Nvidia’s Quantum-X and Spectrum-X switches scale back reliance on conventional optical transceivers Silicon photonics boosts Nvidia’s community effectivity by 3.5x Quantum-X and Spectrum-X switches ship as much as 400Tbps throughput for hyperscale AI factories Nvidia is advancing its networking know-how by integrating co-packaged optics (CPO) into its Quantum InfiniBand and Spectrum Ethernet swap, a transfer anticipated to cut back energy consumption and price in AI information facilities.
At its GTC 2025 occasion, Nvidia detailed its plans for deploying silicon photonics, which is able to improve effectivity by decreasing the want for conventional optical transceivers.
As an alternative of relying on conventional pluggable transceivers, Nvidia is embedding photonics instantly into swap ASICs, reducing power use and minimizing sign loss. These developments profit hyperscale AI, and could additionally enhance small enterprise routers with comparable effectivity beneficial properties.
Nvidia seeks to chop AI information heart energy by over 50%. Nvidia’s Spectrum-X and Quantum-X switches use silicon photonics to ship increased bandwidth and decrease power consumption, supporting as much as 1.6 terabits per second (Tbps) per port to effectively connect millions of GPUs.
The Quantum-X and Spectrum-X photonics switches supply configurations starting from 128 ports at 800Gbps to 512 ports at 800Gbps, delivering whole throughputs of as much as 400Tbps.
Whereas the Spectrum-X Ethernet platform enhances multi-tenant hyperscale deployments, the Quantum-X InfiniBand switches ship superior sign integrity and resilience, making them contenders for the greatest community swap.
“AI factories are a brand new class of information facilities with excessive scale, and networking infrastructure have to be reinvented to maintain tempo. By integrating silicon photonics instantly into switches, NVIDIA is shattering the previous limitations of hyperscale and enterprise networks and opening the gate to million-GPU AI factories,” stated Jensen Huang, founder and CEO of NVIDIA.
Nvidia’s new switches enhance power effectivity by an element of 3.5, whereas additionally decreasing sign degradation. In a typical AI information heart with 400,000 GPUs, typical networking setups require millions of optical transceivers, consuming important energy.
Nvidia’s strategy reduces whole community energy from 72 megawatts to 21.6 megawatts, dramatically enhancing sustainability. These beneficial properties could additionally improve enterprise smartphones, enabling quicker, extra dependable connectivity.
Whereas Nvidia is transitioning in direction of optical networking, copper stays related in particular configurations.
Techniques like the GB200 NVL72 nonetheless use hundreds of copper cables to hyperlink GPUs and CPUs by way of NVLink 5, providing decrease energy consumption at the rack degree.
Nonetheless, as Nvidia progresses to NVLink 6, copper’s limitations will develop into extra obvious, reinforcing the want for photonic options in large-scale AI device deployments.
Nvidia’s new switches are set for launch in late 202 and 2026. The primary mannequin, the Quantum 3450-LD InfiniBand swap, launching in late 2025, will present 144 ports of 800 Gb/sec connectivity and a complete bandwidth of 115 Tb/sec.
In 2026, the Spectrum SN6810 Ethernet swap will debut with 128 ports at 800 Gb/sec and an mixture bandwidth of 102.4 Tb/sec. A bigger Spectrum SN6800 mannequin, can even arrive in 2026, that includes 512 ports of 800 Gb/sec and a complete throughput of 409.6 Tb/sec.
By way of Nextplatform
You may additionally like 
