QumulusAI
QMLS · NASDAQ · Technology
$3.18
$0.29 (-9.14%) today
Market Cap
$106.40M
24h Volume
$561.0K
P/E Ratio
0.0
52w High
$21.43
52w Low
$3.14
Est. AI Revenue
100%
7-Day Sparkline (Static)
Company facts
Performance
About QumulusAI
QumulusAI leases Nvidia GPU clusters to AI companies that need capacity for training and inference workloads. Customers pay for GPU-as-a-Service (GPUaaS) compute time, which is effectively all of its revenue, making it a pure GPU cloud specialist rather than a diversified cloud provider.
Its main product is GPUaaS, delivered through clusters built around Nvidia's Blackwell B300 GPUs. The company has been working to grow its base of direct customer relationships after relying heavily on a single marketplace partner in its early years.
Founded in 2019 and headquartered in Atlanta, Georgia, QumulusAI trades on the NASDAQ Global Market after a direct listing in 2026 and employs a small team of about 29 people under CEO Michael Maniscalco.
QumulusAI in the AI buildout
QumulusAI sits in the Neoclouds layer as a GPU cloud specialist, renting Nvidia GPU capacity to AI companies rather than operating as a general-purpose hyperscaler. Because GPUaaS compute is almost all of its revenue, its fortunes track AI training and inference demand directly. Supercycle scores its AI exposure at 58 out of 100, reflecting that near-total AI dependence against the risks of its small scale, customer concentration history, and unproven access to growth capital.
QMLS is tracked in Neoclouds, rated Strategic on the AI Exposure leaderboard.
By the numbers
52-week range
QMLS trades at $3.18, 0% of the way up its 52-week range of $3.14 to $21.43.
Market cap
Market capitalization is $106.40M, with an AI Exposure Score of 58 out of 100.
Next earnings
QumulusAI is next expected to report earnings on November 10, 2026.
Market data as of the close on . Prices are end of day, not live.
Frequently asked questions
What does QumulusAI do?
QumulusAI leases Nvidia GPU clusters to AI companies that need compute for training and inference. It sells this capacity as GPUaaS, GPU-as-a-Service, which makes up essentially all of its revenue.
How is QumulusAI exposed to AI?
Heavily. Its entire business is renting GPU compute to AI customers, so Supercycle rates its AI Exposure Score at 58 out of 100, a top-two driver tier adjusted down for its small scale and funding risk.
What are QumulusAI's main products?
GPUaaS cloud compute delivered through clusters built on Nvidia Blackwell B300 GPUs, sold to AI companies needing training and inference capacity.
Is QMLS a pure play AI stock?
Its revenue is almost entirely tied to AI compute demand, but Supercycle's score of 58 out of 100 reflects real risk around its small scale and capital needs rather than placing it in the top Pure Play tier.
Where is QumulusAI based and who leads it?
QumulusAI is headquartered in Atlanta, Georgia, and is led by CEO Michael Maniscalco. Founded in 2019, it employs about 29 people and trades on the NASDAQ Global Market.
Profile written 2026-10, updated as the business changes. Editorial content, not investment advice.
Related reading
- Is the AI bubble bursting? What the buildout data says · AI stocks have read Fear for six weeks while the biggest buyers keep raising their spending plans. What the buildout data says about the bubble question.
- Hyperscaler capex roundup: September 2026 · Seven AI builders spent $657B on capex over the last four quarters. The table, the 2026 guidance, what changed since the Q2 earnings round, and the power deals behind it.
- Neocloud stocks to watch: CRWV, NBIS, IREN, APLD · The four most-searched neocloud stocks read through the buildout tracker and the June 13F filings: who each one depends on, and what would change the picture.