Research Finder
Find by Keyword
Supermicro Q4 FY2026: AI Infrastructure Demand Runs Into Deployment Constraints
Power and networking readiness affected Q4 shipments as Supermicro expands its rack-scale portfolio deeper into cooling, networking, and data center infrastructure.
08/12/2026
By the Numbers
- Q4 revenue of $11.1 billion, up 93% year over year and near the low end of the $11.0 billion to $12.5 billion guidance range
- Q4 non-GAAP gross margin of 17.6%, up 750 basis points sequentially from 10.1% in Q3
- FY2026 revenue of $39.1 billion, up 78%, with non-GAAP gross margin of 10.9% compared with 11.2% in FY2025
- FY2026 cash flow used in operations of $6.8 billion, compared with $1.7 billion provided in FY2025
- Cash conversion cycle of 149 days, up from 106 days in Q3
- Q1 FY2027 guidance of $14.5 billion to $15.5 billion in net sales with non-GAAP gross margin of 10.4% to 10.8%
- FY2027 revenue guidance of $65 billion to $72 billion
Key Highlights
- Customer readiness affected Q4 revenue, with power availability and networking among the constraints cited by Supermicro. More than $60 billion in new orders during the quarter highlights the gap between AI infrastructure demand and the rate at which customers can deploy it.
- AI solutions contributed approximately 60% of Q4 revenue, down from more than 80% in Q3 as several large AI projects moved into subsequent quarters. Supermicro expects AI-related solutions to return above 80% based on its backlog.
- Enterprise and Channel revenue reached $5.6 billion, or 50% of Q4 revenue, and grew 172% year over year. OEM appliance and large data center revenue contributed the other 50%.
- Supermicro plans manufacturing capacity of more than 6,000 racks per month, including more than 3,000 direct liquid-cooled racks, while expanding DCBBS into optical photonics and networking.
- Q4 non-GAAP gross margin reached 17.6%, but the company expects 10.4% to 10.8% in Q1 as AI revenue returns above 80%. Product and customer mix remain important to the economics of rack-scale AI.
The News
Supermicro reported fourth quarter and full fiscal year 2026 results on August 11, 2026. Q4 revenue increased 93% year over year to $11.1 billion and landed near the low end of guidance, while non-GAAP gross margin reached 17.6%. FY2026 revenue increased 78% to $39.1 billion. The company guided Q1 FY2027 revenue to $14.5 billion to $15.5 billion and FY2027 revenue to $65 billion to $72 billion. For more Q4 details, see the Supermicro investor site.
Analyst Take
Supermicro received more than $60 billion in new orders during Q4, with approximately 70% tied to GPU-based AI and the remaining 30% to CPU-based, edge, and agentic applications. Revenue still landed near the low end of guidance as customer readiness, including power availability and networking, delayed several large projects into subsequent quarters.
Ordering AI infrastructure and putting it into production are different milestones. Dense GPU systems require facilities capable of supporting their power, cooling, and connectivity requirements. Supermicro's Q4 results provide evidence that those dependencies can affect deployment schedules even when underlying demand remains strong.
The delays also changed the Q4 revenue mix. AI solutions represented approximately 60% of revenue, down from more than 80% in Q3, while Enterprise and Channel increased to 50% of the quarter. Supermicro said it expects AI revenue to return above 80% based on its backlog.
That mix contributed to an unusual margin result. Favorable customer and product mix accounted for approximately 75% of the 750 basis point sequential improvement in non-GAAP gross margin. Lower tariff costs and inventory reserves accounted for the remainder and were characterized as non-recurring. Q1 non-GAAP gross margin guidance of 10.4% to 10.8% accompanies the expected return to a higher concentration of AI revenue.
Supermicro is expanding the infrastructure surrounding the accelerator
Supermicro's DCBBS strategy extends rack-scale systems into cooling, power, networking, management software, and deployment services. The company plans capacity for more than 3,000 direct liquid-cooled racks per month during FY2027, with 240kW racks expected to reach volume and 500kW identified as the next step.
The HyperFRAME Research Lens: State of the Enterprise Infrastructure & Operations (1H 2026) found that 84% of enterprise I&O leaders agree AI deployment has consumed more budget and operational resources than originally planned. DCBBS helps position Supermicro to absorb part of that burden by delivering cooling, power, networking, and management software as integrated content within the rack. A new 32-acre Silicon Valley DCBBS campus will add optical photonics, networking, and data center-scale manufacturing, alongside investment in high-speed switch optical technology.
That investment aligns with a constraint visible in Q4. Networking readiness affected when some customers could accept systems. Expanding DCBBS further into networking gives Supermicro an opportunity to integrate more of the infrastructure required to deploy those systems.
Supermicro previously targeted DCBBS at approximately 20% of gross profit and now says it is approaching that level. FY2026 non-GAAP gross margin still declined 30 basis points to 10.9%. As accelerator platforms become available through multiple system vendors, system design, integration, cooling, networking, and deployment become important areas for differentiation and value capture.
Customer mix shows two infrastructure markets
Supermicro's growth remains concentrated among large infrastructure buyers. One large data center and CSP customer represented 28% of FY2026 revenue, while OEM appliance and large data center customers accounted for 69% and grew 104%. Nine customers generated more than $1 billion each during FY2026, up from four in FY2025. That concentration gives Supermicro significant exposure to the timing of individual AI infrastructure deployments, as Q4 demonstrated when several large projects moved into subsequent quarters.
Enterprise and Channel provide a different growth path. The segment represented 31% of FY2026 revenue and grew 39%, then reached $5.6 billion and 50% of Q4 revenue as customers upgraded compute, storage, and networking infrastructure. The stronger enterprise mix also contributed to the quarter's higher gross margin.
The two customer sets therefore affect more than revenue mix. Large AI infrastructure customers drive rack-scale volume but can create greater quarterly variability as projects depend on data center readiness. Enterprise customers broaden the opportunity across more deployment environments and can contribute a different margin profile. Supermicro's $65 billion to $72 billion FY2027 target depends on scaling both while managing the different economics and deployment cycles behind them.
That scale also places pressure on working capital. Inventory ended FY2026 at $12.9 billion, up from $4.7 billion a year earlier, while the cash conversion cycle increased to 149 days. Supermicro expects improved contract terms in its backlog to help fund growth from the balance sheet, but acknowledged that revenue materially above its guidance range could require additional capital. Converting demand into revenue therefore depends on both customer deployment readiness and Supermicro's ability to finance the infrastructure required to support that growth.
One additional risk remains unresolved. The Board's independent review of certain transactions related to export-control issues remains open, with an update expected shortly. Supermicro indicated that the outcome could affect forecasts, preliminary results, and prior-period results.
What Was Announced
Supermicro guided Q1 FY2027 net sales to $14.5 billion to $15.5 billion, with non-GAAP gross margin of 10.4% to 10.8% and non-GAAP diluted EPS of $1.01 to $1.10. Full-year FY2027 revenue guidance is $65 billion to $72 billion.
The company plans manufacturing capacity exceeding 6,000 racks per month during FY2027, including more than 3,000 direct liquid-cooled racks and approximately 3,000 air-cooled racks. Supermicro expects 240kW racks to reach volume during FY2027 and identified 500kW as the next step. The company reported more than 85MW plus an additional 10MW of power capacity worldwide to support manufacturing and testing.
Supermicro is expanding DCBBS around rack-scale infrastructure. The portfolio includes rack-level cooling and power, 400G and 800G networking, management software, and deployment services. Cooling capabilities extend from in-rack CDUs supporting up to 250kW to megawatt-scale cooling infrastructure.
A new 32-acre Silicon Valley DCBBS campus will add optical photonics, networking, and data center-scale manufacturing. The expansion brings Supermicro's U.S. footprint to nearly four million square feet, with additional global manufacturing expansion planned.
The company is also adding a proactive service model covering failed-unit repair and maintenance, with additional management software capabilities expected next quarter.
The FY2027 platform roadmap spans NVIDIA Vera CPU and Vera Rubin, GB300 NVL72, HGX B300 and RTX 6000 Pro, along with AMD MI350 and MI355. Supermicro also highlighted co-design work with NVIDIA, AMD, Intel, and Arm.
Looking Ahead
Supermicro enters FY2027 with demand ahead of its current revenue run rate. Q4 showed that converting that demand into deployed AI capacity depends on customer readiness, including power and networking. Expanding DCBBS gives Supermicro an opportunity to address more of those dependencies within its rack-scale architecture. Direct liquid cooling will be an important measure of that strategy. We will be watching how quickly DLC scales and how much of the surrounding infrastructure Supermicro delivers as part of the rack.
The data path remains another area to watch. Dense GPU infrastructure depends on moving data fast enough to keep expensive accelerators productive, yet Supermicro provided limited discussion of storage, data movement, or utilization. As its role expands around the GPU, greater visibility into the data path would provide another measure of its rack-scale architecture.
The economics will show how much value Supermicro captures from that expansion. Q1 margin guidance returns closer to FY2026 levels as AI solutions move back above 80% of revenue. FY2027 will show whether greater DCBBS content, deployment scale, and integration can improve margins as competition in rack-scale AI grows.
Don Gentile | Analyst-in-Residence -- Storage & Data Resiliency
Don Gentile brings three decades of experience turning complex enterprise technologies into clear, differentiated narratives that drive competitive relevance and market leadership. He has helped shape iconic infrastructure platforms including IBM z16 and z17 mainframes, HPE ProLiant servers, and HPE GreenLake — guiding strategies that connect technology innovation with customer needs and fast-moving market dynamics.
His current focus spans flash storage, storage area networking, hyperconverged infrastructure (HCI), software-defined storage (SDS), hybrid cloud storage, Ceph/open source, cyber resiliency, and emerging models for integrating AI workloads across storage and compute. By applying deep knowledge of infrastructure technologies with proven skills in positioning, content strategy, and thought leadership, Don helps vendors sharpen their story, differentiate their offerings, and achieve stronger competitive standing across business, media, and technical audiences.