We've secured $594M to build the global AI Inference Cloud with @megaportnetwork
2,048 NVIDIA B300 GPUs coming to our platform this Fall on-demand, globally distributed across Megaport's 1,100+ data centers in 31 countries.
Fast inference, close to your users. No egress games.
Full story โ https://t.co/DPuEUpQ294
Today we're launching the new Activity explorer on OpenRouter.
It's the best way to see how much and your team are spending on every model, along with tokens, cache hit rate, agents, & trends. All updated in real time.
See how our team is using Fable and other models ๐
Day 1 is here. We're LIVE at The AI Summit London. ๐
The energy at The AI Summit today is electric, 5,000+ of the most curious, driven people in AI all under one roof. It's exactly the kind of room we love being in.
Come find us at Booth #411. Weโre here all day and would love to connect.
What are you most excited to see or learn today? Drop it in the comments. ๐
#TheAISummit #AISummit2026 #LondonTechWeek #HydraHost #BareMetal
Even with GPUs at insane prices, CTOs are tempted to "๐ฃ๐ถ๐บ ๐ฐ๐ฏ๐ค๐ฆ, ๐ค๐ณ๐บ ๐ฐ๐ฏ๐ค๐ฆ" to own an AI GPU cloud.
But there ๐๐๐ hidden costs. Power, cooling, and security add up to 40% more than the purchase price of just the AI chips.
Read more: https://t.co/eBX5Qr2QuS
Healthcare CTOs with AI initiatives: GPUSeeker helps you navigate power & datacenter constraints during our current GPU scarcity crisis.
Learn more: https://t.co/PXpeHpXnB0
The AI Summit London is just around the corner, and we want to meet you there.
Hydra Host will be at Booth #411 on June 10โ11 at Tobacco Dock. Whether you're exploring new infrastructure solutions, want to talk shop about AI, or just want to shake hands, come find us.
๐ June 10โ11
๐ Tobacco Dock, London
๐ฏ Booth #411
Who else is going? Tag someone we should connect with. ๐
#TheAISummit #HydraHost #AIFactory #BareMetal #SovereignAI
This month is, unsurprisingly, Cost Reduction Month.
In our data from the last 3 yrs, we commonly see major cost crunches right after the latest breakthrough.
We'll ship major features to help you cut inference costs at least once a week, starting with today.
Running list ๐
๐ฃ Korea's telecom network is becoming national AI infrastructure.
@SKtelecom plans to build a gigawatt-scale AI Cloud in Korea, powered by the NVIDIA DSX platform. ๐ฐ๐ท
The first AI factory comes online in 2027 to support sovereign, physical and agentic AI services for enterprises and industries across Korea โ with plans to expand across Asia.
๐ Read the announcement: https://t.co/knJ5j81cEa
Prices high from a lack of GPU chips? It's also a dearth of HBM and power that constrain 2026 AI compute.
What's the winning strategy for CTOs? We recommend finding partners with neocloud capacity... GPUSeeker can help!
Learn more: https://t.co/xNe0GhjXDE
๐๐จ๐ฐ ๐ข๐ฌ ๐ฒ๐จ๐ฎ๐ซ ๐๐ฎ๐ฌ๐ข๐ง๐๐ฌ๐ฌ ๐ฉ๐ฅ๐๐ง๐ง๐ข๐ง๐ ๐ญ๐จ ๐๐๐ฅ๐๐ง๐๐ ๐๐ฎ๐ญ๐จ๐ง๐จ๐ฆ๐ฒ ๐๐ง๐ ๐๐จ๐ฌ๐ญ ๐๐จ๐ซ ๐๐ ๐๐ ๐๐ง๐ญ ๐ฌ๐ฐ๐๐ซ๐ฆ๐ฌ?
We think recent announcements from NVIDIA and Intel are key to finding the best tradeoffs...
Learn more: https://t.co/b5oithItmM
AI ROI Strategy: ๐ ๐๐ฌ๐ญ๐๐ซ ๐ญ๐ซ๐๐ข๐ง๐ข๐ง๐ ๐ฆ๐๐๐ง๐ฌ ๐ฅ๐จ๐ฐ๐๐ซ ๐๐ข๐ฅ๐ฅ๐ฌ.
Optimize your infra by removing software abstraction layers.
Learn more: https://t.co/dptY4H3MHs
NVIDIA $NVDA AND SK HYNIX JUST ANNOUNCED A MULTIYEAR TECHNOLOGY PARTNERSHIP
SK Hynix will codevelop memory for four specific NVIDIA platforms: Vera Rubin AI supercomputers, Vera CPUs, RTX Spark-powered PCs, and Jetson Thor robotic computing platforms.
SK Hynix will apply AI to semiconductor chip design and manufacturing itself, using NVIDIA's CUDA-X libraries and PhysicsNeMo framework to accelerate chip simulations and engineering workflows.
SK Hynix will build factory digital twins using NVIDIA Omniverse, targeting fully autonomous fab operations.
"AI factories are the engines of the next industrial revolution, and advanced memory is essential to their performance." - Jensen Huang
It is weird to think that we're the only provider on the planet with any on-demand / short-term $AMD MI300x availability right now. We don't have much though. Thankfully, we're about to add another VM server to the mix. Demand is high and people are signing up left and right.
Funny thing is that we've been using that server for the last few days to automate our entire vm image build pipeline, as well as support Ubuntu 26.04. It is wild, we don't even have enough compute to do our own development, so when we have a server free up, we usually take a few minutes to improve our own processes.
We're speaking at Vector Space Day 2026 in San Francisco this June 11.
The Vultr team will be on stage, joining 300+ engineers, researchers, and builders exploring what's next in retrieval, search, and agentic AI.
Grab your spot: https://t.co/F9UxWDzkbH
NVIDIA Hopper GPUs are still VERY viable for inference workloads. But the H100 is not just a faster version of the A100... Don't overspend on hardware that your team cannot fully use.
Read more: https://t.co/bSEFCvC7Gg
Looking for GPUaaS for your AI initiative?
Public pages โ live inventory.
Read our executive guide to securing GPU capacity now. https://t.co/PXpeHpXnB0
The bottleneck isn't GPUs anymore.
It's the three to five clouds most AI teams stitch together to get a model to production.
So as of today, we'll start calling Runpod the AI Developer Cloud.
Read Zhen's take on this through the link below.
https://t.co/gBaKNhZtNL