Dell is introducing the PowerEdge XE7740 server, featuring Intel Gaudi 3
PCIe accelerators, to meet the need for efficient, scalable, and cost-effective
AI infrastructure.
The PowerEdge XE7740 is designed to offer powerful AI acceleration,
flexibility, and performance, according to Dell. The 4U PowerEdge XE7740
delivers the ease of integration of the PowerEdge R-Series with the performance
and scalability of PowerEdge XE-Series, advanced networking and powerful
accelerator configurations to meet the growing needs of AI development.
Whether you're fine-tuning AI models for specific business workflows or
running high-performance inferencing, the XE7740 simplifies deployment,
improves operations, and sets new benchmarks for AI price versus performance,
the company said. Now enterprise business and industries including financial
services, healthcare, manufacturing, telecommunications, and retail have even
more freedom to choose right-sized AI infrastructure.
The XE7740 is highly flexible, accommodating up to eight double-wide or
PCIe accelerators. With configurations supporting up to 8 individual Gaudi 3
accelerators, the XE7740 optionally offers two groups of 4-way bridged
accelerators.
4-way Gaudi 3 PCIe accelerator bridging integrates 900 GB/s combined
throughput for accelerator-to-accelerator communication over 18x 200GbE
interfaces using Remote Direct Memory Access (RDMA) over Converged Ethernet
(RoCE v2). Accelerator-to-accelerator
bridging is crucial for larger-scale AI models and larger memory spaces,
ensuring workflow scalability and flexibility, according to Dell.
This server isn’t just about raw computing power; it delivers flexible
networking with up to a 1:1 accelerator to Network Interface Card (NIC) ratio
through eight full-height PCIe slots and an integrated OCP networking module.
The XE7740 is purpose-built for enterprise AI and seamless integration
into existing air-cooled racks, thanks to innovative Dell Smart Cooling.
Powered by up to 8 Gaudi 3 PCIe accelerators, it supports seamless AI
workflow integration with optimization for popular models such as Llama4,
Llama3, Deepseek, Phi4, Qwen3, Falcon3, and more.
Alongside a diverse catalog of existing models, Gaudi 3 allows
enterprises to accelerate AI adoption effortlessly. Its efficient design
ensures high performance without costly power and cooling upgrades, making it
ideal for organizations looking to scale AI workloads reliably and efficiently,
Dell said.
Dell PowerEdge XE7740 is tailored for various AI workloads, including but
not limited to:
• Large Language Model (LLM)
inferencing and fine-tuning
• Multimodal AI
applications, such as image and speech recognition
• Data-intensive healthcare
analysis and genome sequencing
• Financial services fraud
detection and risk mitigation
• Real-time personalization
in retail and e-commerce
The Intel Gaudi 3 PCIe accelerators integrate smoothly with PyTorch and
platforms such as Hugging Face, enabling organizations to fine-tune models or
infer complex AI models efficiently.
By reducing unpredictable pay-as-you-go accelerator utilization costs and
cloud egress charges, the XE7740 empowers businesses to streamline on-premises
AI projects without compromising performance or data security, Dell said.
Leave A Comment