581 Search Results for "NVIDIA ConnectX-6"
-
Vendor:Cisco
Cisco UCSX-GPU-H100-NVL | H100 NVL PCIe Accelerator, 94GB HBM3, 400W, 2-Slot
UCSX-GPU-H100-NVL
Stock Status: Enquire
Vendor:CiscoCisco UCSX-GPU-H100-NVL | H100 NVL PCIe Accelerator, 94GB HBM3, 400W, 2-Slot
The Cisco UCSX-GPU-H100-NVL is an NVIDIA H100 NVL Tensor Core GPU designed for large language model inference, AI training, and high-performance computing workloads in enterprise data centers. Built on the Hopper architecture with 94GB of HBM3 memory and 3.9TB/s bandwidth, it delivers exceptional performance for transformer models and generative AI applications. The 2-slot FHFL form factor integrates into Cisco UCS X-Series servers with PCIe Gen5 connectivity. -
Vendor:Cisco
Cisco HCI-GPU-A30 | A30 24GB HBM2 GPU, PCIe 4.0, Passive, 165W
HCI-GPU-A30
Stock Status: Enquire
Vendor:CiscoCisco HCI-GPU-A30 | A30 24GB HBM2 GPU, PCIe 4.0, Passive, 165W
The Cisco HCI-GPU-A30 delivers NVIDIA A30 Tensor Core GPU compute acceleration for AI inference, training, and HPC workloads in mainstream enterprise servers. Built on NVIDIA Ampere architecture with 3,584 CUDA cores and 24GB HBM2 memory, it provides 933 GB/s memory bandwidth and supports Multi-Instance GPU technology to partition into up to four isolated instances. The dual-slot passive design with 165W TDP integrates seamlessly into Cisco HCI and UCS C-Series rack servers. -
Vendor:Cisco
Cisco UCSC-GPU-A30 | A30 Tensor Core GPU, 24GB HBM2, 180W Passive
UCSC-GPU-A30
Stock Status: Enquire
Vendor:CiscoCisco UCSC-GPU-A30 | A30 Tensor Core GPU, 24GB HBM2, 180W Passive
The Cisco UCSC-GPU-A30 delivers enterprise-grade AI inference and HPC acceleration for mainstream rack servers. Built on NVIDIA Ampere architecture with 3584 CUDA cores and 224 Tensor Cores, this passive-cooled GPU provides 24GB HBM2 memory with 933GB/s bandwidth in a power-efficient 180W PCIe Gen 4.0 design. Multi-Instance GPU technology enables secure workload partitioning for AI inference, deep learning training, and scientific computing applications. -
Vendor:Blue Cable Co
MC3208011-SX-BCC | NVIDIA Mellanox Compatible 1GB Multi-Mode SFP 550m (1000BASE-SX)
MC3208011-SX-BCC
In stock
Vendor:Blue Cable CoMC3208011-SX-BCC | NVIDIA Mellanox Compatible 1GB Multi-Mode SFP 550m (1000BASE-SX)
The MC3208011-SX-BCC is a Blue Cable Co 1GbE multi-mode SFP optical transceiver, compatible with NVIDIA Mellanox platforms. Using an 850nm VCSEL laser transmitter and duplex LC connector, it supports reaches of up to 550m on 50/125um MMF or up to 300m on 62.5/125um MMF. This plug-and-play module is tested in our state-of-the-art lab in Sydney, Australia to verify NVIDIA Mellanox platform compatibility, and is backed by Blue Cable Co's 5-year advanced hardware replacement warranty.
- Plug-and-play SFP form factor — no additional configuration required
- 850nm oxide VCSEL laser transmitter for reliable short-range multi-mode operation
- Up to 550m reach on 50/125um MMF, up to 300m on 62.5/125um MMF (per datasheet, depends on fibre and link conditions)
- Duplex LC connector for standard multi-mode fibre patch leads
- Hot-pluggable design for live insertion without system downtime
- Operating case temperature range: -20°C to +85°C
- Single 3.3V power supply, low power dissipation
- RoHS compliant and lead free
- Tested in Blue Cable Co's Sydney lab for NVIDIA Mellanox platform compatibility
- Backed by Blue Cable Co's 5-year advanced hardware replacement warranty
-
Vendor:NVIDIA
NVIDIA 980-9IAM3-00X003 | MCA4K50 | 1600G OSFP Active Copper Cable | Reach 3m | 1.6T Ethernet | Connector: OSFP‑to‑OSFP
980-9IAM3-00X003
Stock Status: Enquire
Vendor:NVIDIANVIDIA 980-9IAM3-00X003 | MCA4K50 | 1600G OSFP Active Copper Cable | Reach 3m | 1.6T Ethernet | Connector: OSFP‑to‑OSFP
Active copper OSFP cable delivering 1.6T aggregate bandwidth for dense data center fabrics. Integrated heat sinks on both ends support reliable thermals at 3m. Optimized for short‑reach, low‑latency Ethernet links in AI/HPC clusters and spine‑leaf architectures. -
Vendor:NVIDIA
NVIDIA 980-9I92N-00N003 | 800G to 2x400G OSFP AOC | 850nm MMF | 3 m | OSFP-OSFP | IB/Ethernet breakout
980-9I92N-00N003
Stock Status: Enquire
Vendor:NVIDIANVIDIA 980-9I92N-00N003 | 800G to 2x400G OSFP AOC | 850nm MMF | 3 m | OSFP-OSFP | IB/Ethernet breakout
Active optical breakout cable that splits one 800G OSFP port into two 400G OSFP links. Designed for short-reach data center interconnects with 850 nm multimode optics and hot‑pluggable OSFP ends. Ideal for building high‑bandwidth InfiniBand NDR or 400/800GbE fabrics over a 3 m reach. -
Vendor:NVIDIA
NVIDIA 980-9I580-00N030 | MPO-12 trunk, SMF OS2, 40m, 1310nm PSM4/PLR4, MPO–MPO, yellow jacket
980-9I580-00N030
Stock Status: Enquire
Vendor:NVIDIANVIDIA 980-9I580-00N030 | MPO-12 trunk, SMF OS2, 40m, 1310nm PSM4/PLR4, MPO–MPO, yellow jacket
Single‑mode OS2 MPO‑12 trunk for parallel optics over longer distances. The 40 m yellow cable supports 4‑lane PSM4/PLR4 100G links and parallel single‑mode deployments in campus and data‑center interconnects. Factory‑terminated for consistent performance and straightforward high‑density patching. -
Vendor:NVIDIA
NVIDIA 980-9I149-00CS00 | 100GbE | QSFP28 | 100GBASE-SR4 | 850 nm | 100 m OM4 (70 m OM3) | MPO-12
980-9I149-00CS00
Stock Status: Enquire
Vendor:NVIDIANVIDIA 980-9I149-00CS00 | 100GbE | QSFP28 | 100GBASE-SR4 | 850 nm | 100 m OM4 (70 m OM3) | MPO-12
High-density 100G short‑reach optical module for leaf/spine fabrics and TOR aggregation. QSFP28 SR4 runs at 850 nm over MMF with MPO-12 connector, delivering plug-and-play 100G or 4x25G breakout in campus and data center lanes up to 100 m on OM4. -
Vendor:NVIDIA
NVIDIA 980-9I08O-00W010 | QSFP-DD 400G AOC | 8x50G PAM4 | 850nm MMF | 10m reach | QSFP-DD to QSFP-DD
980-9I08O-00W010
Stock Status: Enquire
Vendor:NVIDIANVIDIA 980-9I08O-00W010 | QSFP-DD 400G AOC | 8x50G PAM4 | 850nm MMF | 10m reach | QSFP-DD to QSFP-DD
10‑meter 400G QSFP‑DD active optical cable designed for flexible rack‑to‑rack or row deployments. Delivers consistent latency and signal integrity using 8x50G PAM4 lanes over integrated multimode fiber at 850nm, minimizing EMI and simplifying installation in high‑density environments. -
Vendor:Juniper
Juniper SRX5K-4XGE-XFP | SRX5000 I/O card | 4x10GbE XFP ports | optics required | HA-ready
SRX5K-4XGE-XFP
Stock Status: Enquire
Vendor:JuniperJuniper SRX5K-4XGE-XFP | SRX5000 I/O card | 4x10GbE XFP ports | optics required | HA-ready
High-density 10GbE Ethernet I/O Card (IOC) for SRX5000 providing 4x10GbE via XFP cages for data center and service provider edge connectivity. Requires XFP transceivers (sold separately). Hot-swappable and cluster-ready for SRX5400/5600/5800 deployments. -
Vendor:NVIDIA
NVIDIA 980-9IAM5-00X001 | MCA4K50 | 1600G OSFP Active Copper Cable | Reach 1.1m | 1.6T Ethernet | Connector: OSFP‑to‑OSFP
980-9IAM5-00X001
Stock Status: Enquire
Vendor:NVIDIANVIDIA 980-9IAM5-00X001 | MCA4K50 | 1600G OSFP Active Copper Cable | Reach 1.1m | 1.6T Ethernet | Connector: OSFP‑to‑OSFP
Short‑reach 1.6T OSFP active copper cable with integrated heat sinks on both ends for superior thermal control in tightly packed racks. The 1.1m length is ideal for within‑rack connections that demand the lowest latency and minimal power. -
Vendor:Blue Cable Co
SFP-1GTXRJ45-T-BCC | Moxa Compatible 1GB Ethernet RJ45 100m (1000BASE-T)
SFP-1GTXRJ45-T-BCC
In stock
Vendor:Blue Cable CoSFP-1GTXRJ45-T-BCC | Moxa Compatible 1GB Ethernet RJ45 100m (1000BASE-T)
The SFP-1GTXRJ45-T-BCC is a drop-in compatible replacement for the Moxa SFP-1GTXRJ45-T. This copper SFP transceiver provides plug-and-play Gigabit Ethernet connectivity for Moxa industrial switches and networking equipment over standard twisted-pair cabling. - IEEE 802.3ab 1000BASE-T compliance guarantees interoperability with any Gigabit Ethernet copper infrastructure - 10/100/1000Mbps auto-negotiation over Cat5 or Cat5e cable, supporting up to 100m transmission distance - Industrial temperature rating from -40 to 75°C for deployment in harsh environments including automation, transportation, and outdoor installations - Hot-pluggable SFP form factor with standard RJ45 connector for immediate integration without downtime - SFP MSA compliant design with metallic enclosure for low EMI - Compatible with Moxa ICS, IKS, EDS, PT, EDR, and IM series Ethernet switches requiring Gigabit copper SFP modules -
Vendor:Blue Cable Co
Q28-100GBIDI-BCC | Dell Compatible 40/100GB Multi-Mode BiDi - 100m (40GBASE-SR-BD / 100GBASE-SR-BD)
Q28-100GBIDI-BCC
In stock
Vendor:Blue Cable CoQ28-100GBIDI-BCC | Dell Compatible 40/100GB Multi-Mode BiDi - 100m (40GBASE-SR-BD / 100GBASE-SR-BD)
The Q28-100GBIDI-BCC is a drop-in compatible replacement for the Dell Q28-100GBIDI and 407-BBYM. This BiDi transceiver delivers plug-and-play 40/100GB connectivity over duplex LC multi-mode fibre on Dell networking platforms. - 40GBASE-SR-BD and 100GBASE-SR-BD compliance ensures interoperability with BiDi-enabled QSFP28 and QSFP ports - QSFP28 form factor, dual-rate 40/100GB operation, 850nm/900nm BiDi wavelengths, duplex LC connector - 100m reach over OM4 fibre (70m over OM3), enabling infrastructure reuse from existing duplex MMF deployments - DDM support provides real-time monitoring of temperature, laser bias current, optical power, and supply voltage - Hot-swappable design with commercial temperature range (0 to 70C) - PAM4 modulation using two bidirectional 50Gbps lanes (2x50G) for 100GB mode -
Vendor:NVIDIA
NVIDIA MFS1S50-H030V | NVIDIA Mellanox active optical cable, 200Gb/s to 2x100Gb/s IB HDR, QSFP56 to 2x QSFP56, 30m
MFS1S50-H030V
Stock Status: Enquire
Vendor:NVIDIANVIDIA MFS1S50-H030V | NVIDIA Mellanox active optical cable, 200Gb/s to 2x100Gb/s IB HDR, QSFP56 to 2x QSFP56, 30m
All products sold by XS Network Tech include a 12-month warranty on both new and used items. Our in-house technical team thoroughly tests used hardware prior to sale to ensure enterprise-grade reliability. -
Vendor:Juniper
Juniper SRX650-BASESRE6-645AP | 4x 1G RJ45; 645W AC PoE PSU (247W PoE budget); NGFW/UTM; SD-WAN capable; LTE optional; license: base
SRX650-BASESRE6-645AP
Stock Status: Enquire
Vendor:JuniperJuniper SRX650-BASESRE6-645AP | 4x 1G RJ45; 645W AC PoE PSU (247W PoE budget); NGFW/UTM; SD-WAN capable; LTE optional; license: base
Branch/perimeter services gateway with SRE 6 compute, four onboard 10/100/1000BASE‑T ports, and an AC PoE power supply delivering a 247 W PoE budget for supported PoE modules. Consolidates routing, VPN, and next‑gen firewall in a compact chassis with rack kit included. -
Vendor:Arista
Arista DCS-7130-96S-F | 7130 Connect 96-Port 10G SFP+ Layer 1 Switch, 6ns
DCS-7130-96S-F
Stock Status: Enquire
Vendor:AristaArista DCS-7130-96S-F | 7130 Connect 96-Port 10G SFP+ Layer 1 Switch, 6ns
The Arista DCS-7130-96S-F is a high-performance Layer 1 switch designed for ultra-low latency network applications. With 96 SFP+ ports delivering port-to-port latency of just 6 nanoseconds, this 2U device is ideal for financial trading, network visibility, and high-frequency data distribution. It features full signal recovery and regeneration with non-blocking architecture, front-to-rear airflow, and redundant hot-swappable power supplies for enterprise-grade reliability. -
Vendor:NVIDIA
NVIDIA MFS1S50-H020V | NVIDIA Mellanox active optical cable, 200Gb/s to 2x100Gb/s IB HDR, QSFP56 to 2x QSFP56, 20m
MFS1S50-H020V
Stock Status: Enquire
Vendor:NVIDIANVIDIA MFS1S50-H020V | NVIDIA Mellanox active optical cable, 200Gb/s to 2x100Gb/s IB HDR, QSFP56 to 2x QSFP56, 20m
All products sold by XS Network Tech include a 12-month warranty on both new and used items. Our in-house technical team thoroughly tests used hardware prior to sale to ensure enterprise-grade reliability. -
Vendor:Cisco
Cisco CAI-GPU-H100-NVL | H100 NVL 94GB HBM3 GPU, 400W, PCIe 5.0 x16, Passive
CAI-GPU-H100-NVL
Stock Status: Enquire
Vendor:CiscoCisco CAI-GPU-H100-NVL | H100 NVL 94GB HBM3 GPU, 400W, PCIe 5.0 x16, Passive
The Cisco NVIDIA H100 NVL is an enterprise-grade GPU accelerator built on the Hopper architecture, designed for AI inference, large language model deployment, and high-performance computing workloads. Featuring 94GB of HBM3 memory with 3.9TB/s bandwidth and fourth-generation Tensor Cores, it delivers exceptional performance for generative AI, deep learning training, and data analytics at scale. The 2-slot full-height, full-length form factor with passive cooling and PCIe Gen5 x16 interface makes it ideal for multi-GPU datacenter server deployments requiring maximum computational throughput. -
Vendor:Cisco
Cisco HX-GPU-P100-12G | Tesla P100 GPU Accelerator, 12GB HBM2, PCIe 3.0
HX-GPU-P100-12G
Stock Status: Enquire
Vendor:CiscoCisco HX-GPU-P100-12G | Tesla P100 GPU Accelerator, 12GB HBM2, PCIe 3.0
The Cisco HX-GPU-P100-12G is an NVIDIA Tesla P100 GPU accelerator designed for Cisco HyperFlex hyperconverged infrastructure. Built on Pascal architecture with 3584 CUDA cores and 12GB HBM2 memory, this PCIe 3.0 x16 card delivers 9.3 teraflops single-precision and 4.7 teraflops double-precision performance for AI, deep learning, and high-performance computing workloads. The passively cooled design integrates seamlessly into data center environments requiring accelerated compute for complex analytical and scientific applications. -
Vendor:Cisco
Cisco UCSC-GPU-H200-NVL | H200 NVL GPU, 141GB HBM3e, 4.8TB/s, 600W PCIe 5.0
UCSC-GPU-H200-NVL
Stock Status: Enquire
Vendor:CiscoCisco UCSC-GPU-H200-NVL | H200 NVL GPU, 141GB HBM3e, 4.8TB/s, 600W PCIe 5.0
The Cisco UCSC-GPU-H200-NVL is an NVIDIA H200 NVL Tensor Core GPU built on the Hopper architecture, delivering 141GB of HBM3e memory with 4.8TB/s bandwidth for AI inference, generative AI, and high-performance computing workloads. This PCIe 5.0 accelerator features 600W TDP, NVLink support for multi-GPU configurations, and fits standard air-cooled enterprise rack designs. Ideal for large language model training, deep learning, scientific simulations, and memory-intensive AI applications requiring massive memory capacity and throughput. -
Vendor:NVIDIA
NVIDIA MFS1S00-H010V | 200G QSFP56 AOC | IB HDR | 850nm | 10m | QSFP56-to-QSFP56
MFS1S00-H010V
Stock Status: Enquire
Vendor:NVIDIANVIDIA MFS1S00-H010V | 200G QSFP56 AOC | IB HDR | 850nm | 10m | QSFP56-to-QSFP56
Longer-reach 200Gb/s InfiniBand HDR active optical cable for GPU clusters, storage fabrics, and spine–leaf deployments. QSFP56-to-QSFP56 AOC simplifies cabling while delivering low latency and high signal integrity across 10 meters. LSZH jacket with black pull tab for safe, maintainable installs. -
Vendor:Cisco
Cisco UCSC-GPUV100SXM32 | Tesla V100 GPU Accelerator, 32GB HBM2, SXM2 300W
UCSC-GPUV100SXM32
Stock Status: Enquire
Vendor:CiscoCisco UCSC-GPUV100SXM32 | Tesla V100 GPU Accelerator, 32GB HBM2, SXM2 300W
The Cisco UCSC-GPUV100SXM32 is an NVIDIA Tesla V100 SXM2 GPU accelerator with 32GB HBM2 memory, designed for AI training, deep learning, and high-performance computing workloads in enterprise data centers. Powered by NVIDIA Volta architecture with 5,120 CUDA cores and 640 Tensor cores, this passively cooled module delivers 125 TFLOPS of tensor performance and 900GB/s memory bandwidth. The SXM2 form factor supports NVLink 2.0 interconnect for multi-GPU scaling up to 300GB/s, ideal for demanding machine learning and scientific simulation applications in Cisco UCS C-Series rack servers. -
Vendor:Cisco
Cisco CAI-NVL2-H100 | NVLink 2-Way Bridge for H100 NVL GPU Interconnect
CAI-NVL2-H100
Stock Status: Enquire
Vendor:CiscoCisco CAI-NVL2-H100 | NVLink 2-Way Bridge for H100 NVL GPU Interconnect
A high-speed NVLink bridge enabling direct GPU-to-GPU communication between two NVIDIA H100 NVL PCIe cards, delivering 600 GB/s bidirectional bandwidth—10x faster than PCIe Gen4. Designed specifically for AI training, large language model inference, and HPC workloads requiring memory-intensive multi-GPU configurations, this bridge maximizes application performance when deploying H100 NVL pairs in standard PCIe servers. -
Vendor:NVIDIA
NVIDIA 980-9I556-00N020 | MPO trunk, 2x MPO-12, OM4 MMF 850nm, 20 m
980-9I556-00N020
Stock Status: Enquire
Vendor:NVIDIANVIDIA 980-9I556-00N020 | MPO trunk, 2x MPO-12, OM4 MMF 850nm, 20 m
High-density OM4 multimode MPO trunk assembly for parallel-optics links and patching. Designed for low-loss connectivity between patch panels, cassettes, and transceiver-breakout applications in modern data centers.
