ASUS RS723Q-E11-RS24: 2U 4-Node Dual-Socket Intel Xeon Server with Liquid Cooling
The ASUS RS723Q-E11-RS24 is a high-density 2U4N (4 independent nodes in 2U) rack server powered by 5th Gen Intel Xeon Scalable processors (Emerald Rapids). Purpose-built for direct-to-chip liquid cooling, it delivers exceptional compute density for HPC, AI inference/training, cloud computing, virtualization, and scale-out infrastructure.
Per-Node Specifications (4 nodes per chassis)
-
Processors: 2x 5th/4th Gen Intel Xeon Scalable (LGA 4677), up to 350W TDP per CPU with liquid cooling
-
Memory: 16x DDR5 DIMM slots (8-channel per CPU) – up to 5600MHz, 4TB per node (using 256GB 3DS RDIMM); supports Intel Optane PMem 300 series
-
Storage per Node: 6x 2.5" drive bays (2x NVMe + 4x SAS/SATA) + 2x M.2 (PCIe Gen4 x2 or SATA)
-
Expansion: 2x PCIe 5.0 slots (1x x16 HHHL + 1x x16/x8 HHHL) + 1x OCP 3.0 Gen5 x16 per node
-
Networking: 1x Intel I210 1GbE LAN + 1x dedicated management LAN per node
-
Rear I/O (per node): 1x USB 3.1 Gen1, 1x mini DisplayPort, 1x GbE LAN, 1x Mgmt LAN
System-Wide Features
-
Total Chassis Storage: 24x 2.5" hot-swap bays (8x NVMe/SAS/SATA + 16x SAS/SATA)
-
Cooling: Direct-to-chip liquid cooling (DLC) – supports 350W TDP CPUs; improves PUE and enables higher rack density
-
Power: 1+1 redundant 3000W/3600W 80 PLUS Titanium PSUs
-
Management: ASMB11-iKVM (ASPEED AST2600 BMC) per node, ASUS Control Center, Redfish API
-
Security: Hardware-level Root-of-Trust per node; optional TPM
-
Dimensions: 444 x 88 x 890 mm (17.48" x 3.46" x 35.04")
-
Net Weight: 37.5 kg (CPU, DRAM, HDD not included)
-
Certifications: BSMI, CB, CE, FCC Class A
Why choose this server?
-
Unmatched compute density: 4 dual-socket nodes in 2U – up to 8 processors, 16 memory channels, 32 DIMMs total
-
Liquid cooling optimized: Direct-to-chip DLC supports 350W TDP processors – enabling higher density than air cooling
-
Massive memory capacity: Up to 4TB DDR5 per node (16 DIMMs) + Intel Optane PMem 300 support
-
High-speed memory: 5600MHz DDR5 across 8 channels per CPU
-
PCIe 5.0 ready: 2x slots + OCP 3.0 per node – 2x bandwidth over PCIe 4.0
-
Scale-out architecture: 24 storage bays (8 NVMe total) – ideal for distributed infrastructure
-
Independent nodes: Each node operates independently with hot-swappable design
-
Power-efficient: 80 PLUS Titanium redundant PSUs – lower PUE and TCO
-
Enhanced performance: 21% better performance per watt for AI workloads versus previous generation
Ideal for organizations seeking maximum compute density, liquid cooling efficiency, and scale-out capabilities for HPC, AI, and cloud data center environments.
