NVSwitch enables every GPU to communicate with every other GPU at full bandwidth of 2.4TB/sec to solve the largest of AI and HPC problems. Every GPU has full access to 0.5TB of aggregate HBM2 memory to handle the most massive of datasets. By enabling a unified server node, NVSwitch dramatically accelerates complex AI and HPC applications.
Designed for Next Generation AI
AI models are exploding in complexity and require large memory, multiple GPUs, and an extremely fast connection between the GPUs to work. With NVSwitch connecting all GPUs and unified memory, HGX-2 provides the power to handle these new models for faster training of advanced AI. A single HGX-2 replaces 300 CPU-powered servers, saving significant cost, space, and energy in the data center.
Ready for Max Density HPC
HPC applications require strong server nodes with the computing power to perform a massive number of calculations per second. Increasing the compute density of each node dramatically reduces the number of servers required, resulting in huge savings in cost, power, and space consumed in the data center. For HPC simulations, high-dimension matrix multiplication requires a processor to fetch data from many neighbors to facilitate computation, making GPUs connected by NVSwitch ideal. A single HGX-2 server replaces 60 CPU-only servers.
X11 Servers Featuring New Intel Skylake Scalable Xeon® Processors
Supermicro’s new X11 servers are engineered to unleash the full performance and rich feature sets on the new Intel® Xeon® Scalable processor family, supporting more cores and higher TDP envelopes of 205 watts and higher, increased number of memory channels and higher bandwidth, more PCI-E 3.0 lanes, 100G/40G/25G/10G Ethernet, 100G EDR InfiniBand (on select servers) and integrated Intel® Omni-Path Architecture networking fabrics. The elevated compute performance, density, I/O capacity, and efficiency are coupled with industry’s most comprehensive support for NVMe NAND Flash and Intel® Optane™ SSDs for unprecedented application responsiveness and agility. For exact sever specifications, please see highlights below and also refer to detailed technical specifications.
“Supermicro’s new SuperServer based on the HGX-2 platform will deliver more than double the performance of current systems, which will help enterprises address the rapidly expanding size of AI models that sometimes require weeks to train,” said Charles Liang, president and CEO of Supermicro. “Our new HGX-2 system will enable efficient training of complex models. It combines sixteen Tesla V100 32GB SXM3 GPUs connected via NVLink and NVSwitch to work as a unified 2 PetaFlop accelerator with half a terabyte of aggregate GPU memory to deliver unmatched compute power.”
Charles Liang, President and CEO of Supermicro
Server Systems Management
Supermicro Server Manager (SSM) provides capabilities to monitor the health of server components including memory, hard drives and RAID controllers. It enables the datacenter administrator to monitor and manage power usage across all Supermicro servers allowing users to maximize their CPU payload while mitigating the risk of tripped circuit. Firmware upgrades on Supermicro servers became easier now with a couple of clicks. Administrators can now mount an ISO image on multiple servers and reboot the servers with those images. The tool also provides pre-defined reports and many more features that will make managing Supermicro servers simpler. Download the SSM_brochure for more info or download Supermicro SuperDoctor® device monitoring and management software.
The Tesla V100 GPU, widely adopted by the world’s leading researchers, has received a 2x memory boost to handle the most memory-intensive deep learning and high performance computing workloads.
Now equipped with 32GB of memory, Tesla V100 GPUs will help data scientists train deeper and larger deep learning models that are more accurate than ever. They can also improve the performance of memory-constrained HPC applications by up to 50 percent compared with the previous 16GB version.
GroundBreaking Volta Architecture
By pairing CUDA Cores and Tensor Cores within a unified architecture, a single server with Tesla V100 GPUs can replace hundreds of commodity CPU servers for traditional HPC and Deep Learning. TENSOR CORE Equipped with 640 Tensor Cores, Tesla V100 delivers 125 TeraFLOPS of deep learning performance. That’s 12X Tensor FLOPS for DL Training, and 6X Tensor FLOPS for DL Inference when compared to NVIDIA Pascal™ GPUs.
Next Generation NVLink
NVIDIA NVLink in Tesla V100 delivers 2X higher throughput compared to the previous generation. Up to eight Tesla V100 accelerators can be interconnected at up to 300 GB/s to unleash the highest application performance possible on a single server. HBM2 With a combination of improved raw bandwidth of 900 GB/s and higher DRAM utilization efficiency at 95%, Tesla V100 delivers 1.5X higher memory bandwidth over Pascal GPUs as measured on STREAM.
Maximum Efficiency Mode
The new maximum efficiency mode allows data centers to achieve up to 40% higher compute capacity per rack within the existing power budget. In this mode, Tesla V100 runs at peak processing efficiency, providing up to 80% of the performance at half the power consumption.
Programmability
Tesla V100 is architected from the ground up to simplify programmability. Its new independent thread scheduling enables finer-grain synchronization and improves GPU utilization by sharing resources among small jobs.
CUDA Ready
CUDA® is a parallel computing platform and programming model developed by NVIDIA for general computing on graphical processing units (GPUs). With CUDA, developers are able to dramatically speed up computing applications by harnessing the power of GPUs.
In GPU-accelerated applications, the sequential part of the workload runs on the CPU – which is optimized for single-threaded performance – while the compute intensive portion of the application runs on thousands of GPU cores in parallel. When using CUDA, developers program in popular languages such as C, C++, Fortran, Python and MATLAB and express parallelism through extensions in the form of a few basic keywords.
The CUDA Toolkit from NVIDIA provides everything you need to develop GPU-accelerated applications. The CUDA Toolkit includes GPU-accelerated libraries, a compiler, development tools and the CUDA runtime.
Performance Specifications for NVIDIA Tesla P4, P40 and V100 Accelerators
Tesla V100: The Universal Datacenter GPU
Tesla P4 for Ultra-Efficient Scale-Out Servers
Tesla P40 for Inference Throughput Servers
Single-Precision Performance (FP32)
14 teraflops (PCIe)
15.7 teraflops (SXM2)
5.5 teraflops
12 teraflops
Half-Precision Performance (FP16)
112 teraflops (PCIe)
125 teraflops (SXM2)
—
—
Integer Operations (INT8)
—
22 TOPS*
47 TOPS*
GPU Memory
16/32 GB HBM2
8 GB
24 GB
Memory Bandwidth
900 GB/s
192 GB/s
346 GB/s
System Interface/Form Factor
Dual-Slot, Full-Height PCI Express Form Factor
SXM2 / NVLink
Low-Profile PCI Express Form Factor
Dual-Slot, Full-Height PCI Express Form Factor
Power
250W (PCIe)
300W (SXM2)
50 W/75 W
250 W
Hardware-Accelerated Video Engine
—
1x Decode Engine, 2x Encode Engines
1x Decode Engine, 2x Encode Engines
*Tera-Operations per Second with Boost Clock Enabled
By pairing CUDA Cores and Tensor Cores within a unified architecture, a single server with Tesla V100 GPUs can replace hundreds of commodity CPU servers for traditional HPC and Deep Learning. TENSOR CORE Equipped with 640 Tensor Cores, Tesla V100 delivers 125 TeraFLOPS of deep learning performance. That’s 12X Tensor FLOPS for DL Training, and 6X Tensor FLOPS for DL Inference when compared to NVIDIA Pascal™ GPUs.
Next Generation NVLink
NVIDIA NVLink in Tesla V100 delivers 2X higher throughput compared to the previous generation. Up to eight Tesla V100 accelerators can be interconnected at up to 300 GB/s to unleash the highest application performance possible on a single server. HBM2 With a combination of improved raw bandwidth of 900 GB/s and higher DRAM utilization efficiency at 95%, Tesla V100 delivers 1.5X higher memory bandwidth over Pascal GPUs as measured on STREAM.
Maximum Efficiency Mode
The new maximum efficiency mode allows data centers to achieve up to 40% higher compute capacity per rack within the existing power budget. In this mode, Tesla V100 runs at peak processing efficiency, providing up to 80% of the performance at half the power consumption.
Programmability
Tesla V100 is architected from the ground up to simplify programmability. Its new independent thread scheduling enables finer-grain synchronization and improves GPU utilization by sharing resources among small jobs.
CUDA Ready
CUDA® is a parallel computing platform and programming model developed by NVIDIA for general computing on graphical processing units (GPUs). With CUDA, developers are able to dramatically speed up computing applications by harnessing the power of GPUs.
In GPU-accelerated applications, the sequential part of the workload runs on the CPU – which is optimized for single-threaded performance – while the compute intensive portion of the application runs on thousands of GPU cores in parallel. When using CUDA, developers program in popular languages such as C, C++, Fortran, Python and MATLAB and express parallelism through extensions in the form of a few basic keywords.
The CUDA Toolkit from NVIDIA provides everything you need to develop GPU-accelerated applications. The CUDA Toolkit includes GPU-accelerated libraries, a compiler, development tools and the CUDA runtime.
Performance Specifications for NVIDIA Tesla P4, P40 and V100 Accelerators
Tesla V100: The Universal Datacenter GPU
Tesla P4 for Ultra-Efficient Scale-Out Servers
Tesla P40 for Inference Throughput Servers
Single-Precision Performance (FP32)
14 teraflops (PCIe)
15.7 teraflops (SXM2)
5.5 teraflops
12 teraflops
Half-Precision Performance (FP16)
112 teraflops (PCIe)
125 teraflops (SXM2)
—
—
Integer Operations (INT8)
—
22 TOPS*
47 TOPS*
GPU Memory
16/32 GB HBM2
8 GB
24 GB
Memory Bandwidth
900 GB/s
192 GB/s
346 GB/s
System Interface/Form Factor
Dual-Slot, Full-Height PCI Express Form Factor
SXM2 / NVLink
Low-Profile PCI Express Form Factor
Dual-Slot, Full-Height PCI Express Form Factor
Power
250W (PCIe)
300W (SXM2)
50 W/75 W
250 W
Hardware-Accelerated Video Engine
—
1x Decode Engine, 2x Encode Engines
1x Decode Engine, 2x Encode Engines
*Tera-Operations per Second with Boost Clock Enabled
Lenovo ThinkSystem servers support GPU technology to accelerate different computing workloads, maximize performance for graphic design, virtualization, artificial intelligence and high performance computing applications in Lenovo servers.
The following table summarizes the server support for the GPUs. The numbers listed in the server columns represent the number of GPUs supported.
GroundBreaking Volta Architecture
By pairing CUDA Cores and Tensor Cores within a unified architecture, a single server with Tesla V100 GPUs can replace hundreds of commodity CPU servers for traditional HPC and Deep Learning. TENSOR CORE Equipped with 640 Tensor Cores, Tesla V100 delivers 125 TeraFLOPS of deep learning performance. That’s 12X Tensor FLOPS for DL Training, and 6X Tensor FLOPS for DL Inference when compared to NVIDIA Pascal™ GPUs.
Next Generation NVLink
NVIDIA NVLink in Tesla V100 delivers 2X higher throughput compared to the previous generation. Up to eight Tesla V100 accelerators can be interconnected at up to 300 GB/s to unleash the highest application performance possible on a single server. HBM2 With a combination of improved raw bandwidth of 900 GB/s and higher DRAM utilization efficiency at 95%, Tesla V100 delivers 1.5X higher memory bandwidth over Pascal GPUs as measured on STREAM.
Maximum Efficiency Mode
The new maximum efficiency mode allows data centers to achieve up to 40% higher compute capacity per rack within the existing power budget. In this mode, Tesla V100 runs at peak processing efficiency, providing up to 80% of the performance at half the power consumption.
Programmability
Tesla V100 is architected from the ground up to simplify programmability. Its new independent thread scheduling enables finer-grain synchronization and improves GPU utilization by sharing resources among small jobs.
CUDA Ready
CUDA® is a parallel computing platform and programming model developed by NVIDIA for general computing on graphical processing units (GPUs). With CUDA, developers are able to dramatically speed up computing applications by harnessing the power of GPUs.
In GPU-accelerated applications, the sequential part of the workload runs on the CPU – which is optimized for single-threaded performance – while the compute intensive portion of the application runs on thousands of GPU cores in parallel. When using CUDA, developers program in popular languages such as C, C++, Fortran, Python and MATLAB and express parallelism through extensions in the form of a few basic keywords.
The CUDA Toolkit from NVIDIA provides everything you need to develop GPU-accelerated applications. The CUDA Toolkit includes GPU-accelerated libraries, a compiler, development tools and the CUDA runtime.
Performance Specifications for NVIDIA Tesla P4, P40 and V100 Accelerators
Tesla V100: The Universal Datacenter GPU
Tesla P4 for Ultra-Efficient Scale-Out Servers
Tesla P40 for Inference Throughput Servers
Single-Precision Performance (FP32)
14 teraflops (PCIe)
15.7 teraflops (SXM2)
5.5 teraflops
12 teraflops
Half-Precision Performance (FP16)
112 teraflops (PCIe)
125 teraflops (SXM2)
—
—
Integer Operations (INT8)
—
22 TOPS*
47 TOPS*
GPU Memory
16/32 GB HBM2
8 GB
24 GB
Memory Bandwidth
900 GB/s
192 GB/s
346 GB/s
System Interface/Form Factor
Dual-Slot, Full-Height PCI Express Form Factor
SXM2 / NVLink
Low-Profile PCI Express Form Factor
Dual-Slot, Full-Height PCI Express Form Factor
Power
250W (PCIe)
300W (SXM2)
50 W/75 W
250 W
Hardware-Accelerated Video Engine
—
1x Decode Engine, 2x Encode Engines
1x Decode Engine, 2x Encode Engines
*Tera-Operations per Second with Boost Clock Enabled
Lenovo ThinkSystem servers support GPU technology to accelerate different computing workloads, maximize performance for graphic design, virtualization, artificial intelligence and high performance computing applications in Lenovo servers.
The following table summarizes the server support for the GPUs. The numbers listed in the server columns represent the number of GPUs supported.
Double Memory than Previous Generation V100
The Tesla V100 GPU, widely adopted by the world’s leading researchers, has received a 2x memory boost to handle the most memory-intensive deep learning and high performance computing workloads.
Now equipped with 32GB of memory, Tesla V100 GPUs will help data scientists train deeper and larger deep learning models that are more accurate than ever. They can also improve the performance of memory-constrained HPC applications by up to 50 percent compared with the previous 16GB version.
GroundBreaking Volta Architecture
By pairing CUDA Cores and Tensor Cores within a unified architecture, a single server with Tesla V100 GPUs can replace hundreds of commodity CPU servers for traditional HPC and Deep Learning. TENSOR CORE Equipped with 640 Tensor Cores, Tesla V100 delivers 125 TeraFLOPS of deep learning performance. That’s 12X Tensor FLOPS for DL Training, and 6X Tensor FLOPS for DL Inference when compared to NVIDIA Pascal™ GPUs.
Next Generation NVLink
NVIDIA NVLink in Tesla V100 delivers 2X higher throughput compared to the previous generation. Up to eight Tesla V100 accelerators can be interconnected at up to 300 GB/s to unleash the highest application performance possible on a single server. HBM2 With a combination of improved raw bandwidth of 900 GB/s and higher DRAM utilization efficiency at 95%, Tesla V100 delivers 1.5X higher memory bandwidth over Pascal GPUs as measured on STREAM.
Maximum Efficiency Mode
The new maximum efficiency mode allows data centers to achieve up to 40% higher compute capacity per rack within the existing power budget. In this mode, Tesla V100 runs at peak processing efficiency, providing up to 80% of the performance at half the power consumption.
Programmability
Tesla V100 is architected from the ground up to simplify programmability. Its new independent thread scheduling enables finer-grain synchronization and improves GPU utilization by sharing resources among small jobs.
CUDA Ready
CUDA® is a parallel computing platform and programming model developed by NVIDIA for general computing on graphical processing units (GPUs). With CUDA, developers are able to dramatically speed up computing applications by harnessing the power of GPUs.
In GPU-accelerated applications, the sequential part of the workload runs on the CPU – which is optimized for single-threaded performance – while the compute intensive portion of the application runs on thousands of GPU cores in parallel. When using CUDA, developers program in popular languages such as C, C++, Fortran, Python and MATLAB and express parallelism through extensions in the form of a few basic keywords.
The CUDA Toolkit from NVIDIA provides everything you need to develop GPU-accelerated applications. The CUDA Toolkit includes GPU-accelerated libraries, a compiler, development tools and the CUDA runtime.
Performance Specifications for NVIDIA Tesla P4, P40 and V100 Accelerators
Tesla V100: The Universal Datacenter GPU
Tesla P4 for Ultra-Efficient Scale-Out Servers
Tesla P40 for Inference Throughput Servers
Single-Precision Performance (FP32)
14 teraflops (PCIe)
15.7 teraflops (SXM2)
5.5 teraflops
12 teraflops
Half-Precision Performance (FP16)
112 teraflops (PCIe)
125 teraflops (SXM2)
—
—
Integer Operations (INT8)
—
22 TOPS*
47 TOPS*
GPU Memory
16/32 GB HBM2
8 GB
24 GB
Memory Bandwidth
900 GB/s
192 GB/s
346 GB/s
System Interface/Form Factor
Dual-Slot, Full-Height PCI Express Form Factor
SXM2 / NVLink
Low-Profile PCI Express Form Factor
Dual-Slot, Full-Height PCI Express Form Factor
Power
250W (PCIe)
300W (SXM2)
50 W/75 W
250 W
Hardware-Accelerated Video Engine
—
1x Decode Engine, 2x Encode Engines
1x Decode Engine, 2x Encode Engines
*Tera-Operations per Second with Boost Clock Enabled
X11 Servers Featuring New Intel Skylake Scalable Xeon® Processors
Supermicro’s new X11 servers are engineered to unleash the full performance and rich feature sets on the new Intel® Xeon® Scalable processor family, supporting more cores and higher TDP envelopes of 205 watts and higher, increased number of memory channels and higher bandwidth, more PCI-E 3.0 lanes, 100G/40G/25G/10G Ethernet, 100G EDR InfiniBand (on select servers) and integrated Intel® Omni-Path Architecture networking fabrics. The elevated compute performance, density, I/O capacity, and efficiency are coupled with industry’s most comprehensive support for NVMe NAND Flash and Intel® Optane™ SSDs for unprecedented application responsiveness and agility. For exact sever specifications, please see highlights below and also refer to detailed technical specifications.
“At Supermicro, we understand that customers need the newest technologies as early as possible to drive leading performance and improved TCO. With the industry’s strongest and broadest product line, our designs not only take full advantage of Xeon Scalable Processors’ new features such as three UPI, faster DIMMs and more core count per socket, but they also fully support NVMe through unique non-blocking architectures to achieve the best data bandwidth and IOPS. For instance, one Supermicro 2U storage server can deliver over 16 million IOPS!”
“Supermicro designs the most application-optimized GPU systems and offers the widest selection of GPU-optimized servers and workstations in the industry. Our high performance computing solutions enable deep learning, engineering and scientific fields to scale out their compute clusters to accelerate their most demanding workloads and achieve fastest time-to-results with maximum performance per watt, per square foot and per dollar. With our latest innovations incorporating the new NVIDIA V100 PCI-E and V100 SXM2 GPUs in performance-optimized 1U and 4U systems with next-generation NVLink, our customers can accelerate their applications and innovations to help solve the world’s most complex and challenging problems.”
Charles Liang, President and CEO of Supermicro
Support for 8 Double Width GPUs for Deep Learning
The 4029GP-TRT2 takes full advantage of the new Xeon Scalable Processor Family PCIe lanes to support 8 double-width GPUs to deliver a very high performance Artificial Intelligence and Deep Learning system suitable for autonomous cars, molecular dynamics, computational biology, fluid simulation, advanced physics and Internet of Things (IoT) and Big Data Analytics etc. With NVIDIA Tesla cards, this server delivers unparalleled acceleration for compute intensive applications.
Server Systems Management
Supermicro Server Manager (SSM) provides capabilities to monitor the health of server components including memory, hard drives and RAID controllers. It enables the datacenter administrator to monitor and manage power usage across all Supermicro servers allowing users to maximize their CPU payload while mitigating the risk of tripped circuit. Firmware upgrades on Supermicro servers became easier now with a couple of clicks. Administrators can now mount an ISO image on multiple servers and reboot the servers with those images. The tool also provides pre-defined reports and many more features that will make managing Supermicro servers simpler. Download the SSM_brochure for more info or download Supermicro SuperDoctor® device monitoring and management software.
Technical Specifications
Mfr Part #
SYS-4029GP-TRT2
Motherboard
Super X11DPG-OT-CPU
CPU
Dual Socket P (LGA 3647); Intel® Xeon® Scalable Processors,
Dual UPI up to 10.4GT/s; Dual UPI up to 10.4GT/s; Support CPU TDP 70-205W2 x Intel Skylake Xeon Silver 4116 2.1 GHz 12 Core CPU Installed
Cores
Up to 28 Cores with Intel® HT Technology
GPU / Coprocessor Support
Please refer to: Compatible GPU list
Memory Capacity
24 DIMM slots; Up to 3TB ECC 3DS LRDIMM, 1TB ECC RDIMM, DDR4 up to 2666MHz
128 GB DDR4-2666MHz (32GB x 4) Installed
Memory Type
2666/2400/2133MHz ECC DDR4 SDRAM
Chipset
Intel® C622 chipset
SATA
SATA3 (6Gbps) with RAID 0, 1, 5, 10
Network Controllers
Dual Port 10GbE from C622
IPMI
Support for Intelligent Platform Management Interface v.2.0; IPMI 2.0 with virtual media over LAN and KVM-over-LAN support
Graphics
ASPEED AST2500 BMC
SATA
10 SATA3 (6Gbps) ports
LAN
2 RJ45 10GBase-T LAN ports; 1 RJ45 Dedicated IPMI LAN port
X11 Servers Featuring New Intel Skylake Scalable Xeon® Processors
Supermicro’s new X11 servers are engineered to unleash the full performance and rich feature sets on the new Intel® Xeon® Scalable processor family, supporting more cores and higher TDP envelopes of 205 watts and higher, increased number of memory channels and higher bandwidth, more PCI-E 3.0 lanes, 100G/40G/25G/10G Ethernet, 100G EDR InfiniBand (on select servers) and integrated Intel® Omni-Path Architecture networking fabrics. The elevated compute performance, density, I/O capacity, and efficiency are coupled with industry’s most comprehensive support for NVMe NAND Flash and Intel® Optane™ SSDs for unprecedented application responsiveness and agility. For exact sever specifications, please see highlights below and also refer to detailed technical specifications.
“At Supermicro, we understand that customers need the newest technologies as early as possible to drive leading performance and improved TCO. With the industry’s strongest and broadest product line, our designs not only take full advantage of Xeon Scalable Processors’ new features such as three UPI, faster DIMMs and more core count per socket, but they also fully support NVMe through unique non-blocking architectures to achieve the best data bandwidth and IOPS. For instance, one Supermicro 2U storage server can deliver over 16 million IOPS!”
“Supermicro designs the most application-optimized GPU systems and offers the widest selection of GPU-optimized servers and workstations in the industry. Our high performance computing solutions enable deep learning, engineering and scientific fields to scale out their compute clusters to accelerate their most demanding workloads and achieve fastest time-to-results with maximum performance per watt, per square foot and per dollar. With our latest innovations incorporating the new NVIDIA V100 PCI-E and V100 SXM2 GPUs in performance-optimized 1U and 4U systems with next-generation NVLink, our customers can accelerate their applications and innovations to help solve the world’s most complex and challenging problems.”
Charles Liang, President and CEO of Supermicro
Support for 8 Double Width GPUs for Deep Learning
The 4029GP-TRT2 takes full advantage of the new Xeon Scalable Processor Family PCIe lanes to support 8 double-width GPUs to deliver a very high performance Artificial Intelligence and Deep Learning system suitable for autonomous cars, molecular dynamics, computational biology, fluid simulation, advanced physics and Internet of Things (IoT) and Big Data Analytics etc. With NVIDIA Tesla cards, this server delivers unparalleled acceleration for compute intensive applications.
Server Systems Management
Supermicro Server Manager (SSM) provides capabilities to monitor the health of server components including memory, hard drives and RAID controllers. It enables the datacenter administrator to monitor and manage power usage across all Supermicro servers allowing users to maximize their CPU payload while mitigating the risk of tripped circuit. Firmware upgrades on Supermicro servers became easier now with a couple of clicks. Administrators can now mount an ISO image on multiple servers and reboot the servers with those images. The tool also provides pre-defined reports and many more features that will make managing Supermicro servers simpler. Download the SSM_brochure for more info or download Supermicro SuperDoctor® device monitoring and management software.
Technical Specifications
Mfr Part #
SYS-4029GP-TRT2
Motherboard
Super X11DPG-OT-CPU
CPU
Dual Socket P (LGA 3647); Intel® Xeon® Scalable Processors,
Dual UPI up to 10.4GT/s; Dual UPI up to 10.4GT/s; Support CPU TDP 70-205W2 x Intel Skylake Xeon Silver 4116 2.1 GHz 12 Core CPU Installed
Cores
Up to 28 Cores with Intel® HT Technology
GPU / Coprocessor Support
Please refer to: Compatible GPU list
Memory Capacity
24 DIMM slots; Up to 3TB ECC 3DS LRDIMM, 1TB ECC RDIMM, DDR4 up to 2666MHz
256 GB DDR4-2666MHz (32GB x 8) Installed
Memory Type
2666/2400/2133MHz ECC DDR4 SDRAM
Chipset
Intel® C622 chipset
SATA
SATA3 (6Gbps) with RAID 0, 1, 5, 10
Network Controllers
Dual Port 10GbE from C622
IPMI
Support for Intelligent Platform Management Interface v.2.0; IPMI 2.0 with virtual media over LAN and KVM-over-LAN support
Graphics
ASPEED AST2500 BMC
SATA
10 SATA3 (6Gbps) ports
LAN
2 RJ45 10GBase-T LAN ports; 1 RJ45 Dedicated IPMI LAN port
X11 Servers Featuring New Intel Skylake Scalable Xeon® Processors
Supermicro’s new X11 servers are engineered to unleash the full performance and rich feature sets on the new Intel® Xeon® Scalable processor family, supporting more cores and higher TDP envelopes of 205 watts and higher, increased number of memory channels and higher bandwidth, more PCI-E 3.0 lanes, 100G/40G/25G/10G Ethernet, 100G EDR InfiniBand (on select servers) and integrated Intel® Omni-Path Architecture networking fabrics. The elevated compute performance, density, I/O capacity, and efficiency are coupled with industry’s most comprehensive support for NVMe NAND Flash and Intel® Optane™ SSDs for unprecedented application responsiveness and agility. For exact sever specifications, please see highlights below and also refer to detailed technical specifications.
“At Supermicro, we understand that customers need the newest technologies as early as possible to drive leading performance and improved TCO. With the industry’s strongest and broadest product line, our designs not only take full advantage of Xeon Scalable Processors’ new features such as three UPI, faster DIMMs and more core count per socket, but they also fully support NVMe through unique non-blocking architectures to achieve the best data bandwidth and IOPS. For instance, one Supermicro 2U storage server can deliver over 16 million IOPS!”
“We are excited to preview our X11 Ultra, TwinPro™, BigTwin™, SuperBlade® and many more new designs based on the new Intel® Xeon® Processor Scalable Family processors with CPUs that can provide up to 3.9x higher virtualized throughput,”
Charles Liang, President and CEO of Supermicro
Server Systems Management
Supermicro Server Manager (SSM) provides capabilities to monitor the health of server components including memory, hard drives and RAID controllers. It enables the datacenter administrator to monitor and manage power usage across all Supermicro servers allowing users to maximize their CPU payload while mitigating the risk of tripped circuit. Firmware upgrades on Supermicro servers became easier now with a couple of clicks. Administrators can now mount an ISO image on multiple servers and reboot the servers with those images. The tool also provides pre-defined reports and many more features that will make managing Supermicro servers simpler. Download the SSM_brochure for more info or download Supermicro SuperDoctor® device monitoring and management software.
Technical Specifications
Mfr Part #
SYS-7049GP-TRT (Black)
CPU
Dual Socket P (LGA 3647); Intel® Xeon® Scalable Processors,
3 UPI up to 10.4GT/s; Support CPU TDP 70-205W with IVR 2 x Intel Xeon Skylake Gold 4114 10 Core 2.2 GHz CPU Installed
Cores
Up to 28 Cores with Intel® HT Technology
GPU
GPU Support Matrix
Memory Capacity
16 DIMM slots; Up to 2TB ECC 3DS LRDIMM, 1TB ECC RDIMM, DDR4 up to 2666MHz
96GB ECC DDR4-2666MHz installed (8GB x 12); expandable up to 2TB
Memory Type
2666/2400/2133MHz ECC DDR4 SDRAM
Chipset
Intel® C621 chipset
SATA
SATA3 (6Gbps); RAID 0, 1, 5, 10
Network Controllers
Intel® X550 Dual Port 10GBase-T; Virtual Machine Device Queues reduce I/O overhead; Supports 10GBASE-T, 100BASE-TX, and 1000BASE-T, RJ45 output
IPMI
Support for Intelligent Platform Management Interface v.2.0; IPMI 2.0 with virtual media over LAN and KVM-over-LAN support
Graphics
ASPEED AST2500 BMC
SATA
10 SATA3 (6Gbps) ports
LAN
2 RJ45 10GBase-T ports; 1 RJ45 Dedicated IPMI LAN port
USB
5 USB 3.0 ports (2 rear, 2 via header, 1 Type A); 4 USB 2.0 ports (2 rear, 2 via headers)
1 x Seagate 2.5″ 480GB SATA 6Gb/s, 7.0mm, 16nm,0.6 DWPD SSD Installed
1 x Toshiba 3.5″ 4TB SATA 6Gb/s 7.2K RPM 128M 512E (Tomcat) HDD Installed
Fans
4 Heavy duty fans; 4 Rear exhaust fans
Heatsink
2 Active heatsink with optimal fan speed control
Total Output Power and Input
1200W with Input 100-127Vac; 1800W with Input 200-220Vac; 1980W with Input 220-230Vac; 2090W with Input 230-240Vac; 2090W with Input 200-220Vac; 2200W with Input 220-240Vac (for UL/cUL use only); 2090W with Input 230-240Vdc (for CCC only)
TYAN offers a wide range of GPU (graphics processing unit) computing platforms that are designed for High Performance Computing (HPC) and massive parallel computing environments. TYAN’s GA88B8021 (B8021G88V2HR-2T-N) GPU platform supports up to 6 GPUs inside traditional 1U rackmount chassis, which helps reduce the rack space requirements, required for high performance applications. With the right workload, these platforms can offer higher performance and reduce power consumption when compared to traditional CPU centric platforms.
Support for NVIDIA® Tesla® GPU Accelerators:
TYAN and NVIDIA have joined together to deliver a GPU computing platform that provides double-precision performance that is both cost-effective and energy-efficient. With support for NVIDIA Tesla GPUs, TYAN’s GPU Computing platforms offer massively parallel compute power to solve the most computationally-intensive challenges.
Target applications for TYAN GPU platforms:
High Performance GPU frameworks
Cryptography
Oil & Gas Discovery
Particle Physics
Genetic Sequencing
Cryptocurrency Mining
Large Scale Facial Recognition
Machine Learning and Artificial Intelligence
Featuring New AMD EPYC™ 7000 Series Processors
Tyan’s new AMD servers are engineered to unleash the full performance and rich feature sets on the new AMD EPYC™ 7000 Series processor family with more cores, more memory, more I/O, and more security. AMD EPYC processors are based on the 14nm “Zen” x86 core architecture with 32 cores and 64 threads, featuring 8 memory channels and up to16 DIMMs per socket. The new powerful processor allows Tyan to offer the optimal balance of computing, memory and I/O for the customers. By adopting the latest AMD EPYC processor technology, Tyan brings high memory channel bandwidth and PCI Express high speed I/O connectivity features to enterprises and data centers. With servers built on EPYC technology, cloud environments can drive greater scale and performance, virtualized data centers can further increase consolidation ratios while delivering better performing virtual machines and Big Data, and analytics environments can collect and analyze larger data sets much faster. High performance applications in research labs can solve complex problem sets in a significantly accelerated manner. EPYC, with all the critical compute, memory, I/O, and security resources brought together in the SoC with the right ratios, delivers industry-leading performance and enables lower TCO. With the flexibility to choose from 8 to 32 cores, EPYC enables you to deploy the right hardware platforms to meet your workload needs from virtualized infrastructure to large-scale big-data and analytics platforms and legacy line-of-business applications. For exact sever specifications, please see highlights below and also refer to detailed technical specifications.
Tyan’s Transport HX GA88-B8021 offers outstanding computational power for a single-socket server, with support for 8, 16, 24, or 32-Core processors with up to 64 threads. The memory capacity of the server is also incredibly high for a single socket server, able to support up to 2TB of RAM with every processor in the AMD EPYC SKU stack.
Technical Specifications
TYAN GA88B8021 (B8021G88V2HR-2T-N) Specifications
System
Form Factor
1U Rackmount
Chassis Model
GA88
Dimension (D x W x H)
34.84″ x 17.24″ x 1.71″ (885 x 438 x 43.5mm)
Motherboard Name
S8021GM2NR-2T
Gross Weight
26.5 kg (58.5 lbs)
Net weight
13.5 kg (30 lbs)
Front Panel
Buttons
(1) RST / (1) PWR w/ LED / (1) UID
LEDs
(1) PWR / (2) LAN / (1) UID / (1) BMC event
I/O Ports
(2) USB 3.0 ports
External Drive Bay
Q’ty / Type
(2) 2.5″ Hot-Swap SSD/HDD
HDD Backplane Support
SAS 12Gb/s /SATA 6Gb/s
Supported HDD Interface
(2) SAS 12Gb/s / SATA 6Gb/s
System Cooling Configuration
Redundancy
Yes
FAN
(10) 4cm fans
Heat Sink
(1) Passive CPU Heat sink
Power Supply
Type
ERP1U
Input Range
AC 100-127V/12A / AC 200-240V/9.48A
Frequency
47-63 Hz
Output Watts
1600 Watts
Efficiency
80 plus Platinum
Redundancy
1+1
Processor
Q’ty / Socket Type
(1) AMD Socket SP3
Supported CPU Series
(1) AMD EPYC™ 7000 Series Processor
System Bus
1066/ 800/ 533 MHz
Average CPU Power (ACP) Wattage
Max up to 180W
Memory
Supported DIMM Qty
(16) DIMM slots
DIMM Type / Speed
DDR4 ECC RDIMM/LRDIMM 2666/2400
Capacity
Up to 2,048GB RDIMM/ 1,024GB LRDIMM/ 2,048GB LRDIMM 3DS
*Actual memory speed depends on memory population topology
Hardware Monitor / Boot from USB device/PXE via LAN/Storage / Console Redirection / ACPI 6.1 / SMBIOS 3.1/PnP/Wake on LAN / ACPI sleeping states S0, S5 / FAN speed control automatic
Operating System
OS supported list
Please refer to our AVL support lists.
Regulation
FCC (SDoC)
Class A
CE (DoC)
Class A
CB/LVD
Yes
RCM
Class A
VCCI
Class A
Operating Environment
Operating Temp.
10° C ~ 35° C (50° F~ 95° F)
Non-operating Temp.
– 40° C ~ 70° C (-40° F ~ 158° F)
In/Non-operating Humidity
90%, non-condensing at 35° C
RoHS
RoHS 6/6 Compliant
Yes
Package Contains
Barebone
(1) GA88-B8021 Barebone
Manual
(1) Quick Installation Guide
Installation CD
(1) TYAN Device Driver CD
Select Product Category
Flexible Payment Options
Order Online & Pay Offline!
Besides using Credit Card, PayPal or Amazon Pay, you have the option to place an order online without instant payment using Bank/Wire Transfer method at checkout and then transfer funds after receiving order confirmation. Note - order will only ship after receipt of payment, except for approved Net Terms orders. This option is suitable for larger orders and for international customers.
Why Dihuni?
Our Partners
Our Customers
Our deep understanding of markets, technology and applications, partner relationships in addition to working closely with customers on product requirements has helped us serve customers ranging from Fortune Corporations to Universities, Schools, Startups, Small and Medium Businesses, Government and Individual Buyers. Our customers are among :
Top 5 Semiconductor Companies
Top 3 RideShare Companies
Top 5 IT Infrastructure Companies
Ivy League Universities
Top Defense Contractors, US Navy etc
Top IoT, AI and Deep Learning Startups
“With Dihuni’s prompt service, we got up to speed with NVIDIA Tesla V100 32GB GPU for our Deep Learning application. The product was just released by NVIDIA and Dihuni delivered on time for our project and without hassle!“
Engineering Manager, Tech Co, USA
“We purchased a variety of systems for low-power applications from Dihuni. Dihuni engaged with us to ensure we selected the best options for our requirements and gave us access to newly introduced systems. We would recommend Dihuni to other buyers because of the quality of their customer care.”
Strategic Marketing Manager, Top 15 Semiconductor Co.
“We’ve purchased AMD EPYC Servers from Dihuni, and their service and response was top notch. They clarified exactly what we needed, what type of motherboard will fit best our needs, and delivered the servers on time and in tip top condition. Was a pleasure working with them!”
CTO, Tech Co., Tel Aviv, Israel
“We ordered the SuperMicro 5019A-FTN4 SuperServer featuring the latest “Denverton” Intel Atom C3758 processor from Dihuni within days of the product being available. Working with Dihuni was an absolute pleasure, they kept us informed/updated on the order process and made sure to get the product shipped in time to meet our tight deadline. We are running several virtual machines on the Windows 2016 Hyper-V platform: Palo Alto Networks VM-100 firewall, Aruba Networks Virtual Mobility Controller, BlueCat DNS/DHCP servers and a few Windows Server 2012 & 2016 servers in our POC lab. We configured our system with 64GB of RAM and it’s easily handled the workload we’ve thrown at it. I highly recommend purchasing all of your SuperMicro server equipment from Dihuni, you won’t be disappointed!”
Robert Tesdall, Sr. Network Engineer, Enterprise Networks Solutions, USA
Can’t Find Your Product?
Our partners carry hundreds and thousands of products and it is possible we don’t have the one you are looking for listed online here. Please contact us with the product name and part number and we will respond with a quote asap.
“Supermicro’s new SuperServer based on the HGX-2 platform will deliver more than double the performance of current systems, which will help enterprises address the rapidly expanding size of AI models that sometimes require weeks to train,” said Charles Liang, president and CEO of Supermicro. “Our new HGX-2 system will enable efficient training of complex models. It combines sixteen Tesla V100 32GB SXM3 GPUs connected via NVLink and NVSwitch to work as a unified 2 PetaFlop accelerator with half a terabyte of aggregate GPU memory to deliver unmatched compute power.”
