你的購物車是空的

${{ (item.variation.media ? item.variation.media.alt_translations : item.product.cover_media.alt_translations) | translateModel }}$ ${{ (item.variation.media ? item.variation.media.alt_translations : item.product.cover_media.alt_translations) | translateModel }}$

{{ childProduct.title_translations | translateModel }}

{{ getChildVariationShorthand(childProduct.child_variation) }}

{{ getSelectedItemDetail(selectedChildProduct, item).childProductName }} x {{ selectedChildProduct.quantity || 1 }}

{{ getSelectedItemDetail(selectedChildProduct, item).childVariationName }}

{{item.quantity}}x HK$0 {{ item.unit_point }} 點

刪除

${{addonItem.product.cover_media.alt_translations | translateModel}}$

{{addonItem.quantity}}x {{ mainConfig.merchantData.base_currency.alternate_symbol + "0" }}

刪除

NVIDIA MCA4K50 980-9IAM5-00X001 1600Gbps to 1600Gbps OSFP Active Copper Cable

The NVIDIA MCA4K50 (980-9IAM5-00X001) is a 1.1m Active Copper Cable built for InfiniBand XDR clusters. It delivers a 1600Gb/s point-to-point connection over eight 200G-PAM4 lanes with a symmetrical IHS-to-IHS thermal profile optimized for linking Quantum-3 switches.

全店，Free Shipping

{{shoplineProductReview.avg_score}}( {{shoplineProductReview.total}} {{'product.product_review.reviews' | translate}} )

{{amazonProductReview.avg_rating}}( ({{amazonProductReview.total_comment_count}} {{'product.product_review.reviews' | translate}}) )

一次最大商品購買數量限制為 99999

該數量不適用，請填入有效的數量。

售完

若想購買，請聯絡我們。

商品描述

送貨及付款方式

顧客評價

商品描述

NVIDIA MCA4K50 Active Copper Cable

Name: NVIDIA MCA4K50 980-9IAM5-00X001 1600Gbps to 1600Gbps OSFP Active Copper Cable
Brand: Nvidia
SKU: 6948ad78b6735540d266076c
Availability: InStock

1600Gb/s | OSFP to OSFP | 1.1m | IHS to IHS

Product Overview

The NVIDIA MCA4K50 (MPN: 980-9IAM5-00X001) is a flagship, ultra-high-density Active Copper Cable (ACC) specifically engineered to orchestrate the backend fabric of next-generation AI pipelines and high-performance computing (HPC) ecosystems. Providing a staggering point-to-point throughput of 1600Gb/s (1.6Tb/s), this premium 1.1-meter link allows high-capacity data center devices to communicate seamlessly at bare-metal speeds.

Utilizing eight high-speed electrical copper pairs operating on advanced 200G-PAM4 modulation, the MCA4K50 features specialized onboard active signal conditioning. This enables stable multi-terabit transmission over its 1.1-meter length without the significant power draw, thermal overhead, and premium cost associated with traditional optical transceivers.

Key Features

Massive 1.6Tb/s Line Rate: Supports full wire-speed performance utilizing 8 independent channels of state-of-the-art 200G-PAM4 technology.
InfiniBand XDR Optimized: Onboard firmware is explicitly tailored and tuned to unleash the full potential of low-latency InfiniBand XDR architecture.
Symmetrical IHS-to-IHS Form Factor: Equipped with dual Integrated Heat Sink (IHS / Closed Finned Top) connectors on both ends, purpose-built to slide into finned-cage ports.
Active Equalization Integration: Active internal circuitry counteracts signal attenuation and jitter, guaranteeing a pristine Bit Error Rate (BER) across high-frequency interconnections.

Technical Specifications

Parameter	Specification
Manufacturer	NVIDIA
Part Number (MPN)	980-9IAM5-00X001
Model	MCA4K50
Cable Type	Active Copper Cable (ACC / Active DAC)
Data Rate	1600Gb/s (1.6T Aggregate)
Modulation	8x 200G-PAM4
Connector Configuration	OSFP to OSFP (Point-to-Point)
Thermal Profile	Closed Finned Top to Closed Finned Top (IHS to IHS)
Length	1.1 Meters (3.61 feet)
Wire Gauge	26 AWG
Jacket Material	LSZH (Low Smoke Zero Halogen)
Power Consumption	Maximum: 27W (Fully Active Subsystem)
Standards Compliance	OSFP MSA, CMIS, SFF-8665, SFF-8636, RoHS

Hardware Compatibility & Applications

Quantum-3 Switch Interconnects: Purpose-built to link adjacent NVIDIA Quantum-3 XDR InfiniBand switches together in scale-up structures where both interconnecting port cages feature a finned profile.

Enterprise AI Supercomputers: Serves as the critical, zero-latency physical plumbing required to build high-capacity switch fabrics for large-scale GPU clustering and deep learning training workloads.