NVIDIA
NVIDIA Corporation designs and manufactures graphics processing units (GPUs) and system on chip units (SoCs) for the gaming and professional markets, as well as automotive, robotics, and other industries. They develop AI-powered software and hardware solutions. For data centres, NVIDIA supplies AI and high-performance computing (HPC) infrastructure. This includes DGX and HGX platforms, modular server systems (MGX), and OVX systems for AI workloads. They also provide networking solutions like InfiniBand and Ethernet, along with DPUs and networking software to accelerate data centre operations.
At a glance
Facilities tracked
1
Countries
1
About NVIDIA
NVIDIA Corporation is a data center company that specializes in developing high-performance graphics processing units (GPUs) and related technologies. The company's products are widely used in data centers for tasks such as artificial intelligence, machine learning, and high-performance computing. NVIDIA has a strong presence in the data center industry, with its GPUs powering many of the world's leading cloud computing platforms and research institutions.
NVIDIA is a publicly traded semiconductor company specializing in GPUs and AI chips critical for data centers, cloud computing, telecom, and AI infrastructure.
Company Overview
NVIDIA facilities
Every NVIDIA site with a page on DC Atlas.
Israel
- Nvidia Israel data centerYokne'am · Announced
Browse every operator in Israel
Recent Activity
LG Energy Solution joins NVIDIA DSX Ready program for AI data centre energy storage
LG Energy Solution has been selected as one of three global battery energy storage system (BESS) partners in NVIDIA's DSX Ready program, a qualification programme for AI factory design, construction and operations. The company's JF2 AC Link product, based on high-capacity LFP batteries, meets DSX Ready requirements for AC energy storage solutions and includes load-balancing and grid-forming control capabilities. LG Energy Solution plans to expand AI data centre energy storage business leveraging five North American ESS production facilities and system integration capabilities.
Nvidia invests additional $1.5 billion in SB Energy data-centre project
Det amerikanska halvledarbolaget Nvidia investerar ytterligare 1,5 miljarder dollar i energibolaget SB Energy, ägt av japanska Softbank, som planerar...
GreenSquare signs Sharon AI as first customer for SYD1 Sydney data centre
Australian data centre operator GreenSquare has signed a long-term contract with AI infrastructure provider Sharon AI for capacity at its SYD1 campus in Western Sydney. Sharon AI will deploy approximately 8,200 NVIDIA Blackwell Ultra GPUs at the site, with initial operations targeted for Q4 2026. The agreement anchors the first tranche of AI-ready capacity at SYD1, which is being developed as a 100MW-plus campus in Norwest Business Park through adaptive reuse of an existing building.
Crusoe raises $3.9bn for AI data center expansion
Crusoe raised $3.9 billion from investors including Nvidia for AI data center expansion.
Nvidia, Google and Anthropic launch alliance for flexible AI data centres
Nvidia announced the formation of the AI Energy Management Alliance (AEMA) with Google, Anthropic, Analog Devices, and 18 other companies from AI, semiconductor, and energy sectors. The alliance aims to redesign how AI data centres consume electricity, enabling them to adjust power usage based on grid conditions through load shifting, battery storage, and on-site generation. The initiative addresses electricity shortage as a major constraint on AI infrastructure expansion and proposes revising grid connection criteria to recognize data-centre demand flexibility rather than constant maximum power consumption.
Bitdeer AI signs 10-year agreement for 65.1 MW data center in Malaysia
Bitdeer AI has signed a 10-year data center services agreement for a 65.1 MW AI cloud data center facility (A202) in Johor Bahru, Malaysia, to be energized in Q3 2027. The facility will be located on the same campus as Bitdeer's existing 21.7 MW A201 facility, bringing total campus capacity to 86.8 MW when both are operational. The facility is designed for liquid-cooled, rack-scale NVIDIA GPU systems and will support GPU cloud services and data hosting.
Hardware
NVIDIA in the Hardware Index4 equipment classes · 9 models
Data Center Switches4
| Render | Model | Platform type | Switching capacity (as published) | Rack units |
|---|---|---|---|---|
 | NVIDIA Spectrum-4 SN5400 | fixed | Throughput 25.6 Tbps | 2 RU |
 | NVIDIA Spectrum-4 SN5600 | fixed | Throughput 51.2 Tbps | 2 RU |
 | NVIDIA Spectrum-4 SN5600D | fixed | Throughput 51.2 Tbps | 2 RU |
 | NVIDIA Spectrum-4 SN5610 | fixed | Throughput 51.2 Tbps | 2 RU |
GPU Servers2
| Render | Model | GPU model | GPU count | Form factor |
|---|---|---|---|---|
 | NVIDIA DGX B200 | B200 | 8 | eia rack mount |
 | NVIDIA DGX H100 | H100 | 8 | eia rack mount |
Optical Transceivers1
| Render | Model | Nominal data rate | Form factor (as published) | Maximum power |
|---|---|---|---|---|
 | NVIDIA MMS4X00-NM | 800 Gb/s | Twin-port OSFP, finned top | 17 W |
Rack-Scale AI Systems2
| Render | Model | Accelerator | Accelerators per rack | Cooling architecture |
|---|---|---|---|---|
 | NVIDIA DGX GB200 | NVIDIA Blackwell | 72 | dlc hybrid air liquid |
| NVIDIA DGX GB300 | NVIDIA Blackwell Ultra | 72 | dlc liquid share unstated |
Renders, not photographs. Product images are generated to DC Atlas art direction from each manufacturer’s own product image, so form, panel layout and controls follow the real machine. Vendor branding is not reproduced. Where a manufacturer publishes no image of an individual model, the render follows another unit in the same range.