델 테크놀로지스(Dell Technologies)가 엔비디아와의 협력을 확대하며, 엔비디아 블랙웰 아키텍처 기반의 서버 신제품 ‘델 파워엣지 XE9680L’를 선보였다. 이 제품은 컴팩트한 폼 팩터에서 빠른 처리 성능을 제공하는 다이렉트 리퀴드 냉각 방식을 채용했고, 8개 엔비디아 블랙웰 텐서코어 GPU를 탑재해 기업들의 AI 활용 능력을 비약적으로 향상 시킬 것으로 기대된다.
New Server and Edge Solutions/Workstations Unveiled
Dell PowerEdge XE9680L with NVIDIA Blackwell
Dell Technologies has expanded its collaboration with NVIDIA and introduced a new server product based on NVIDIA's Blackwell architecture, the Dell PowerEdge XE9680L. This product features direct liquid cooling that delivers fast processing performance in a compact form factor and is equipped with eight NVIDIA Blackwell Tensor Core GPUs, which is expected to dramatically improve companies' AI utilization capabilities.
Dell announced on the 22nd that it is adding new servers, edges, workstations, solutions, and services to the 'Dell AI Factory with NVIDIA' to support rapid AI adoption.
'Dell AI Factory powered by NVIDIA' is an offering that integrates Dell's leading AI portfolio with the 'NVIDIA AI Enterprise' software platform and is based on technologies such as NVIDIA Tensor Core GPUs, NVIDIA Spectrum-X Ethernet networking fabric and NVIDIA Bluefield DPU.
Customers can purchase customized integrations or pre-validated full-stack solutions based on their needs to address AI use cases such as augmented search generation (RAG), model training, and inference.
The new product, 'Dell PowerEdge XE9680L', provides high performance by mounting eight NVIDIA Blackwell architecture GPUs in a smaller 4U form factor than the existing model.
This new product delivers the highest rack-scale density of any server with NVIDIA GPUs in an industry-standard x86 rack, delivering 33 percent more GPU density per node than previous models.
It also features 20% more 5th generation PCIe slots and doubles the north/south network expansion capacity.
In particular, the cooling capacity of the CPU and GPU has been increased with the 'direct liquid cooling' technology, improving overall efficiency.
Designed for easy serviceability, it is shipped fully configured from the factory for easy rack-scale deployment and field installation.
Dell accelerates time to value for large-scale GPU acceleration deployments by delivering the industry’s most dense and energy-efficient rack-scale turnkey solutions. Dell and NVIDIA will offer a variety of options in the future, including an air-cooled design that supports 64 GPUs in a single rack, or a liquid-cooled version that supports 72 NVIDIA Blackwell GPUs in a single rack.
'Dell NativeEdge' is the first edge orchestration platform that automates the deployment of 'NVIDIA AI Enterprise' software, allowing developers and IT operators to easily build AI applications and solutions at the edge.
A wide range of companies across manufacturing and retail industries can now quickly and accurately analyze edge data with new Dell NativeEdge blueprints that include NVIDIA Metropolis video analytics, NVIDIA Riva speech and translation capabilities, and NVIDIA NIM inference microservices.
The newly released 'Dell Generative AI Solution for Digital Assistants' is based on a full-stack solution from Dell and NVIDIA, helping users quickly build digital assistants to experience personalized self-service.
The accompanying Implementation Services for Digital Assistants will help you design, plan, implement, test and scale digital assistant solutions.
Dell AI Factory powered by NVIDIA, a full-stack automation solution designed through joint engineering between the two companies, helps to quickly build AI environments for various use cases. It accelerates time-to-value by up to 85% and, when used with NVIDIA Inferencing Microservices (NIM), dramatically reduces the overall time from delivery to execution of inference jobs.
'Dell Accelerator Services for RAG on Precision AI Workstations' shortens AI development cycles and improves AI application performance. RAG on Dell Precision Workstations equipped with the NVIDIA AI Workbench toolkit enables rapid prototyping in a safe environment by utilizing custom large-scale language models.
Kim Kyung-jin, General Manager of Dell Technologies Korea, said, “IT simplification is emerging as a key task for AI. As AI workloads increase, we face more issues such as data quality, security, consistency, and various selection options, so only by applying proven infrastructure and software can we effectively manage risks and costs. The collaboration between Dell and NVIDIA will be an important milestone in simplifying AI workloads.”