Solid State Drive (SSD)

M.2
SATA
SAS 12G/24G NEW
U.2 NEW
PCIE
Memory Module or RAM (Random Access Memory)

UDIMM
ECC DIMM
SODIMM
RDIMM NEW
LRDIMM
Network Card / Intelligent Network Card

ConnectX-4 Ethernet Card
ConnectX-5 Ethernet Card NEW
ConnectX-6 Ethernet Card NEW
ConnectX-5 VPI InfiniBand Network Card
ConnectX-6 VPI InfiniBand Network Card
Mellonox - DPU
Bus HBA / Disk RAID

12Gb/s Sata/SAS HBA Disk
12Gb/s Sata/SAS/NVMe HBA Disk
24Gb/s Sata/SAS/NVMe HBA Disk NEW
12Gb/s Sata/SAS RAID Disk
12Gb/s Sata/SAS/NVMe RAID Disk
24Gb/s Sata/SAS/NVMe RAID Disk NEW
Switch

SB78X0 EDR InfiniBand Switch
QM87X0 EDR InfiniBand Switch
Quantum MetroX-2
QM97X0 NDR InfiniBand Switch NEW
SN3000 Ethernet Switch
SN4000 Ethernet Switch NEW
SN5000 Ethernet Switch
End-of-Life (EOL)

SAMSUNG
SOLIDIGM
NVIDIA
BROADCOM
INTEL

Samsung releases Gen5 SSD: Performance uation and application advantages of two PM1743 and PM9D3a models

With the rapid development of big data, cloud computing, and 5G technology, the demand for data computing and efficient storage is constantly increasing. As a leading global supplier of storage device technology, Samsung has officially released the PM9D3a solid-state drive. As a cutting-edge PCIe Gen5.0 NVMe SSD, the PM9D3a provides unparalleled performance and energy efficiency for enterprise servers and multi-core processing data systems. With Samsung's eighth generation 3D V-NAND technology, the PM9D3a can achieve sequential read and write speeds of 12000MB/s and 6800MB/s, respectively. Its capacity of up to 15.36TB and diverse specification options ensure that it can meet the storage needs of various data centers. The new generation PM9D3a not only improves performance, but also helps data centers reduce operating costs and effectively address environmental challenges through improved heat dissipation and power efficiency design. This SSD is an ideal choice for adapting to AI, machine learning, and large-scale data processing, ushering in a new era of data storage.

Router and gateway system

By utilizing InfiniBand routers, InfiniBand Long Distance Connectivity (NVIDIA MetroX®-2), and InfiniBand to Ethernet gateway systems (NVIDIA Skyway™), InfiniBand systems can deliver exceptional scalability and subnet isolation capabilities. When needed, the latter can efficiently and scalably connect high-performance, low-latency InfiniBand data centers to external Ethernet infrastructure, such as Ethernet storage.
InfiniBand Switch

InfiniBand switch systems provide exceptionally high performance and port density. Innovative features such as NVIDIA Scalable Hierarchical Aggregation and Reduction Protocol (SHARP)™, advanced management capabilities including self-healing network functionality, quality of service, enhanced virtual channel mapping, and NVIDIA network computing acceleration engine enhance the performance of industrial, AI, and scientific applications.
Data Processor (DPU)

NVIDIA ® BlueField ® With powerful computing power, high-speed network and extensive programmability, DPU can provide software defined and Hardware acceleration solutions for demanding workloads. From accelerating AI and scientific computing to cloud native supercomputing, BlueField redefines possibilities.
InfiniBand network card

InfiniBand Network Card (HCA) provides ultra-low latency, ultra-high throughput, and innovative NVIDIA network computing engine, providing the acceleration, scalability, and feature-rich technology required for today's modern workloads.

News

[Exhibition Preview] Saiten invites you to gather at the 2025AI Computing Conference!
The hosting of the 2025 AI Computing Industry CITE Conference is not only a comprehensive display of the development achievements of China's electronic information industry in the past year, but also a profound insight into the future trends of AI technology development. In this era full of opportunities and challenges, let us look forward to the arrival of the 2025 AI Computing Industry Conference and CITE 2025, witness the glorious moments of the AI industry together, and explore the infinite possibilities of AI technology together.
Saiten | Construction starts in 2025!!
The Lunar New Year is a festival full of hope and joy for every Chinese child. At this moment of reunion and celebration, the eighth day of the lunar new year is particularly important because it marks the start of work after the holiday and symbolizes the flourishing development of career and the health and happiness of family in the new year.

Faq

You can query the solutions to our product-related problems and conduct business consultation here
Driver

For all product driver inquiries or downloads, please click here for detailed information.
Datasheet

For all product specification inquiries or downloads, please click below for more details.
Anti-Counterfeiting

Anti-Counterfeiting Verification, coming soon! Currently under development.
After-ale Service

No matter where you purchase and use our products, if there is a malfunction within the warranty scope, you can present the purchase invoice

Culture

Our core values are: loyalty, gratitude, learning, innovation, and hard work!
distributors

Our reseller community spans the world, which makes it easy for you to get the products you need to drive your business forward. You can find an LR-LINK reseller using the list below.
Contact Us

For business cooperation or other questions, please click here to contact us.

Back Home Product Center Network Card / Intelligent Network Card ConnectX-5 Ethernet Card

MCX512A-ACUT

Nvidia MCX512A-ACUT ConnectX-5 EN Adapter Card 10/25GbE Dual Port Ethernet Network Card

Key Features

The ConnectX-5 Ethernet network interface card has up to two 10/25GbE connection ports, 750ns latency, up to 200 million messages per second (Mpps), and a development kit (DPDK). For storage workloads, ConnectX-5 provides a series of innovative accelerations, such as signature switching (T10-DIF) in hardware, embedded PCIe switches, and NVMe over Fabric target offloading. The ConnectX-5 adapter card also brings advanced Open vSwitch offloading to telecommunications and cloud data centers to drive extremely high packet rates and throughput, while reducing CPU resource consumption and improving the efficiency of data center infrastructure.

Datasheet:

Overview
Specifications
Download
Driver
Relative Products

Product characteristics:
Label matching and assembly point unloading
Reliable adaptive routing for transmission
Burst buffer offloading for backend checkpoints
NVMe over Fabric uninstallation
Embedded PCIe
Enhanced vSwitch/vRouter uninstallation
Flexible pipeline
Overlay network RoCE
PCIe Gen 4.0 support
Compliant with RoHS standards
ODCC compatible
Product advantages:
Industry leading throughput, low latency, low CPU utilization, and high message rate
Innovative rack design for storage and machine learning based on host chain technology
Intelligent interconnection for x86, Power, Arm, and GPU based computing and storage platforms
Advanced storage features, including NVMe over Fabric offloading
Intelligent network adapter supporting flexible pipeline programmability
The cutting-edge performance of virtualized networks, including Network Function Virtualization (NFV)
Efficient I/O consolidation, reducing data center costs and complexity
&Nbsp;
Note: The copyright (in whole or in part) of images or videos related to NVIDIA products belongs to NVIDIA Corporation.

Model Name	MCX512A-ACUT
Rate	1/10/25GbE
Interface Type	Dual Port SFP28
Application Area	Ethernet
Hardware Interface	PCIe 3.0 x8
RoCE Support	Yes
Chip Type	ConnectX-5 EN
RoHS	Yes
Operating System	RHEL/CentOS, Windows, FreeBSD, VMware, OFED, WinOF-2
Link Rate	8.0 GT/s