Supermicro Reveals DCBBS® with New NVIDIA Vera Rubin NVL72, HGX Rubin NVL8, and Vera CPU Systems, Designed to Accelerate Customer Time-to-Market

Rhea-AI Summary

Supermicro (NASDAQ: SMCI) unveiled systems built for the NVIDIA Vera Rubin platform on March 16, 2026, using its DCBBS liquid-cooling stack to accelerate AI factory deployments.

Key specs include the Vera Rubin NVL72 rack (up to 3.6 Exaflops inference, 75TB fast memory, 1.6 PB/s HBM4 bandwidth), claims of up to 10x throughput per watt and one-tenth token cost vs Blackwell, a 2U HGX Rubin NVL8 supporting up to 72 Rubin GPUs per rack, and a 2U Vera CPU node with dual Vera CPUs and up to 6 RTX PRO 4500 GPUs. Supermicro also previewed a BlueField-4 powered context memory storage platform and announced continued availability of Blackwell-based systems for immediate deployment.

Positive

• Vera Rubin NVL72: 3.6 Exaflops inference capacity
• Vera Rubin NVL72: 75TB fast memory capacity
• 1.6 PB/s HBM4 bandwidth for NVL72
• HGX Rubin NVL8: up to 72 Rubin GPUs per rack
• Claimed up to 10x throughput per watt vs Blackwell

Negative

• Vera Rubin platforms will be fully liquid-cooled from this generation
• Next-generation Vera systems are in development, not immediately available

News Market Reaction – SMCI

-1.10%

-1.10% Session close to close

In the Mar 17 session, SMCI declined 1.10%, reflecting a mild negative market reaction.

Data tracked by StockTitan Argus on the day of publication.

Market Context

This announcement highlights SMCI’s push into NVIDIA Vera Rubin-based AI factories, with claims of u...

Analysis

This announcement highlights SMCI’s push into NVIDIA Vera Rubin-based AI factories, with claims of up to 10x throughput per watt, 3.6 Exaflops of inference, and dense liquid-cooled racks. Recent history shows both rallies and pullbacks around similar AI and server launches. Investors may focus on how quickly these systems reach volume deployment, customer adoption versus existing Blackwell platforms, and whether margins hold as next-generation infrastructure ramps.

Key Figures

Throughput efficiency: up to 10x throughput per watt Token cost: one-tenth the token cost Rack GPU density: 72 Rubin GPUs per rack +5 more

8 metrics

Throughput efficiency up to 10x throughput per watt Vera Rubin NVL72 vs NVIDIA Blackwell solutions

Token cost one-tenth the token cost Vera Rubin NVL72 vs NVIDIA Blackwell solutions

Rack GPU density 72 Rubin GPUs per rack 2U HGX Rubin NVL8 systems per rack

GPU count per node up to 6 RTX PRO 4500 GPUs NVIDIA Vera CPU 2U server

Inference performance up to 3.6 Exaflops Vera Rubin NVL72 rack-scale accelerator

Fast memory 75TB fast memory Vera Rubin NVL72 configuration

HBM4 bandwidth 1.6 PB/s HBM4 bandwidth Vera Rubin NVL72 supercomputer

HGX systems per rack 9 HGX Rubin NVL8 systems Per rack, supporting 72 Rubin GPUs total

Historical Context

5 past events · Latest: Mar 02 (Positive)

Pattern 5 events

Date	Event	Sentiment	24h Move	Catalyst
Mar 02	AI infrastructure update	Positive	-1.7%	Expanded support for AI-RAN and sovereign AI infrastructure with new systems.
Feb 26	Server product launch	Positive	-3.9%	Introduced high-density AMD EPYC 4005 MicroBlade platform for cloud and edge.
Feb 25	AI data platform	Positive	+7.9%	Launched CNode-X enterprise AI data platform with NVIDIA and VAST Data.
Feb 25	AI demand response test	Positive	+7.9%	Demonstrated GPU-based AI compute providing fast demand response for power markets.
Feb 13	Investor conferences	Neutral	+0.4%	Announced participation in multiple March investor conferences with 1x1 meetings.

24h Move is the share-price change in the day after each event; other market factors may also have contributed.

Pattern Detected

Recent AI and product announcements often moved the stock materially, with both strong gains and occasional selloffs on positive news.

Recent Company History

Over the last few weeks, SMCI has repeatedly highlighted new AI and high-density infrastructure offerings. AI-focused announcements on Feb 25 around an enterprise AI data platform and demand response capabilities saw +7.93% reactions, while a March AI-RAN and sovereign AI update on Mar 2 coincided with a -1.73% move. A dense AMD EPYC MicroBlade launch on Feb 26 saw a -3.93% reaction, showing that product news can trigger both rallies and pullbacks.

Key Terms

liquid-cooling, coolant distribution units (cdus), mixture-of-experts (moe), dpu, +4 more

8 terms

liquid-cooling technical

"built on the DCBBS liquid-cooling stack, targeting up to 10x throughput"

Liquid-cooling is a method of removing heat from electronic equipment, batteries or machinery by circulating a liquid (like water or a special coolant) that absorbs heat and carries it away, similar to how a car radiator keeps an engine from overheating. For investors, it matters because more effective cooling can improve performance, energy efficiency, reliability and lifetime of high‑value assets such as data centers, electric vehicles and industrial systems, which affects operating costs, maintenance needs and resale value.

coolant distribution units (cdus) technical

"includes in-rack and in-row components such as coolant distribution units (CDUs),"

Coolant Distribution Units (CDUs) are devices that manage and direct the flow of cooling liquids in large data centers or industrial systems. They help keep equipment from overheating by efficiently distributing coolant to where it's needed, ensuring smooth operation and preventing damage, much like a thermostat controls temperature in your home.

mixture-of-experts (moe) technical

"agentic reasoning, long-context AI, and Mixture-of-Experts (MoE) workloads are driving"

A mixture-of-experts (MoE) is a machine‑learning design that uses many specialist submodels and routes each task to the few experts best suited to handle it, rather than running one large model for everything. For investors, MoE can deliver faster, cheaper and more scalable AI performance—like calling a team of specialists instead of hiring one person who knows a little about everything—so it can lower operating costs and improve product competitiveness.

dpu technical

"Rubin GPU, Vera CPU, NVIDIA NVLink 6, NVIDIA ConnectX-9 SuperNIC, NVIDIA BlueField-4 DPU,"

Distributions per unit (DPU) is the amount of cash or income paid to each unit holder of a trust, real estate investment trust (REIT) or similar pooled investment for a given period. It tells investors how much cash income they received per unit, like getting a fixed slice of a pie for every share you own, and helps compare yield and judge whether the payout level is steady, growing or at risk.

pcie technical

"LPDDR5X memory subsystem and PCIe GPU acceleration in a space-efficient footprint."

PCIe (PCI Express) is a high-speed connection standard used inside computers and servers to link components like graphics cards, storage drives, and network adapters so they can send data quickly to each other. Investors care because faster, more efficient PCIe support can make a product more competitive—think of it as wider, faster highway lanes for data—which affects device performance, upgrade flexibility, manufacturing cost and customer demand.

lpddr5x technical

"It is a high-bandwidth LPDDR5X memory subsystem and PCIe GPU acceleration"

LPDDR5X is a modern, low-power type of volatile memory used mostly in smartphones, tablets and other energy-sensitive devices to store working data temporarily while apps run. Think of it as a faster, more efficient short-term memory for a device — its speed and lower power use can improve performance and battery life, so adoption trends affect makers of chips and devices and can signal competitiveness in hardware design.

kv cache technical

"tier that extends GPU KV cache capacity and serves long-context inference data"

A kv cache is a small, fast memory store that keeps recently used “keys” (identifiers) and their associated “values” (data) so a computer system can look them up instantly instead of re-calculating or re-fetching them. For investors, a kv cache matters because it can cut latency and computing costs and improve the responsiveness of services like algorithmic trading, market data feeds, or AI-driven analysis—similar to a clerk who remembers recent lookups so customers don’t wait.

rag technical

"data path offload that large-scale AI inference pipelines and RAG workloads require."

A RAG (red-amber-green) status is a simple color-coded system used in reports and dashboards to show the health, progress, or risk level of a project, metric, or business area. Think of it as a traffic light: green means on track, amber means caution or potential issues, and red means serious problems. Investors use RAG indicators to quickly spot emerging risks or improvements that could affect future performance and value.

AI-generated analysis. How Rhea-AI works. Not financial advice.

See more from StockTitan in Google Search and AI answers. Adds StockTitan as a preferred source · opens Google

Add on Google Not now

• Supermicro's NVIDIA Vera Rubin NVL72 and HGX Rubin NVL8 systems are built on the DCBBS liquid-cooling stack, targeting up to 10x throughput per watt and one-tenth the token cost, compared to NVIDIA Blackwell solutions.
• Supermicro's 2U HGX Rubin NVL8 system is the most flexible platform supporting NVIDIA Vera and next-generation x86 CPUs, scaling to 72 Rubin GPUs per rack, as well as a DCBBS Liquid-to-Air (L2A) Sidecar CDU option for data centers without liquid-cooling.
• Supermicro's new NVIDIA Vera CPU systems include a 2U server supporting up to 6 RTX PRO 4500 Blackwell Server Edition GPUs and a new AI storage system for context memory extension integrated with NVIDIA BlueField-4 DPU.

SAN JOSE, Calif., March 16, 2026 /PRNewswire/ -- Super Micro Computer, Inc. (NASDAQ: SMCI), a Total IT Solution Provider for Cloud Computing, AI/ML, Storage, and 5G/Edge, today unveiled its upcoming system portfolio powered by the NVIDIA Vera Rubin platform. As data centers transform into AI factories, producing intelligence at massive scale, agentic reasoning, long-context AI, and Mixture-of-Experts (MoE) workloads are driving demand for an entirely new class of compute and storage infrastructure. Supermicro's NVIDIA Vera Rubin NVL72, NVIDIA HGX Rubin NVL8, NVIDIA Vera CPU systems are being designed and built with Supermicro's Data Center Building Block Solutions (DCBBS) advanced liquid-cooling technology stack to accelerate time-to-market for customers.

"We are entering a new era where every organization requires an AI factory to win in the marketplace, as the demand for inference workloads is reshaping what data center infrastructure must deliver," said Charles Liang, president and CEO of Supermicro. "Supermicro's DCBBS technology stack is being engineered to empower upcoming NVIDIA Vera Rubin NVL72, HGX Rubin NVL8, and Vera CPU systems to give our customers a fast, clear path to deploying next-generation AI factories, at scale. We are excited to provide an early look at these solutions as a testament to being first to market with the infrastructure that will power the next frontier of AI."

For more information visit NVIDIA Vera Rubin | Supermicro.

Supermicro DCBBS for NVIDIA Vera Rubin and Rubin Platforms

Delivering AI factory performance at scale requires much more than just compute — It demands power, cooling, and networking infrastructure that performs seamlessly. Supermicro's modular DCBBS approach enables data center operators to deploy validated, pre-engineered rack solutions rather than custom-building infrastructure for each project — reducing time-to-online, minimizing integration risk, and lowering total cost of ownership across AI factory deployments of any scale.

Supermicro's DCBBS are engineered specifically to meet the evolving thermal, power, and networking demands needed to enable rapid and robust deployment of upcoming NVIDIA Vera Rubin NVL72, NVIDIA HGX Rubin, and NVIDIA Vera CPU infrastructure. To meet the needs of Vera Rubin platforms, which will be fully liquid-cooled from this generation forward, DCBBS includes a full suite of validated liquid-cooling infrastructure. This expansion includes in-rack and in-row components such as coolant distribution units (CDUs), manifolds, and liquid-to-air sidecar. Also included are infrastructure solutions such as cooling towers and cabling design and implementation services — designed to integrate seamlessly with Supermicro's next-generation system portfolio.

Supermicro NVIDIA Vera Rubin NVL72 SuperCluster

Supermicro is engineering its NVIDIA Vera Rubin NVL72 with new DCBBS liquid-cooling components to fully support the power and thermal envelope at rack and cluster scale. This includes the manufacturing of optimized NVIDIA MGX racks, in-rack or in-row CDU, RDHx and L2A sidecar to streamline production and deployment of the rack-scale AI supercomputer at scale. The Vera Rubin NVL72 operates as a single rack-scale accelerator, unifying six co-designed chips — Rubin GPU, Vera CPU, NVIDIA NVLink 6, NVIDIA ConnectX-9 SuperNIC, NVIDIA BlueField-4 DPU, and NVIDIA Spectrum-X Ethernet — to deliver up to 3.6 Exaflops of inference, 75TB of fast memory, and 1.6 PB/s of HBM4 bandwidth, targeting up to 10x the throughput per watt and one-tenth the token cost compared to NVIDIA Blackwell.

NVIDIA HGX Rubin NVL8 System

The 2U HGX Rubin NVL8 system provides the densest and most flexible HGX platform — and the first HGX platform to offer greater flexibility in CPU selections including NVIDIA Vera CPUs alongside next-generation AMD and Intel x86 processors. Built on the NVIDIA MGX rack architecture with Supermicro's blind mate busbar and manifold for tool-free rack integration, it gives customers the freedom to pair eight Rubin GPUs with the CPU platform that best fits their workload and software stack.

The design supports 9 HGX Rubin NVL8 systems per rack — up to 72 Rubin GPUs total — for large-scale AI training, inference, and accelerated HPC. DCBBS provides in-rack CDU, in-row CDU, RDHx and an optional Liquid-to-air (L2A) sidecar for customers deploying in liquid-cooled or air-cooled data center environments.

NVIDIA Vera CPU System with RTX PRO

Supermicro's Vera CPU next-generation agentic AI system is being engineered as a versatile AI compute node for organizations targeting next-generation agentic AI deployments. The system features dual NVIDIA Vera CPUs supporting up to 6 RTX PRO 4500 Blackwell Server Edition GPUs in a compact 2U chassis — delivering the compute density and energy efficiency that enterprise AI inference, agentic workloads, visualization, and adding accelerated computing to all enterprise workloads. It is a high-bandwidth LPDDR5X memory subsystem and PCIe GPU acceleration in a space-efficient footprint.

NVIDIA BlueField-4 STX Context Memory Storage Platform

Supermicro's upcoming Context Memory Storage Platform (CMX) introduces a new class of AI-native storage for context memory — architected as an intelligent pod-level context memory storage tier that extends GPU KV cache capacity and serves long-context inference data at the throughput that Vera Rubin NVL72 super pod clusters demand. Powered by NVIDIA BlueField-4 processor, NVIDIA Vera CPUs, NVIDIA ConnectX-9 SuperNICs, Spectrum-X Ethernet NVIDIA DOCA, and NVIDIA Dynamo, the system provides the high-bandwidth, low-latency fabric and intelligent data path offload that large-scale AI inference pipelines and RAG workloads require.

Supermicro NVIDIA Blackwell Solutions — Available Now

With next-generation systems in rapid development, Supermicro's current portfolio of NVIDIA Blackwell-based systems is in full production and available for immediate deployment through Supermicro's US and global manufacturing capacity, enabling customers to build and scale production AI infrastructure today. Supermicro is investing across both its current Blackwell lineup and its next-generation systems to ensure customers have the right platform at every stage of this transformation.

Visit Supermicro at GTC San Jose 2026

Supermicro will be unveiling early previews of its Vera Rubin platform systems alongside its current production Blackwell portfolio. Supermicro experts will be available to discuss current procurement options, roadmap planning, and deployment timelines for both near-term and next-generation AI infrastructure in Supermicro booth #1113.

About Super Micro Computer, Inc.

Supermicro (NASDAQ: SMCI) is a global leader in Application-Optimized Total IT Solutions. Founded and operating in San Jose, California, Supermicro is committed to delivering first-to-market innovation for Enterprise, Cloud, AI, and 5G Telco/Edge IT Infrastructure. We are a Total IT Solutions provider with server, AI, storage, IoT, switch systems, software, and support services. Supermicro's motherboard, power, and chassis design expertise further enables our development and production, enabling next-generation innovation from cloud to edge for our global customers. Our products are designed and manufactured in-house (in the US, Taiwan, and the Netherlands), leveraging global operations for scale and efficiency and optimized to improve TCO and reduce environmental impact (Green Computing). The award-winning portfolio of Server Building Block Solutions® allows customers to optimize for their exact workload and application by selecting from a broad family of systems built from our flexible and reusable building blocks that support a comprehensive set of form factors, processors, memory, GPUs, storage, networking, power, and cooling solutions (air-conditioned, free air cooling or liquid cooling).

Supermicro, Server Building Block Solutions, and We Keep IT Green are trademarks and/or registered trademarks of Super Micro Computer, Inc.

All other brands, names, and trademarks are the property of their respective owners.

View original content to download multimedia:https://www.prnewswire.com/news-releases/supermicro-reveals-dcbbs-with-new-nvidia-vera-rubin-nvl72-hgx-rubin-nvl8-and-vera-cpu-systems-designed-to-accelerate-customer-time-to-market-302714873.html

SOURCE Super Micro Computer, Inc.

FAQ

What did Supermicro (SMCI) announce on March 16, 2026 about Vera Rubin NVL72?

Supermicro announced the Vera Rubin NVL72 rack with up to 3.6 Exaflops inference performance. According to the company, NVL72 unifies Rubin GPUs, Vera CPUs, NVLink 6, BlueField-4 DPU and Spectrum-X to deliver high memory and bandwidth for large-scale inference.

How many Rubin GPUs per rack does the HGX Rubin NVL8 support for SMCI customers?

The HGX Rubin NVL8 design supports up to 72 Rubin GPUs per rack using nine 2U systems per rack. According to the company, the platform allows flexible CPU choices and tool-free rack integration for dense AI training and inference.

What energy-efficiency and cost claims did Supermicro make for Vera Rubin versus Blackwell?

Supermicro claims Vera Rubin can achieve up to 10x throughput per watt and one-tenth token cost versus Blackwell. According to the company, these targets reflect co-designed hardware and DCBBS liquid-cooling at rack and cluster scale.

Does SMCI require liquid cooling for the Vera Rubin platforms and what are options?

Vera Rubin platforms are designed to be fully liquid-cooled from this generation forward, with a Liquid-to-Air sidecar available. According to the company, DCBBS offers in-rack/in-row CDUs and an L2A sidecar for air-cooled data centers.

What are the specs of the Supermicro Vera CPU 2U system announced by SMCI?

The 2U Vera CPU system supports dual NVIDIA Vera CPUs and up to 6 RTX PRO 4500 Blackwell Server Edition GPUs. According to the company, it targets agentic AI inference, visualization, and enterprise accelerated workloads with high-bandwidth LPDDR5X memory.

When can customers deploy Supermicro Blackwell and upcoming Vera systems from SMCI?

Blackwell-based Supermicro systems are available now for immediate deployment through global manufacturing. According to the company, next-generation Vera systems are being previewed and remain in development for future delivery.