
TITAN G640R-G7
AI & GPU Server
- 4U
- Dual Intel® Xeon® 6700P-series, 6500P-series, and 6700E-series Processors
- Supports 10 dual-slot GPUs with 600W TDP each
- Up to 4TB DDR5-6400 MT/s memory
- 18 x 2.5" NVMe hot-swap and 16 x 2.5" SATA/SAS hot-swap drive bays
- ASPEED AST2600 management
- 3200W (3+1) 80 PLUS Titanium Power Supply
- Perfect for AI Training, AI Inference, Data Analytics and more

Product Overview
The CIARA TITAN G640R-G7 is a 4U AI & GPU server built for intensive AI, HPC, and data-processing workloads. It is powered by dual Intel® Xeon® 6th Gen processors (up to 400 W TDP each), featuring Intel AMX and DL Boost technologies, to accelerate AI inference and data processing directly on the CPU.
Supporting up to ten NVIDIA GPUs, including the RTX Pro™ 6000 SE, H200 NVL, H100 NVL, and L40s, it delivers the parallel compute performance required for training, inference, and large-scale simulations.
Its 8-channel DDR5 architecture with 32 DIMM slots provides high bandwidth for smooth data exchange between CPUs, GPUs, and storage. PCIe Gen5 connectivity and NVMe storage deliver fast data access and reduce latency, enabling faster performance across demanding workloads.


Key Features and Benefits
High- Performance Intel Compute
The dual Intel® Xeon® 6th Gen processors deliver strong multi-core performance for AI, HPC, and analytics workloads. Built-in Intel AMX and DL Boost technologies accelerate AI inference and data processing on the CPU. These technologies work alongside GPUs to enhance overall performance.
Multi-GPU Acceleration
The CIARA TITAN G640R-G7 supports up to ten NVIDIA GPUs (600 W TDP each), including the RTX Pro™ 6000 SE, H200 NVL, H100 NVL, and L40s, enabling the computing performance required for deep learning training, generative AI, simulation workloads, and large-scale inference. This configuration provides exceptional GPU density within a 4U form factor.
Flexible Memory Architecture
Up to 4 TB of DDR5 memory across 32 DIMM slots provides the bandwidth and capacity needed for large datasets, complex models, and memory-intensive analytics. This design supports smooth data flow between CPUs, GPUs, and storage resources.
Expansion and Connectivity
Ten PCIe 5.0 x16 dual-slot positions, an additional x16 slot, and two FHHL PCIe lanes support high-speed networking, accelerators, and future expansion. Dual 1 GbE LAN ports and a dedicated management interface enable reliable connectivity and streamlined remote administration.
Perfect For
AI Training
AI Inference
Data Analytics
Simulation
Research
Tech Specs
Resources
Product Downloads
Service & Support
Related Products
















