NVIDIA, alongside top global computer manufacturers, has introduced a range of systems featuring the NVIDIA Blackwell architecture. These systems, equipped with Grace CPUs and NVIDIA networking, are designed to drive the next wave of generative AI advancements in AI factories and data centers.
NVIDIA, alongside top global computer manufacturers, has introduced a range of systems featuring the NVIDIA Blackwell architecture. These systems, equipped with Grace CPUs and NVIDIA networking, are designed to drive the next wave of generative AI advancements in AI factories and data centers.
NVIDIA Blackwell
During his keynote at COMPUTEX, NVIDIA founder and CEO Jensen Huang announced that companies including ASRock Rack, ASUS, GIGABYTE, Ingrasys, Inventec, Pegatron, QCT, Supermicro, Wistron, and Wiwynn will deliver a variety of AI systems. These systems will range from cloud and on-premises solutions to embedded and edge AI systems, all powered by NVIDIA GPUs and networking technologies.
Huang emphasized AI’s transformative potential: “The next industrial revolution has begun.” He added companies and countries are transitioning from traditional data centers to accelerated computing, building AI factories to produce artificial intelligence.
Diverse System Configurations and Advanced Technologies
The new offerings will cater to a wide range of applications, with configurations varying from single to multi-GPU setups, x86 to Grace-based processors, and incorporating both air and liquid cooling technologies. The NVIDIA MGX modular reference design platform now supports NVIDIA Blackwell products to expedite the development of these diverse systems. This includes the new NVIDIA GB200 NVL2 platform, optimized for large language model inference, retrieval-augmented generation, and data processing.
The GB200 NVL2 platform, with its high-bandwidth memory and NVLink-C2C, interconnects, accelerates data processing by up to 18 times, and offers eight times better energy efficiency than x86 CPUs. This makes it ideal for data analytics, where companies invest heavily.
NVIDIA MGX: A Modular Reference Architecture
The NVIDIA MGX platform provides computer manufacturers with a flexible reference architecture to efficiently build over 100 distinct system configurations. Manufacturers can start with a basic system architecture and customize it with their choice of GPU, DPU, and CPU to address various workloads. This approach has significantly reduced development costs and time, enabling rapid deployment of advanced computing systems.
NVIDIA MGX
AMD and Intel support the MGX architecture and plan to deliver their CPU host processor module designs, including AMD’s next-generation Turin platform and Intel’s Xeon 6 processor. These reference designs help ensure consistent performance and design while reducing development time.
Ecosystem and Partner Collaborations
NVIDIA’s comprehensive partner ecosystem includes TSMC, the leading semiconductor manufacturer, and global electronics makers providing critical components for AI factories. Companies like Amphenol, Asia Vital Components, Cooler Master, Colder Products Company, Danfoss, Delta Electronics, and LITEON contribute innovations in server racks, power delivery, and cooling solutions.
NVIDIA Spectrum-X
New data center infrastructure can be rapidly developed and deployed to meet enterprise needs, enhanced by NVIDIA’s Blackwell technology, Quantum-2 or Quantum-X800 InfiniBand networking, Spectrum-X Ethernet networking, and BlueField-3 DPUs. Leading system makers such as Dell Technologies, Hewlett Packard Enterprise, and Lenovo will offer servers incorporating these advanced technologies. Additionally, enterprises can leverage the NVIDIA AI Enterprise software platform, which includes NVIDIA NIM inference microservices, to develop and run production-grade generative AI applications.
Engage with StorageReview
Newsletter | YouTube | Podcast iTunes/Spotify | Instagram | Twitter | TikTok | RSS Feed
