The KR4268-E2 is a 4U AI server optimized for high-performance computing, AI model training, and large-scale inference. Supporting 8 FHFL double-width GPUs with PCIe 5.0 bandwidth, it ensures maximum throughput and scalability for demanding AI applications.
Powered by dual AMD EPYC™ 9004 processors, the system provides exceptional compute power and memory bandwidth for heterogeneous workloads. With support for 24x 2.5” SSD drives (up to 16x NVMe U.2) and 11 PCIe 5.0 x16 expansion slots, the KR4268-E2 offers flexible configuration options to meet the needs of data-intensive applications.
The KR4268-E2 is ideal for enterprise AI deployments, HPC clusters, and research institutions, delivering enterprise-grade reliability, density, and scalability in a compact 4U chassis.
Main Application Areas:
-
AI Inference & Training: High-performance GPU computing for deep learning workloads
-
High-Performance Computing (HPC): Scientific simulations, modeling, and large-scale analytics
-
Generative AI & Research: Model fine-tuning and multi-GPU accelerated experiments
-
Data-Intensive Workloads: Big data pipelines, cloud AI, and multi-node computing
Price on request – daily updates apply. Please check availability before ordering.
.png "KR4268-E2 – 4U AI Server with 8 GPUs and PCIe 5.0")
.png "KR4268-E2 – 4U AI Server with 8 GPUs and PCIe 5.0")
.png "KR4268-E2 – 4U AI Server with 8 GPUs and PCIe 5.0")
.png "Preview: KR4268-E2 – 4U AI Server with 8 GPUs and PCIe 5.0")
.png "Preview: KR4268-E2 – 4U AI Server with 8 GPUs and PCIe 5.0")
.png "Preview: KR4268-E2 – 4U AI Server with 8 GPUs and PCIe 5.0")
.png "KR4268-E2 – 4U AI Server with 8 GPUs and PCIe 5.0")
.png "KR4268-E2 – 4U AI Server with 8 GPUs and PCIe 5.0")
.png "KR4268-E2 – 4U AI Server with 8 GPUs and PCIe 5.0")