RS620SA-E10-RS12

    RS620SA-E10-RS12

    AMD EPYC 2U6N high-density server with support for up to 6 sockets, 48 DIMMs, 12 NVMe SSDs, 12 M.2 connectors, 12 U.2 connectors, 6 OCP 3.0 cards, and 6 PCIe 4.0 slots

    • Optimized performance for the data center with a single-socket AMD EPYC 7003&7002 processor and eight DIMMs per node
    • Up to six NVIDIA T4 GPUs support VDI, deep learning, AI training and inferencing
    • Unique data-cable routing-tunnel design provides organized cable management and protection
    • High power-efficiency with redundant 3000W or 2200W Platinum power supplies
    • ASUS thermal-control technology offers more efficient power consumption for large-scale environments
    • ASUS performance-tuning technology improves workload throughput to achieve optimal server performance
    • Complete IT infrastructure remote management with ASMB9 (out-of-band) and ASUS Control Center (in-band)
      Product Image
      ASUS RS620SA-E10-RS12 - World's First EPYC 2U6N server ASUS RS620SA-E10-RS12 - World's First EPYC 2U6N server

      Transform your data center TCO with the world's first six-node high-density server powered by AMD EPYC 7003 processors. With an innovative design that fits six nodes in a 2U chassis, RS620SA-E10-RS12 accommodates the highest 280W TDP CPUs, PCIe 4.0 and OCP 3.0 expansions to satisfy the increasing demands for efficiency, density and TCO-optimization for data centers, web servers, virtualization, cloud and hyperscale environments.

      Each
      node
      supports

      High CPU
      Performance

      1 socket

      GPU
      Acceleration

      1 slot

      PCIe 4.0
      Expansion

      1 slot

      OCP 3.0

      1 Mezzanine

      DDR4 3200

      8 DIMM

      NVMe Drive

      2 bays

      Download Datasheet

      * Each node features one PCIe 4.0 slot for a choice of a GPU card or add-on card.
      * The PCIe 4.0 slot fits half-length and half-height (low-profile) GPU cards.
      * Servers with a 280W TDP CPU are required to be configured with an EVAC heatsink.

      Innovation design

      Innovative architecture design
      Data-cable routing-tunnel design

      Scalability meets adaptability

      A one-of-a-kind, ultra-dense 2U6N design delivers breakthrough performance with more compute, storage, networking and PCIe 4.0 expansion options than traditional multi-node systems, allowing you to flexibly configure your server for specific workloads.

      Node 1
      Node 2
      Node 3
      Node 4
      Rear Panel
      Node 5
      Node 6
      Front Panel

      Easy connectivity and serviceability

      Cabling infrastructure is the foundation of every data center. A unique data-cable routing-tunnel design keeps cables together and protected inside the chassis, and connects the rear panel to the networking ports on front panel for organized installation and easier maintenance and troubleshooting.

      An L-shaped routing-tunnel design makes it easy to route and manage cables in the chassis.
      Pull-out tray for easy cable management
      An L-shaped routing-tunnel design makes it easy to route and manage cables in the chassis.

      Product Layout Overview

      System
      Node
      Panels
      1
      2
      3
      4
      5
      6
      7
      8
      9
      10
      11
      12
      13
      1. 12 x 2.5” 15mm hot-swap storage bays
      2. 4 x Fan
      3. 2200W Platinum or 3000W Titanium 1+1 Redundant CRPS Support
      4. 1 x Management port
      5. 1 x 1GbE LAN (i210)
      6. 1 x VGA port
      7. 2 x USB 3.0 port
      8. Power Switch
      9. Unique data cable routing tunnel design
      10. AMD EPYC™ 7003 Processor (up to 280W)
      11. 8 x DDR4 3200/2933/2666 RDIMM/LR-DIMM/3DS
      12. 1 x PCI-E x16 (Gen4 x16 link), LP, HL
      13. 1 x OCP 3.0 type 1 Mezzanine (Gen4 x16 link)
      1
      2
      3
      4
      5
      6
      7
      1. AMD EPYC™ 7003 Processor (up to 280W)
      2. 1 x Management port
      3. 1 x 1GbE LAN (i210)
      4. 1 x PCI-E x16 (Gen4 x16 link), LP, HL
      5. 8 x DDR4 3200/2933/2666 RDIMM/LR-DIMM/3DS
      6. 1 x VGA port
      7. 1 x OCP 3.0 type 1 Mezzanine (Gen4 x16 link)
      Front Panel
      1
      2
      3
      4
      5
      6
      7
      8
      9
      1. 1 x PCI-E x16 (Gen4 x16 link), LP, HL
      2. 1 x OCP 3.0 type 1 Mezzanine (Gen4 x16 link)
      3. VGA Port
      4. 12 x 2.5” 15mm hot-swap storage bays
      5. 1 x Management port
      6. 1 x 1GbE LAN (i210)
      7. 2 x USB 3.0 Port
      8. Port 80
      9. Power Switch
      Rear Panel
      1
      2
      3
      4
      5
      6
      7
      8
      9
      1. 1 x PCI-E x16 (Gen4 x16 link), LP, HL
      2. 1 x OCP 3.0 type 1 Mezzanine (Gen4 x16 link)
      3. VGA Port
      4. 2200W Platinum or 3000W Titanium 1+1 Redundant CRPS Support
      5. 1 x Management port
      6. 1 x 1GbE LAN (i210)
      7. 2 x USB 3.0 Port
      8. Port 80
      9. Power Switch

      AMD EPYC single-socket processor per node

      Each node is powered by the latest single-socket AMD EPYC 7003 processor that delivers higher computing power density of up to 280W TDP and supports up to eight DIMMs, challenging traditional dual-socket server performance, while greatly reducing operating costs.

      AMD EPYC single-socket processor per node
      Single-socket
      per node

      Ultrafast Networking Speed

      RS620SA-E10-RS12 features one OCP 3.0 mezzanine card per node to deliver networking speeds of up to 200 Gbps with PCIe x16 bandwidth. The server is PCIe 4.0 ready and designed for applications that require high-bandwidth, low-latency network support for connecting multiple nodes and clusters.

      Rear Panel Front Panel
      Ultrafast Networking Speed
      6 x OCP 3.0 mezzanine slots (Gen 4.0 x16 link)

      NVIDIA T4 GPU Support

      One PCIe 4.0 expansion slot per node is available for a low-profile GPU, to accelerate a diverse range of modern AI applications. Support for up to six NVIDIA T4 GPUs in a 2U chassis provides distributed computing performance in a space-efficient and cost-effective solution.

      NVIDIA T4 GPU Support
      NVIDIA T4 GPU Support

      * Servers configured with an NVIDIA T4 GPU must also be configured with a CPU with a 280W TDP or less and an EVAC heatsink.

      Applications

      Content Delivery Network
      Web Hosting

      Content Delivery Network

      The six-node architecture design is ideal for content delivery network (CDN) applications that require highly reduced bandwidth costs and improved page load times through a globally distributed network of web servers or points of presence (PoP). Each node on RS620SA-E10-RS12 can serve different functions, such as delivery, storage, original or control nodes to provide more efficient network-resource utilization.

      Content Delivery Network

      Web hosting

      RS620SA-E10-RS12 takes advantage of the single-socket AMD EPYC processor architecture with up to 64 cores and 280W TDP to provide more cores for high-speed computing tasks and better thermal efficiency in data centers.

      Web hosting

      Comprehensive IT infrastructure management

      RS620SA-E10-RS12 features an embedded iKVM module and is bundled with ASUS Control Center to provide comprehensive out-of-band and in-band management features for any scale of businesses.

      ASUS Control Center
      BMC

      ASUS Control Center (ACC) is a centralized and integrated IT management platform for monitoring and controlling ASUS servers, workstations and Mini PCs. ACC enables remote BIOS updates, monitoring of multiple systems via mobile devices, and one-click software updates and dispatching, allowing easier server management for any IT infrastructure.

      Learn more about ASUS Control Center

      ASUS Control Center

      The embedded ASMB9-iKVM module supports remote BIOS updates, fan control, standalone KVMs, video recording, and BSOD capture, providing around-the-clock remote monitoring and diagnostics — even if the operating system is down or offline — through a user-friendly, web-based graphical interface that works with all major browsers.

      BMC

      Performance Boost technology

      ASUS servers feature exclusive Performance Boost technology to achieve the best server performance and agility by tuning servers to match the requirements of workloads, letting you gain greater control of your server environment. This technology improves workload throughput by maximizing processor frequency and boost power, ideal for time-sensitive applications such as financial services or data center operations. In the BIOS you can choose from pre-configured server profiles optimized for specific workloads, maximizing overall performance and reducing server-configuration time.

      Core Optimizer

      Core Optimizer

      Maximizes the processor frequency in multi-core operations, avoiding frequency shifting for reduced latency.

      Engine Boost

      Engine Boost

      Automatic power acceleration with an innovative voltage design to increase server overall performance.

      Workload Presets

      Workload Presets

      Preconfigured BIOS server profiles based on workloads and benchmarks for improved performance and efficiency.

      Thermal control technology

      ASUS features server thermal-control technologies to provide real-time server temperature monitoring, intelligent fan-curve adjustments, and continue delivering the value of optimized power-efficiency.

      ASUS Thermal Radar 2.0
      ASUS Power Balancer

      ASUS Thermal Radar 2.0

      ASUS Thermal Radar 2.0 takes advantage of placing up to 56 ambient sensors on the front panel of a server as well as in components for temperature monitoring. It can perform intelligent fan-curve adjustments to fit the real-time server usage patterns, lowering fan power-consumption by up to 36% and reducing total cost of ownership — ideal for data centers with extreme workloads.

      More Sensors

      More Sensors

      Sensors are embedded on front panel, CPU, NVMe SSD, OCP card, PSU and memory. Up to 56 ambient sensors on selected models.

      Intelligent Adjustment

      Intelligent Adjustment

      Fans are grouped for dynamic fan curve adjustment in different fan zones to achieve more precise thermal monitoring.

      Intelligent Adjustment

      Economic and Eco-friendly

      More precise thermal information can reduce power consumption and lower down TCO.

      ASUS Power Balancer

      ASUS Power Balancer is an exclusive technology that enables ASUS servers to adjust overall loading automatically based on real-time monitoring in order to decrease overall power consumption for improved efficiency and cost/performance optimization.

      Dynamic Monitoring

      Dynamic Monitoring

      Monitor and feedback CPU loading to have more accurate management on the system.

      Auto Frequency Adjustment

      Auto Frequency Adjustment

      Dynamically and automatically adjust CPU frequency based on current utilization to reduce power consumption.

      Enhanced Power Efficiency

      Enhanced Power Efficiency

      Optimize performance per watt for a better power efficiency system.