI Agents, Illinois Tech joining our academic network, TDROP rewards on every dollar of compute spend, @alibaba_cloud & @CloudicianTech joining as an Enterprise Validator & more.

Read the roundup here: https://t.co/RHhuFvYanG

[Theta Network @Theta_Network]

ICYMI: Theta EdgeCloud AI Agents can now be seamlessly integrated into Twitch 👾

[Theta Network @Theta_Network]

Once you hit 1,000 TDROP, you can either convert it to EdgeCloud credit and put it straight back towards future usage, or withdraw it to your own wallet.

Head to your EdgeCloud dashboard to check your balance. The rebate is live today.

👉 https://t.co/4W1GNWLBxN

[Theta Network @Theta_Network]

As of today, Theta EdgeCloud now rewards you for your compute spend. Every dollar you put towards GPU usage earns you 5% back in TDROP automatically. No opt-in required. (1/2)

[Theta Network @Theta_Network]

Your GPU has idle hours. Theta EdgeCloud puts them to work on jobs consumer hardware handles well: inference, fine-tuning, batch processing, video tasks.

Run an Edge Node on Windows, Mac or Linux. Contribute spare cycles, earn TFUEL 🔥

👉 https://t.co/I6UehAV0E4

[Theta Network @Theta_Network]

And run EdgeCloud client from here: https://t.co/rLUyBJXvIe

[Theta Network @Theta_Network]

Discoverability is one of the quieter but more important shifts happening in the neocloud category right now.

Find us here: https://t.co/Wtj9A8vdbS
Read more here: https://t.co/ytwKpA9UAm

[Theta Network @Theta_Network]

The FP8 precision format makes this possible at the edge. Lower memory requirements mean consumer-grade GPUs can participate in serving a 32B parameter model that would otherwise need dedicated data center hardware.

[Theta Network @Theta_Network]

Decentralized AI inference, running in production on distributed community hardware.

Read more here: https://t.co/6XTdI7ZArI