Exascale Labs (XLAB) Registration filing summary

Company overview and business model

• Asset-light, software-defined GPU compute platform supplies reserved and on-demand GPU capacity from third-party data centers; supports LLM training, fine-tuning and high-concurrency inference.

• GPU-as-a-Service includes bare-metal and VM configurations; infrastructure services include revenue-generating GPU cluster management for AI data center operators.

• Modular data centers, liquid cooling, HVDC power, interconnectivity and energy storage are planned as complementary offerings, pursued through partnerships and integration; these capabilities had not generated revenue as of the prospectus date.

• Reserved compute contracts generally run 3 months to 3 years, with most having one-year terms; on-demand service is pay-as-you-go and generally has no fixed term or minimum usage.

Financial performance and metrics

• FY2026 revenue was $14.8M versus $7.0M in FY2025; intelligent computing services contributed $14.7M, or 98.9% of FY2026 revenue.

• FY2026 gross profit was $2.4M and gross margin 16.3%, versus $1.1M and 15.8% in FY2025; operating loss was $4.8M versus $3.0M.

• FY2026 net loss was $12.2M versus $7.7M in FY2025; accumulated deficit was $25.4M at June 30, 2026.

• At June 30, 2026, cash and USDC totaled $4.9M and current liabilities were $31.9M; the Business Combination provided ~$11.8M cash and converted ~$29.1M of SAFE liabilities to equity.

• Deployed GPU utilization was 91.0% in FY2026 versus 92.4% in FY2025; average customer-contracted GPU capacity was 1,050.8 GPUs versus 754.7 GPUs.

Use of proceeds and capital allocation

• Selling holders receive resale proceeds; no cash proceeds flow to the issuer from those sales, apart from cash received if registered warrants are exercised.

• Net cash proceeds from warrant exercises are intended for general corporate purposes; management has broad discretion and holders are not required to exercise.