Share on
NVIDIA Blackwell HGX B300 NVL8
1 Unit
$1,001,000MOQ:
1 Unit
Payment Term:
T/T
Trade Term:
FOB
Port Of Loading:
Hongkong, China, Cn (HKG/HKN)
Key Takeaways
Freshdi AI is analyzing this product...
NVIDIA Blackwell HGX B300 NVL8 AI System
800Gbps AI Training Server
High-Performance AI Supercomputer
Product Specifications
Boot Storage Capacity:
1.92tb
Total Cpu Cores:
128
Memory Module Capacity:
128gb
Primary Storage Total Capacity:
15.36tb
Total Storage Capacity:
19.2tb
Boot Storage Count:
2
Cpu Count:
2
Warranty Parts:
3 Years
Warranty Labor:
3 Years
Primary Storage Capacity:
3.84tb
Boot Storage Total Capacity:
3.84tb
Memory Module Count:
32
Primary Storage Count:
4
Memory Capacity:
4tb
Gpu Count:
8
Networking Bandwidth:
800gbps
Form Factor:
8u Rackmount
Electronic Component:
Active
Intended Use:
Ai Inference, Deep Learning Training, Generative Ai, Large Language Models Llms, Scientific Computing
Gpu Architecture:
Blackwell
Memory Type:
Ddr5-6400 Ecc Rdimm
Cpu Model:
Intel Xeon 6767p
Networking Interface:
Nvidia Connectx-8
Gpu Model:
Nvidia Hgx B300 Nvl8
Gpu Interconnect:
Nvlink
Boot Storage Model:
Samsung Pm9a3
Primary Storage Model:
Samsung Pm9d3a
System Model:
Supermicro Sys-822gs-nb3rt
4.3
United States
1 yrs
Piper Glen DR
Activity signals
Avg Response Time
≤ 24h
Response Rate
4%
Inquiry Open Rate
10%
Total inquiries received
294
Export Markets
Based on the information provided by the supplier.
China
East Asia
100%
Verification information
Verify business documents
By Freshdi Verification System
Verify work email
By Freshdi.com
Verify phone number
Buyer Verification Team
Verify account email
By Freshdi.com
Product descriptions from the supplier
AI Server Configuration – NVIDIA Blackwell HGX B300 NVL8
This configuration describes a high-end enterprise AI server built on the NVIDIA Blackwell HGX B300 NVL8 platform, designed for large-scale AI training, inference, and high-performance computing (HPC).
System Overview
Model: Supermicro SYS-822GS-NB3RT
* 8U rackmount AI server
* Optimized for NVIDIA HGX Blackwell GPU architecture
* Designed for hyperscale AI and data center deployments
GPU Configuration
* 8 × NVIDIA HGX B300 NVL8 GPUs
* Blackwell architecture
* High-speed NVLink interconnect between GPUs
* Suitable for:
* Large Language Models (LLMs)
* Generative AI
* Deep Learning Training
* AI Inference
* Scientific Computing
CPU Configuration
* 2 × Intel Xeon 6767P Processors
* 64 cores per CPU
* Total: 128 physical CPU cores
Memory
* 32 × 128GB DDR5-6400 ECC RDIMM
* Total Installed Memory: 4TB RAM
Storage
Primary NVMe Storage
* 4 × Samsung PM9D3a 3.84TB NVMe SSD
* Total: 15.36TB
Boot / Additional Storage
* 2 × Samsung PM9A3 1.92TB NVMe SSD
* Total: 3.84TB
Combined Storage Capacity
* Approximately 19.2TB NVMe SSD
Networking
* NVIDIA ConnectX-8 onboard networking
* Up to 800Gbps connectivity
* Designed for AI clusters and high-performance distributed computing
Support & Warranty
* 3-Year Parts Warranty
* 3-Year Labor Warranty
* NVIDIA Enterprise Support Services
User Reviews
No reviews yet
Be the first to share your experience - your feedback helps other buyers trade with confidence.
Related searches
You are viewing details for NVIDIA Blackwell HGX B300 NVL8. The information provided here is either posted directly by the supplier, who is responsible for its accuracy, or collected by Freshdi AI from publicly available sources. Freshdi strives to make product information as useful and reliable as possible, but we do not guarantee complete accuracy. If you notice any incorrect or misleading details, please use the Report Product option so our team can review and address it quickly.
$1,001,000
/ Unit
Min. order: 1 Unit
Yonhy Santos
Active
Response Rate: 4%
Avg Response Time: ≤ 24h
Year Established: 2023
Main Export Markets
