AI outpaces today’s infrastructure. Hyperspeed compute closes the gap: distributed, modular GPU access without long buildouts or procurement drag. We're QumulusAI, and we're breaking AI’s biggest barriers. #AI #Hyperspeed #GPU #Infrastructure

"What’s interesting about this announcement is not just the size of the agreements but the logic behind it. AI infrastructure is entering a second phase where differentiation comes from utilization and economics, not just raw accelerator counts." — @zkerravala, @SiliconANGLE > Read the full SiliconANGLE analysis: siliconangle.com/2026/06/11/…

Over $124M in new 3-year customer agreements, including @hyperbolic_labs. NVIDIA Blackwell. ~20% lower inference costs than standard builds, because we designed for inference first, not whatever whenever. > Read the full press release: qumulusai.com/articles/qumul…

Our new HQ is inside @TechSqATL at @GeorgiaTech. QumulusAI builds full-stack AI infrastructure, from the molecule of gas at the power source to the orchestration layer on top. Tech Square puts us inside one of the country's largest sources of AI talent, next to the research and startups defining what's next. Steve Gertz sat down with Tim Lieuwen, EVP of Research at Georgia Tech, to talk about what's being built here and why proximity matters most in a field that changes weekly.

The question is not which GPUs to use. It is what combination of compute, storage, software, deployment model, and timing creates the best productivity per dollar. Mike Maniscalco (QumulusAI)and James Coomer (DDN) discuss the blueprint problem in enterprise AI. > Watch the full conversation on YouTube: youtube.com/watch?v=raOQgncu…

🔴 Live NOW — watch here: ciscolive.com/global.html Our CEO Mike Maniscalco is on stage with @Cisco @Cirrascale at #CiscoLive on AgenticOps and end-to-end AI networking, from silos to scale.

▶️ Watch live today! Starts in one hour. Our CEO Mike Maniscalco joins @Cisco @Cirrascale at #CiscoLive: "AgenticOps and End-to-End AI Networking: From Silos to Scale." ciscolive.com/global.html 10:30 AM PT / 1:30 PM ET

Training frontier models needs 100,000 GPUs in one place. Agentic workloads don't. So we're building distributed dense compute where demand actually lives. Catch our CTO at @Cisco Live!

We're at #CiscoLive with opinions on AI infrastructure. Catch our CEO, Mike Maniscalco, on Center Stage (Bayside B, 10:30 AM Wed): "AgenticOps and End-to-End AI Networking." Ryan DiRocco, Jim Marino, and Stephen Hunton are on the ground too. Find us on LinkedIn. Easy to find, hard to out-engineer.

Inference platforms aren't waiting on models. They're waiting on capacity. 680 NVIDIA H200 GPUs. Two clusters. A 2-year contract with two of the fastest-scaling inference platforms in production. Built with @shadeformai. Read the press release > qumulusai.com/articles/qumul…

AI plans move fast. GPU capacity usually doesn’t. Stephen Hunton and Mike Halles will be at @Dell Technologies World for QumulusAI. If your team needs capacity for training, inference, fine-tuning, or private AI infrastructure, connect with them on LinkedIn to meet onsite.

Tech Square is one of the densest AI ecosystems in the country. As of June 1, QumulusAI is headquartered in the middle of it. New corporate HQ inside The Biltmore Innovation Center at @GeorgiaTech. Proximity isn't a perk. It's the strategy. > Read More: qumulusai.com/articles/qumul…

The GPU problem isn't supply. It's capital. QumulusAI just closed $26M in lease financing for a 50-node NVIDIA B200 cluster with TFC (subsidiary of Kingsbridge Holdings). Built to scale. Built to repeat. > Read more: qumulusai.com/articles/qumul…

Our clients are working across different hardware, network stacks, and deployment requirements. We believe that means the ability to test, prototype, and validate in realistic environments is an essential advantage. In this conversation with Lukas Gentele from @vcluster, Ryan DiRocco talks about building an infrastructure lab environment that makes that kind of work easier. As the stack keeps changing, real-world validation becomes part of the value. > Watch the full conversation: youtube.com/playlist?list=PL…

AI infrastructure is hard to evaluate when most of the critical pieces are invisible. That’s part of what makes the @DDNintelligence AI Factory on Wheels interesting. At NVIDIA GTC 2026, Mike Maniscalco, CEO of QumulusAI, joined James Coomer, SVP Product at DDN, for a conversation inside the nationally touring DDN AI Factory on Wheels. The exhibit brings AI infrastructure to life through real-world use cases, digital simulations, blueprints, containers, NVIDIA microservices, and deployment models. For enterprise leaders, that matters. Because getting from “we need AI” to “we have a working AI factory” requires more than access to GPUs. It requires a clearer understanding of how compute, storage, software, pipelines, and deployment strategy come together. QumulusAI is featured alongside DDN and other technology providers in the AI Factory on Wheels ecosystem as enterprises look for faster, more flexible paths to production AI. Watch the full conversation on YouTube: youtu.be/raOQgncuHko Contact QumulusAI sales: qumulusai.com/contact

"This new $45 million facility is less about the total dollar amount and more about the velocity of capital. In an industry where procurement delays can stretch into several quarters, the ability to lock in hardware and power capacity ahead of the curve is a vital tactical advantage." — @StevenDickens3 | CEO @HyperFRAME_Res > Read the full analyst article: hyperframeresearch.com/2026/…

You'll see a lot of AI infrastructure deployment case studies where everything went perfectly smooth — no friction. That's not how this story started. In this conversation with @vcluster, we discuss an early customer engagement that came in with real urgency: a bare metal environment, a Kubernetes requirement, and a short window to get a proof of concept stood up. This is the story of how real value was created by turning an urgent implementation into a repeatable delivery motion that serves future clients with less friction.

AI roadmaps don’t fail because of ambition. They fail because compute shows up late. QumulusAI secured a $45M convertible note facility from ATW Partners, with $15M funded to date, to accelerate GPU procurement and infrastructure deployment. For enterprise AI teams, speed to live cluster matters. That’s the story. > Read more: qumulusai.com/articles/qumul…

Serious AI teams are still being forced to make business decisions through infrastructure constraints they did not choose. That is part of what makes speed so consequential in this market. In this conversation with vCluster, we talk about why infrastructure speed is no longer just an operational advantage. It shapes how quickly teams can deploy, test, train, fine-tune, and move into production. When customers expect a cloud-like experience, but the underlying hardware market is constrained and supply chains are tight, providers have to do more than simply offer access. They have to reduce burden and help customers move. That is the focus of this discussion. > Watch here: youtube.com/playlist?list=PL… #AIInfrastructure #Kubernetes #CloudInfrastructure #MLOps #GPUCloud

RT @USDai_Official: Origination alert: $16,395,614 loan executed with @QumulusAI. 600 H200 GPUs. 75 servers. 80% LTV, 12% APR. Dec 2024 -…