{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreiaqqermtmjnv5jx5j6n4g6oyize4334jt7e2vlsa24zxekkgv7xna",
    "uri": "at://did:plc:qzjwstutqk2cy7df7jbzd2hx/app.bsky.feed.post/3mhymkkfa36o2"
  },
  "coverImage": {
    "$type": "blob",
    "ref": {
      "$link": "bafkreif6l477dh256eps36jgcs6wq5qdfkybcqcojvcwf3egraec6pwox4"
    },
    "mimeType": "image/jpeg",
    "size": 10483233
  },
  "path": "/article/4150229/ais-need-for-speed-optical-connectivity-in-focus-at-ofc-2026.html",
  "publishedAt": "2026-03-26T17:13:02.000Z",
  "site": "https://www.networkworld.com",
  "tags": [
    "Artificial Intelligence, Data Center, High-Performance Computing, Networking",
    "opticalization",
    "OFC 2026",
    "Sebastien Naji",
    "the firm’s OFC 2026 observations",
    "Woodside Capital Partners",
    "report on the conference",
    "Bill Gartner",
    "The optical imperative and Nokia’s vision to close the AI gap",
    "Nvidia’s GTC event",
    "making CPO a key component of future large GPU configurations",
    "stated",
    "statement",
    "challenge",
    "Cisco rolled out its Open Transport 3000 Series",
    "Arista has developed a 12.8 Tbps liquid-cooled optics module",
    "Thermo-Electric Generator"
  ],
  "textContent": "The need for high-throughput and energy-efficient optical infrastructure, driven by AI demands, was a recurring theme at this month’s Optical Fiber Communications (OFC) conference in Los Angeles. The so-called opticalization of the network, where fiber replaces copper and optical connectivity becomes imperative, is critical in the AI era, according to OFC 2026 participants and industry watchers.\n\n“The excitement across the ecosystem was palpable, as AI drives significant demand for connectivity solutions across scale-up, scale-out, and scale-across domains,” wrote Sebastien Naji, a research analyst with William Blair, in a report on the firm’s OFC 2026 observations. “As AI has shifted the focus from individual chips to entire computing systems, the ability to … drive high performance has become increasingly tied to how these systems are connected together. While debates rage around what form-factor will gain the most traction (e.g., copper versus optics, co-packaged versus pluggable), the main throughline from our perspective is that demand for reliable and high-performance connectivity continues to accelerate as clusters get larger and racks become denser.”\n\nOptical infrastructure is now a “first-order challenge” in the AI era, concluded Woodside Capital Partners. At this year’s OFC conference, “the center of gravity was unmistakably AI data centers, and the optical backbone required to scale them. Its speed and scale exceed OFC’s traditional telecom orientation by at least an order of magnitude,” the corporate finance advisory firm wrote in its report on the conference.\n\nAI optics will be the main growth contributor to the data center optics market over the next 5 years, according to Bill Gartner, senior vice president and general manager of Cisco’s optical systems and optics group. By 2030, the AI optics total addressable market (TAM) will exceed $20 billion a year with AI applications requiring high-speed optical throughputs supporting 400G, 800G, 1.6T, and 3.2T, Gartner stated.\n\n(Read also: The optical imperative and Nokia’s vision to close the AI gap)\n\n### Co-packaged optics a key focus\n\nCPO always enjoys the spotlight at OFC, and this year it got even more attention, thanks in part to Nvidia. At Nvidia’s GTC event, CEO Jensen Huang reiterated Nvidia’s interest in making CPO a key component of future large GPU configurations built around its Spectrum X platform. Nvidia plans to introduce CPO-based scale-up its Feynman rack, which is due for release in 2028, noted analyst Naji from William Blair.\n\n“While the scale-up domain today is largely serviced by passive copper, data rates and rack densities are necessitating a shift to alternatives,” Naji wrote. “While many of the optical providers like Marvell (following its acquisition of Celestial AI), Broadcom, and Nvidia believe that co-packaged optic is the right solution, others point to pluggable copper, microLEDs, linear optics, and near-packaged optics as potentially better alternatives for scaling up (for their power efficiency, flexibility, and cost). We expect that as connections slowly shift away from copper over the five years, there will be a broad mix of these solutions that succeed, depending on the use-case and scale of a particular system or deployment.”\n\n### The need for speed drives bandwidth density\n\nWoodside Capital Partners predicts that 1.6T is expected to surpass 800G as the primary port speed in new AI backend fabrics by 2027. “The shift to 224G lanes, the push toward 448G, and the unprecedented production ramp rate of 1.6T optical modules highlight how aggressively the roadmap is being pulled forward,” the firm wrote.\n\nOne example of the speed ramp-up is Broadcom’s Taurus BCM83640, a 400G/lane optical DSP built for 1.6T transceiver applications. The Taurus platform of DSPs will enable the next-generation 3.2T optical transceiver modules.\n\n400G/lane technology is the next evolution of 200G/lane architectures, enabling a critical step in scaling bandwidth for high-performance networking and AI infrastructure, according to Broadcom. “1.6T pluggable modules using the Taurus BCM83640 double the bandwidth per optical lane, effectively enabling 102.4T switching capacity in a 1RU system to improve bandwidth density in AI optical interconnects. Further, the adoption of 400G/lane optical interfaces lays the foundation for the eventual deployment of 3.2T module solutions with 400G/lane electrical interfaces for 204.8T switches,” Broadcom stated.\n\n“We expect more than 100 million units of 1.6T and 3.2T optical transceivers to be shipped over the next 5 years with close to half of these using 400G optics,” said Vladimir Kozlov, CEO and founder of analyst firm LightCounting, in a statement related to Broadcom’s Taurus announcement. “High speed optical interconnects are essential for operation of AI clusters. Doubling of the lane rates has been a proven strategy to keep up with the bandwidth growth and it is great to see the first 400G per lane solutions becoming available.”\n\n### Energy efficiency for the AI era\n\nAI’s energy requirements are a challenge for enterprise and hyperscaler organizations. A number of OFC exhibitors looked to address this issue in different ways. For example, Cisco rolled out its Open Transport 3000 Series, a multi-rail open line system that integrates optical components for multiple fiber rails into a single line card, providing improvements in power and density for hyperscalers, neocloud operators and very high-end enterprise AI applications, the vendor stated.\n\nFor its part, Arista has developed a 12.8 Tbps liquid-cooled optics module that it says will help address the power and performance needed for AI data center network development. The eXtra-dense Pluggable Optics (XPO) offers 12.8Tbps of bandwidth using 64 electrical lanes and includes an integrated liquid-cooled cold plate capable of supporting 400W+ module power consumption, according to Arista.\n\nArista also said it has assembled some 45 leading optics module suppliers as part of a multi-source agreement to build and support XPO (it identified three of those vendors: Lightmatter, Eoptolink Technology and TeraHop).\n\nAnother rollout is Coherent’s Thermo-Electric Generator, which harvests thermal energy and converts waste heat into usable electrical power, improving system-level efficiency in next-generation AI datacenters.",
  "title": "AI’s need for speed, optical connectivity in focus at OFC 2026"
}