# Co-Packaged Optics Displace Pluggable Transceivers

As AI data centers hit power and reliability limits, switch vendors move optics onto the switch package, shifting value from pluggable-module makers to laser, packaging and switch-silicon suppliers, and each new GPU generation widens the adoption.

- Conviction: 30 / 100 (weakening)
- Horizon: Emerging (watchlist)
- Tracking since: 2026-11-17T00:00:00.000Z
- Last updated: 2026-11-22T06:36:09.478Z
- Canonical: https://polylog.news/ai/trends/co-packaged-optics-displace-pluggables
- Publisher: Polylog
- Affected regions: Global

## Recent score history

- 2026-11-21: 32
- 2026-11-22: 30

## Recent evidence

- [confirms] Lambda says more than 10,000 Blackwell Ultra GPUs now run on Nvidia co-packaged optics switches with lower latency than pluggable optics (2026-11-17): Lambda says more than 10,000 Blackwell Ultra GPUs now run on Nvidia co-packaged optics switches. It reports a latency edge over pluggable optics that grows from 3.8% at rack scale to 10.1% at pod scale, and puts switch power at 3.95 kW against 7.0 kW for a conventional unit. That is production-scale evidence for both the power and performance reasons to adopt.
