QumulusAI

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

"Becoming an NVIDIA Cloud Partner affirms what our customers already experience — enterprise-grade @nvidia compute, delivered at hyperspeed.” — Mike Maniscalco, CEO of QumulusAI > Read the full announcement: qumulusai.com/articles/qumul…

While we kept the East Coast busy with our first day of trading on the @Nasdaq today, our CEO, Mike Maniscalco, kept busy on the West Coast meeting with clients and partners. Still, he found a moment to reflect on where we've been, where we're headed, and to offer a token of gratitude to everyone inside the Q.

"We started QumulusAI on a simple conviction: as AI moves from research to production, the winners won’t be whoever buys the most GPUs. They’ll be whoever can bring AI infrastructure online with discipline, transparency, and speed. That conviction drives everything we do, and the momentum behind it has never been stronger. " — Mike Maniscalco, CEO of QumulusAI > Read the full letter: qumulusai.com/articles/a-let…

“QumulusAI was built for how AI is actually deployed — inference-first, distributed, and close to demand,” — Mike Maniscalco, CEO of QumulusAI. Common shares have begun trading on the Nasdaq Global Market under the ticker symbol QMLS. > Read more: qumulusai.com/articles/qumul…

Welcome to the Starting Line. Today, QumulusAI begins trading on @Nasdaq under the ticker $QMLS. We’re grateful for the trust of our clients, the dedication of our team, and the conviction of our investors. This milestone marks the beginning of our next chapter—one we’re ready to build at hyperspeed.

The first few years looked nothing like the company you see today. It was a small team, a big idea, and a handful of people willing to bet on it early. Our founders, Chris Bissell and Houston Aderhold, incorporated that company in Georgia exactly seven years ago. Now look where we are... and where we're headed next. Happy Birthday to Q!

"We believe AI demand continues to outpace infrastructure supply, and we designed our hyperspeed and capital-efficient model to close that gap. This direct listing provides us with the platform to scale available infrastructure, engage with a broader investor base and continue delivering accelerated compute to the enterprises building the future of AI." — Mike Maniscalco, CEO of QumulusAI. qumulusai.com/articles/qumul…

Our latest Investor Presentation just dropped at investors.QumulusAI.com. Dig into the details, then subscribe to email notifications for every update as we break AI's biggest barriers.

We brought 360 Blackwell GPUs online in June. 45 new Blackwell and Blackwell Ultra nodes, live in Philadelphia. That's a month of deployment for builders who need the highest class of compute not soon, but now.

A new $15M @QumulusAI loan has moved into escrow. - 272 B300 GPUs - Earning 7% in escrow - Will earn 15% once funded This is part of USDAI’s $500M facility with Qumulus, following prior $4.3M and $16.4M financings. More to build together.

Our fleet of 2,500+ GPUs is live across Kansas City, Philadelphia, and Atlanta. It's a distributed U.S. footprint, deployed and running today, putting real capacity close to the teams that depend on it. And there's even more coming online soon.

256 @nvidia B200 GPUs, live in Philadelphia. A 32-node Blackwell cluster on a 400G RoCE fabric, online now and ready for a team building what's next for AI. The most sought-after silicon on the market is deployed and running inside QumulusAI.

AI demand is only getting more complex from here. Between long-term commitments, short-term spikes, internal business needs, and agent-driven application, the infrastructure layer MUST become more dynamic. In this conversation with @vcluster, we discuss how multi-tenancy is evolving, why isolation matters more, and how infrastructure models may need to adapt as AI systems become more autonomous and more variable in how they consume resources. Full conversation available on YouTube.

Enterprise AI is moving from pilots and experiments to full production systems that create value. Mike Maniscalco of QumulusAI and James Coomer of DDN discuss why the next phase is industrialization. > Watch the full conversation on YouTube: youtube.com/watch?v=raOQgn…

We're on the floor at AI Engineer's World Fair this week. Mark Jackson, our Sr. PM, is the guy to find. Look for the most "lit" backpack in the crowd. 👀 If you're figuring out your H2 GPU strategy, Blackwell clusters, or inference capacity, this is the conversation to have. B300 clusters available Q4. Frontier open models, private deployment, cheaper tokens for your agents. Find Mark. Ask him anything. 📍 San Francisco, CA — June 29–July 2

The more your organization uses an LLM, the more it costs you. How much more? You'll find out at the end of the month. Mazda Marvasti, Ph.D., CEO of Amberd.ai and QumulusAI client, on why unpredictable AI pricing is a dealbreaker for the enterprise — and why fixed monthly cost was non-negotiable when he chose his infrastructure. Watch the full playlist on YouTube: youtube.com/playlist?list=…

The AI infrastructure market is full of claims. So, how do you know which partners and components actually hold up in delivery? In this conversation with @vcluster , we talk about ecosystem fit: why flexibility, trust, security, and performance matter not just at the product level, but across the broader solution customers depend on. > Watch full conversation here: youtube.com/watch?v=0hScOQ…

@GeorgiaTech is one of the largest sources of AI talent in the country. Tim Lieuwen says every graduating student is AI-native by design. That talent pipeline is exactly why @nvidia is so intentional about its presence on campus.