“Leather Case iPhone 11 Pro” has been added to your cart. View cart

-4%

Quanta QCT QuantaGrid D75L-3U/S8L HGX B300 NVL8 270GB Partner-Cooled (8) Intel Granite Rapids SP (2) AI Graphics Card GPU Server - Image 1

Quanta QCT QuantaGrid D75L-3U/S8L HGX B300 NVL8 270GB Partner-Cooled (8) Intel Granite Rapids SP (2) AI Graphics Card GPU Server - Image 2

Click to enlarge

Quanta QCT QuantaGrid D75L-3U/S8L HGX B300 NVL8 270GB Partner-Cooled (8) Intel Granite Rapids SP (2) AI Graphics Card GPU Server

QuantaGrid D75L-3U

Maximum GPU Density for Your AI Factories

3U chassis compatible with 19” MGX 1.1 Rack, 1400A busbar with power shelves, DC-SCM Support
Dual Intel® Xeon® 6700 series processors with P-cores, up to 350W TDP
NVIDIA HGX B300 8-GPU with 5th Generation NVLink® 1.8TB/s, 2.3TB of HBM3e GPU memory per system
8x NVIDIA ConnectX®-8 SuperNICs, up to 800Gb/s, up to three NICs for North/South traffic

Original price was: $1,969,900.00.Current price is: $1,899,900.00.

Add to compare

Add to wishlist

Guaranteed Safe Checkout

Features & Compatibility

The Quanta D75L-3U is a premier rack-level AI powerhouse, purpose-built for the next generation of AI Factories. By leveraging the NVIDIA MGX™ architecture and advanced liquid-cooling technology, the D75L-3U unlocks the full potential of the NVIDIA Blackwell Ultra platform. Its highly efficient, ultra-dense 3U design enables an industry-leading density of 72 GPUs per MGX rack, providing the foundation for exascale computing.

Pioneering Performance with Blackwell Ultra

The Power of Blackwell: Seamlessly supports the NVIDIA HGX™ B300 platform, powered by NVIDIA Blackwell Ultra GPUs. It is engineered to master the world’s most complex Generative AI workloads and Large Language Models (LLM).
800G Fabric with ConnectX-8: Features onboard ConnectX-8 to enable blazing-fast 800G East-West data transfer. This massive bandwidth is critical for building massive-scale GPU clusters, ensuring zero-bottleneck communication across the fabric.

Unmatched Compute Density

Optimized Space Efficiency: The D75L-3U integrates the immense power of the HGX™ B300 into a compact 3U chassis, significantly reducing the physical data center footprint compared to traditional 8U or 10U air-cooled alternatives.
Extreme Rack-Level Scaling: Purpose-built for the NVIDIA MGX™ rack, the system accommodates up to 9 nodes per rack, achieving an optimal density of 72 GPUs.

Leveraging the NVIDIA MGX™ Ecosystem

Accelerate Time-to-Market: By adhering to NVIDIA MGX™ ecosystem, we ensure a stable and reliable supply of critical components—including the rack, power distribution, and cooling manifolds—drastically reducing time-to-market for large-scale deployments.
Serviceability & Standardized Compatibility: The D75L-3U integrates seamlessly with the MGX ecosystem, providing superior serviceability and ensuring your infrastructure remains flexible to evolve with future AI hardware generations.

Processor
Processor Family	Intel® Xeon® 6 processors
Max. TDP Support	350W
Number of Processors	2 Processors

Form Factor
Form Factor	3U

Dimensions
W x H x D (mm)	438 x 131.7 x 864.5 mm

Storage
Default Configuration	(8) hot-swappable E1.S PCIe Gen5 SSD

Memory
Total Slots	(32) DDR5 RDIMM up to 6,400 MHz, (16) MRDIMM up to 8,000 MHz

Expansion Slot
Default Configuration	(3) FHHL SW PCIe Gen5 x16 slots

Fan
Fan	(8) 6056 fans

Operating Environment
Operating Environment	Operating temperature: 5°C to 35°C Non-operating temperature: -40°C to 70°C Operating relative humidity: 20% to 85%RH Non-operating relative humidity: 10% to 95%RH

Additional information

Customer Reviews

Reviews

There are no reviews yet.

Be the first to review “Quanta QCT QuantaGrid D75L-3U/S8L HGX B300 NVL8 270GB Partner-Cooled (8) Intel Granite Rapids SP (2) AI Graphics Card GPU Server”