Welcome to the NVIDIA product page of Magna Solutions!
With us you will find high-quality NVIDIA products that will take your data center or network to the next level. Whether you are looking for lightning-fast Infiniband solutions, powerful ConnectX adapters, advanced BlueField DPUs or reliable switches — We help you with personal advice, fast delivery and excellent service.
Explore our offerings, configure your own configuration, and experience NVIDIA performance, backed by the expertise of Magna Solutions.
All products are new!
NVIDIA - BLUEFIELD
NVIDIA BlueField-2 E-Series DPU 100GbE Dual-Port QSFP56, integrated BMC, PCIe Gen4 x16, Secure Boot
HPC: MBF2M516C-CESOT
NVIDIA - Connext X
NVIDIA ConnectX-7 HHHL Adapter Card, 100GbE, Dual-port QSFP112, PCIe 5.0 x16, Crypto Disabled, Secu
HPC: MCX713106AS-CEAT
NVIDIA ConnectX-7 HHHL Adapter Card, 100GbE, Dual-port QSFP112, PCIe 5.0 x16, Crypto Disabled, Secu
HPC: MCX713106AS-CEAT
Thanks to the tall bracket, it fits effortlessly in both full-height and low-profile chassis, perfect for modern data center and edge applications that require reliability and high throughput speeds.
Key Features
High-quality connectivity: 2×100 GbE (up to 200 Gbps aggregate) via QSFP112, backward compatible with 50/40/25/10/1 GbE
Top performances: PCIe 5.0 x16 (32 GT/s) – also Gen 4.0 compatible (16 GT/s)
Compact & flexible: HHHL format with tall bracket for full-height and low-profile enclosures
Secure & reliable: Hardware root-of-trust Secure Boot; inline encryption offload disabled for cost efficiency
Ultra-low latency: RDMA over Converged Ethernet (RoCE), NVIDIA GPUDirect Storage, ASAP²
Energy efficient: Only 16.8 W typical consumption (PCIe Gen 5.0 x16 + passive cables)
CONNECTX®-6 LX AND ADAPTER CARD, 25GBE, DUAL-PORT SFP28, PCIE 4.0 X8, CRYPTO AND SECURE BOOT, TALL BR
HPC: MCX631102AC-ADAT
Key Features:
High-quality connectivity: Dual-port SFP28 supports 1/10/25 GbE for flexible deployment in data centers
Lightning-fast PCIe: PCIe Gen 4.0 x8 (SERDES @ 16 GT/s) with backward compatibility to Gen 3.0
Secured & encrypted: IPsec inline crypto offload and hardware root of trust Secure Boot for maximum data integrity
Ultra-low latency: Advanced RoCE and ASAP² technology with line rate up to 75 Mpps for real-time performance
Compact & flexible: HHHL‑low‑profile card (96.30 × 68.90 mm) with tall bracket, suitable for full‑height and low‑profile chassis
Efficient & scalable: Low power consumption and high port density, perfect for modern HPC, cloud and edge infrastructures
NVIDIA BlueField-2 E-Series DPU 100GbE Dual-Port QSFP56, integrated BMC, PCIe Gen4 x16, Secure Boot
HPC: MCX75310AAS-NEAT
1×400 Gb/s InfiniBand & Ethernet (NDR/400 GbE): Single OSFP port supports both InfiniBand NDR and Ethernet speeds up to 400/200/100/50/40/10/1 Gb/s
PCIe 5.0 x16 (Gen 4 & 5): SERDES @ 16 GT/s (Gen 4) and 32 GT/s (Gen 5), backward compatible with Gen 3.0
Compact HHHL‑tall bracket: 68.90 × 167.65 mm, fits full-height and low-profile chassis; including tall bracket
Security & crypto-disabled: Hardware root-of-trust Secure Boot; inline encryption disabled for optimal price/performance ratio
Energy-efficient design: Typical 24.9 W consumption at PCIe Gen 5.0 x16 with passive cables; power supply via 12 V + 3.3 V AUX (max. 100 mA)
Robust conditions: Operating at 0°C–55°C; non-operating at –40°C–70°C
Ultra-low latency & offloads: Supports RDMA over Converged Ethernet, NVIDIA GPUDirect Storage, and ASAP² technology for maximum throughput
CONNECTX-6 LX AND ADAPTER CARD, 25GBE, DUAL-PORT SFP28, PCIE 4.0 X8, SECURE BOOT, NO CRYPTO, TALL BRA
HPC: MCX631102AS-ADAT
Optimize your network infrastructure with the half-height/half-long NVIDIA ConnectX‑6 Lx EN:
Dual-port 25 GbE SFP28: Flexible support for 1/10/25 GbE
PCIe 4.0 x8: Up to 16 GT/s per lane, backwards compatible
Secure design: Hardware‑root‑of‑trust Secure Boot
Crypto offload: Inline IPsec encryption for CPU offloading
Ultra-low latency: Advanced RoCE and ASAP² technology, up to 75 Mpps
Compact & flexible: HHHL format with tall bracket, fits full-height and low-profile chassis
Energy efficient: Low power consumption, ideal for densely populated rack environments
Perfect for demanding cloud, virtualization and HPC applications where reliability, security and throughput are essential.
CONNECTX®-6 DX AND ADAPTER CARD, 100GBE, DUAL-PORT QSFP56, PCIE 4.0 X16, NO CRYPTO, TALL BRACKET
HPC: MCX623106AN-CDAT
2×100 GbE via QSFP56: Dual‑port QSFP56 supports 1/10/25/40/50/100 GbE (up to 200 Gb/s aggregate)
PCIe 4.0 x16: 16 lanes @ 16 GT/s, also compatible with PCIe 3.0/2.0/1.1
Security: Inline crypto offload disabled, Secure Boot not supported (No Crypto, No Secure Boot)
Form Factor: Half-height/half-long (HHHL) with tall bracket; fits in full-height and low-profile chassis
Ultra-low latency & offloads: Supports RDMA over Converged Ethernet (RoCE), NVIDIA GPUDirect Storage, and ASAP² for CPU offloading
Power specifications: Typically ~19.5 W with PCIe 4.0 & passive cables; power supply via 3.3 V AUX @ 100 mA
Environment & Compliance: Operating 0°C–55°C; RoHS, CE, FCC compliant; altitude up to 3,050 m
Ideal for demanding cloud, virtualization and HPC applications where high throughput, low latency and cost-efficiency are key.
CONNECTX-5 AND NETWORK INTERFACE CARD FOR OCP, WITH HOST MANAGEMENT, 25GBE DUAL-PORT SFP28, PCIE3.0
HPC: MCX542B-ACAN
Optimize data center and cloud platforms with this compact, halogen-free, bracketless adapter card for OCP servers.
Dual-port 25 GbE SFP28 (1/10/25 GbE) via 2× SFP28 ports
OCP 2.0 Type 1: halogen-free, no-bracket form factor specifically for OCP rack servers
PCIe 3.0 x8: 8 lanes @8 GT/s, backward compatible with PCIe 2.0/1.1
Sub-750 ns latency & up to 200 Mpps: ultra-low latency and high message volumes with DPDK
Advanced offloads: NVMe‑over‑Fabric target, enhanced vSwitch/vRouter offloads, tag matching, rendezvous, RoCE
Host Management: NC‑SI over MCTP, PLDM for monitor & control, integrated BMC interface
Broad ecosystem: Support for RHEL/CentOS, Windows, VMware, OFED and WinOF‑2
Applications:
Ideal for Web 2.0, cloud, NFV, and storage workloads where reliability, low latency, and hardware offloads are critical.
NVIDIA - Switches Cumulus / Infiniband
SPECTRUM 100GBE 1U SWITCH W/CUMULUS LINUX, 16 QSFP28 PORTS, 2 AC PSUS, X86 2CORE, SHORT DEPTH, P2C A
HPC: MSN2100-CB2FC
High‑density 100 GbE connectivity: 16 QSFP28 ports support 1/10/25/40/50/56/100 GbE speeds for maximum flexibility and port density
Massive throughput & forwarding: 3.2 Tbps switching capacity combined with 2.38 billion packets per second line rate
Ultra-low latency: True cut-through architecture delivers only 300 ns port-to-port latency
Efficient energy consumption: Less than 6W per port (typically 94.3W with passive cables) for lower operating costs
Cumulus Linux-ready: Open Ethernet platform with Cumulus support via ONIE for maximum software freedom
Compact 1U short-depth design: 1.72″ × 7.87″ × 20″ (43.8 × 200 × 508 mm) ideal for cramped rack environments; P2C airflow for optimal cooling
Redundant power supply & built-in CPU: Dual AC power supplies and a Rangeley 2-core x86 host processor for in-switch services and management
Advanced network technologies: Support for RDMA, NVGRE, VXLAN, MLAG, QoS, SDN via Onyx/Cumulus, suitable for both spine and ToR roles
SPECTRUM BASED 25GBE/100GBE 1U OPEN ETHERNET SWITCH WITH CUMULUS LINUX, 18 SFP28 PORTS AND 4 QSFP28
HPC: MSN2010-CB2RC
High‑density port configuration: 18× SFP28 ports (1/10/25 GbE) and 4× QSFP28 ports (1/10/25/40/50/100 GbE) – fully splittable for maximum flexibility
Massive throughput speed: Up to 1.7 Tbps switching capacity and 1.26 Bpps line-rate forwarding for latency-sensitive workloads
Software‑defined networking: Powered by Cumulus Linux via ONIE, with freedom of NOS choice and no vendor lock-in
Compact & efficient cooling system: 1U short-depth chassis with C2P airflow, ideal for tight rack environments
Reliable hardware: Dual AC power supplies and built-in x86 quad-core CPU for in-switch services and management
Advanced networking features: Supports RDMA over Converged Ethernet (RoCE), NVMe‑over‑Fabric, MLAG, EVPN‑VXLAN, and extensive QoS mechanisms for modern data center architectures
Unlimited scalability & performance: Zero packet loss, sub-300 ns cut-through latency, and dynamic shared buffering for optimal resource allocation
Low energy consumption: Less than 5W per 100GbE port, contributing to lower operational costs
NVIDIA Spectrum-2 based 200GbE 1U Open Switch with Cumulus Linux, 32 QSF P56 ports, 2 Power Supplies
HPC: MSN3700-VS2FC
32×200 GbE QSFP56 ports: 32 QSFP56 ports with 10/25/40/50/100/200 GbE support per port via splitter cables
12.8 Tb/s switching & 8.33 Bpps: Massive throughput and line-rate packet processing for demanding data center environments
Ultra-low latency: Cut-through architecture delivers approximately 425 ns port-to-port latency
Compact 1U P2C chassis: 1-unit rackmount with Power-to-Connector airflow, standard depth and included airflow rail kit
Cumulus Linux via ONIE: Open networking through ONIE bootloader for maximum NOS freedom of choice
Redundant AC power supplies & x86 CPU: Dual hot-swap AC PSUs with integrated x86 processor for management and in-switch services
NVIDIA Quantum 2 based NDR InfiniBand Switch, 64 NDR ports, 32 OSFP ports, 2 Power Supplies (AC), S
HPC: MQM9700-NS2F
64x400 Gb/s NDR ports via 32 OSFP cages (splittable up to 128×200 Gb/s) with a non-blocking aggregate throughput of 25.6 Tbps
On-board management CPU: Intel® Core™ i3 Coffee Lake for in-switch management and monitoring
Ultra-high packet rate: More than 66.5 billion packets per second forwarding capacity for latency-sensitive HPC workloads
1U P2C airflow chassis: Power-to-Connector (forward) airflow in a 17.2″×1.7″×26″ (438 × 43.6 × 660 mm) rack-mount housing
Integrated management: Fully managed with MLNX‑OS and NVIDIA UFM for provisioning, troubleshooting, and fabric monitoring
Flexible cable options & power: Typically 640 W with passive cables; up to 1720 W with active cables; powered by dual hot-swap AC PSUs
Robust operating range: 0°C–35°C operational on forward airflow, non-op 40°C–70°C; up to 3,050 m altitude warranty
NVIDIA Quantum 2 based NDR InfiniBand Switch, 64 NDR ports, 32 OSFP ports, 2 Power Supplies (AC), S
HPC:MQM9700-NS2R
64×400 Gb/s NDR via 32 OSFP cages: Non‑blocking design with 64 ports NDR (400 Gb/s per port) across 32 OSFP ports, with a bidirectional aggregate bandwidth of 51.2 Tb/s.
Extremely high packet rate: Over 66.5 billion packets per second (Bpps) for line-rate forwarding in latency-sensitive workloads.
Managed switch with on-board subnet manager: Simple, out-of-the-box bring-up and management for up to 2,000 nodes.
Connector-to-Power (C2P) reverse airflow: Reverse airflow design (Connector‑to‑Power) for optimal cooling in cramped rack environments.
Redundant, hot-swappable power supply: 1+1 PSUs for maximum availability and maintenance without downtime.
Advanced In-Network Computing Offloads: Supports RDMA, NVIDIA SHARP™ acceleration, adaptive routing, congestion control, self-healing fabric, GPUDirect RDMA & Storage.
NVIDIA Services & Support
NVIDIA Enterprise Support Renewal – 1 Year
NVIDIA Professional Services – Installation & Configuration
NVIDIA Professional Services – Health Check & Optimization
Keep your NVIDIA environment up-to-date and supported with an Enterprise Support contract extension. This includes access to new firmware, software updates, and priority technical support from NVIDIA specialists.
Suitable for: All NVIDIA Enterprise products.
What can you expect?
Official NVIDIA Enterprise Support Extension
Access to new software updates & firmware upgrades
Priority access to NVIDIA technical support
Prevent support gaps and unexpected costs
One point of contact for support questions
Continuity and security for your NVIDIA environment
Have your new NVIDIA hardware or software professionally installed by certified engineers. We ensure a smooth implementation and configuration tailored to your environment.
Advantages: Faster live, fewer errors, optimal performance.
What can you expect?
Customised advice & architecture
Design of scalable GPU infrastructure
Installation & configuration by certified engineers
Best practices for maximum performance
Rapid implementation without downtime
Knowledge transfer & training IT team
One point of contact throughout the entire project
Annual inspection and tuning of your NVIDIA environment. Our specialists analyze performance, firmware updates, and configuration to ensure you always get the most out of your investment.
Inclusive: Reporting and improvement advice.
What can you expect?
Annual performance and usage analysis
Checking and applying the latest firmware updates
Configuration & architecture review
Report with concrete optimization advice
Recommendations for cost savings and growth
Proactive detection of risks or bottlenecks
Improved uptime and reliability
Bee Magna Solutions you will find everything you need to prepare your data center and corporate network for the future with high-quality NVIDIA hardwareWhether you want to expand, upgrade, or refresh your infrastructure, our solutions are designed for maximum performance, scalability, and reliability.
Complete NVIDIA solutions for your data center
Our offer includes, among other things:
NVIDIA Infiniband: For ultra-fast network connections with minimal latency, ideal for High Performance Computing (HPC), AI training, deep learning and data centers where speed and reliability are essential.
ConnectX network adapters: Powerful network cards that integrate seamlessly with your servers and switches, delivering high throughput and low CPU usage.
NVIDIA BlueField DPUsThe new standard for data center acceleration. Relieve your CPU, strengthen your security, and improve your performance with intelligent data-plane processing.
Datacenter switches: Scalable and efficient switches that work perfectly with Infiniband and Ethernet networks.
Why choose NVIDIA at Magna Solutions?
Our experts will help you select the right products for your specific workload and infrastructure. Whether it's a single server expansion or a complete HPC cluster, we'll ensure you benefit from the latest NVIDIA technology and configurations that truly meet your needs.
Bee Magna Solutions We combine our in-depth technical expertise with personalized service and fast delivery. You'll also benefit from:
Expert advice and custom configurations
Fast quotes and transparent prices
Implementation and maintenance support
Reliable delivery throughout the Netherlands and beyond
Contact us today!
Ready to harness the power of NVIDIA Infiniband, ConnectX, BlueField, and switches for your data center or network? Feel free to contact us for a free consultation or a competitive quote. We're happy to help!
Magna Solutions — Your partner for NVIDIA hardware, data center networking, and high-performance IT solutions.
Discover the benefits of NVIDIA Professional Services and Support Extensions
Bee Magna Solutions We go beyond just delivering NVIDIA products. With our NVIDIA Professional Services Benefit from expert advice, implementation, and optimization of your GPU infrastructure. We also help organizations with extending NVIDIA support contracts, so you are always assured of up-to-date support and maximum performance.
Whether you are looking for a NVIDIA contract renewalWhether you need advice on a new GPU solution or an annual health check for your existing NVIDIA environment, Magna Solutions offers everything under one roof. Contact us today for a customized quote and get the most out of your NVIDIA investment.