Imagens apenas para referência
BCM56850A2KFSBG
+Lista de MateriaisIC ETHERNET SWTCH 10G MULT LAYER
-
FabricanteBroadcom Limited
-
Peça do Fabricante #BCM56850A2KFSBG
-
Pacote BGA
-
Em Estoque5807
365 Garantia de Qualidade em 1 dia
7*24 Garantia de serviço em 24 horas
90-Garantia de pós-venda em 1 dia
Garantia de Produto Genuíno
Especificações
| Atributo | Valor |
| Package / Case | Tray |
| Part Status | Obsolete |
| Protocol | Ethernet |
| Function | Switch |
Visão Geral
Description
Key features:
- Advanced programmability with support for P4 and Broadcom’s SDK.
- Low latency & high efficiency for AI/ML workloads.
- Flexible port configurations (1G to 400G).
- Integrated deep buffering for congestion management.
- Hardware-based telemetry for real-time analytics.
Built on 16nm process technology, it optimizes power efficiency while delivering high throughput. It supports RoCEv2, VXLAN, and EVPN for modern cloud architectures.
Applications:
- Cloud data centers
- AI/ML clusters
- High-performance computing (HPC)
This chip is part of Broadcom’s Tomahawk+ lineup, succeeding the Tomahawk series with enhanced scalability and programmability.
Equivalent
- Broadcom BCM56850 series variants (e.g., BCM56850A0, BCM56850A1)
- Marvell (formerly Cavium) XPliant CNX880xx series
- Mellanox Spectrum (now NVIDIA) for similar high-speed switching
- Intel (Barefoot Networks) Tofino for programmable switches
These offer comparable features like 10/40/100G Ethernet switching. Exact replacements depend on specific requirements.
Features
- Port Configuration: 32x 100GbE, 64x 50GbE, or 128x 25GbE ports.
- Throughput: Up to 3.2 Tbps full-duplex.
- Advanced Switching: Supports Layer 2/3 forwarding, VXLAN/NVGRE tunneling, and ECMP.
- Buffering: Deep packet buffers for congestion management.
- Flexible Pipeline: Programmable packet processing.
- Low Latency: Optimized for high-speed data centers.
- Energy Efficiency: Advanced power-saving modes.
- SDN Support: OpenFlow and other programmable interfaces.
Ideal for cloud, hyperscale, and enterprise networks.
Pinout
- High-speed switching: Supports up to 3.2 Tbps throughput.
- Port interfaces: 64x 10G/25G/40G/50G/100G (via SerDes).
- Protocols: Ethernet, VLAN, QoS, ACLs, and advanced telemetry.
- Integration: On-chip packet buffering, traffic management, and deep packet inspection.
Used in data center and enterprise switches for scalable, high-performance networking.
Manufacturer
Application
- Data Center Networking: High-speed switching for cloud and enterprise data centers.
- Enterprise Networks: Core and aggregation layer switching.
- Service Provider Infrastructure: Carrier-grade Ethernet solutions.
- High-Performance Computing (HPC): Low-latency interconnects.
- AI/ML Clusters: Scalable fabric for distributed workloads.
It supports advanced features like 100G/400G Ethernet, deep buffering, and programmable pipelines, making it ideal for scalable, high-bandwidth environments.