Summary

  • Ciena's WaveLogic family page states 1.6 Tb/s only inside metro ROADM deployments, 1.2 Tb/s across 1,000 km, and 800G across “virtually all network links—even across transpacific distances” with no distance attached to that last clause at all.
  • The Form 10-Q for the quarter ended 1 August 2026 reports total revenue of $1,671,129 thousand against research and development expense of $236,673 thousand — 14.2% of revenue, against 17.4% in the quarter ended 2 August 2025.

The number is the ticket; the condition is the price

Product pages are usually read vertically: fastest first. Read horizontally, Ciena's WaveLogic family page prints a rate card with the qualifiers still attached:

“WaveLogic™ 6 Extreme offers programmable 200 Gb/s–1.6 Tb/s line rates, supporting massive 1.6 Tb/s single-carrier wavelengths for metro ROADM deployments and 1.2 Tb/s across 1,000 km distances while extending 800G connectivity across virtually all network links—even across transpacific distances. WaveLogic 6 Nano enables fit-for-purpose 800G pluggables to meet diverse customer and application requirements, from 10 km to 1,000 km, as well as a 1.6T Coherent-Lite pluggable for data center applications.”

Three things in that paragraph are worth pulling apart.

The 1.6 Tb/s figure arrives inside an application: metro ROADM deployments. It is a ceiling with an address, not a universal rate. The 1.2 Tb/s figure arrives with the hardest number on the page — 1,000 km. And the 800G clause, the one reaching toward the longest links, arrives with a superlative instead of a distance. “Virtually all network links—even across transpacific distances” tells a reader that the reach is large. It does not tell a reader how large.

The gap between the first two rates is the trade-off, priced in my own division and nothing else: 1.2 divided by 1.6 is 0.75. Reaching 1,000 km costs a quarter of the top single-carrier rate, on the vendor's own pairing of the two numbers. The longest-link clause, read the same way, sits at half the top rate — 800 Gb/s against 1,600 Gb/s — with no distance to measure it against.

The 2023 sentence was written in the future tense

The dated document is a launch release, and on this point it is more careful than the product page: it is dated 21 February 2023, and it uses a roadmap verb.

“Ciena's WaveLogic 6 will support up to 1.6Tb/s single-carrier wavelengths for metro ROADM deployments, 800Gb/s over the longest links”

“Will support” is not a record of a shipped link or a customer deployment. The pairing, though, was already fixed three years earlier: the top rate belongs to the metro, and the longest links get the lower rate. What changed between that announcement and today's product page is that 800Gb/s “over the longest links” acquired the phrase “even across transpacific distances”, and the page added a third data point — 1.2 Tb/s at exactly 1,000 km — which the release did not carry. The 1,000 km in the release attaches to energy-efficient 800G pluggables, a different object with a different power envelope.

That distinction is not pedantry. It is the difference between a claim a network planner can design against and a claim only a marketing page can defend: one rate, one distance, one product.

One label, two bodies

The same page reuses “1.6T” for two different objects. The first is a single-carrier coherent wavelength for metro ROADM. The second is a Coherent-Lite pluggable aimed at data centre applications — a different form factor serving a different job. Both are labelled 1.6T. Only one of them is the wavelength a long-haul route is designed around.

Anyone comparing Ciena's terabit claims across documents, or across product families, walks into this: the label is shared, the object is not. The figure stays the same; the distance, the power and the application do not.

What the filed quarter actually measures

Ciena's Form 10-Q for the quarterly period ended 1 August 2026, filed 3 September 2026, is a different kind of document from either page, and it does not price the curve at all. It prices the effort.

Total revenue was $1,671,129 thousand. Research and development expense was $236,673 thousand. A year earlier, in the quarter ended 2 August 2025, the same two lines read $1,219,385 thousand and $211,898 thousand. The period is worth stating exactly, because I got it wrong first: the quarter ended 1 August, not 31 July, and the filing says so on its cover.

Divided out — my arithmetic, not the filing's: R&D was 14.2% of revenue in the quarter just filed, against 17.4% a year earlier. Revenue grew 37.0%; R&D grew 11.7%. The spend line lagged the top line by roughly 25 points.

That is neither scandal nor strategy. It is a consolidated ratio, and the filing shows nothing finer. What can be said is narrow and worth saying precisely: in the one quarter examined, the effort behind the portfolio grew at about a third of the rate of the revenue the portfolio produced. What cannot be said is where inside the portfolio that effort went, because no product-level or segment-level allocation was obtained.

What is not on the page

The family page carries no publication date. I read it on 13 November 2026 and cannot date the claims it makes; treat it as current at reading, not as a dated commitment.

No datasheet, no spectral-efficiency figure in bit/s/Hz, and no watts-per-Gb/s figure for either module was obtained across three attempts at that material. That is a limit on what I obtained, set beside the claim it qualifies — not a finding that Ciena does not publish such figures. The same restraint applies to the conditions a planner would actually need: no route survey, no amplifier-spacing specification, no fibre type, no channel-load assumption and no customer acceptance record was obtained, so nothing here establishes that any buyer reaches any of these rates on any route.

A buyer reading a rate card should therefore ask for the pairing rather than the peak: which distance, which fibre, how many channels lit, what the line card dissipates. The 1.6 Tb/s figure exists. On the vendor's own page, it does not travel alone.