Eoptolink
A future bet on optical transceivers for data centers
KEY TAKEAWAY: Eoptolink of Chengdu has built a $13 billion market cap company as the technology leader in optical transceivers. It is leading the move to 1.6 terabyte per second ("1.6T") transceivers that are crucial to the next generation of data center infrastructure.
Large data centers use millions of optical transceivers, which cost a few thousand dollars per unit. Switching costs for large data centers are high enough to create a short-term competitive moat. Opening bottlenecks to massively improve performance is worth these costs. Buying cheaper optical receivers to save money is not.
_______________________________
Optical transceivers convert electrical signals generated by servers, switches and routers into optical signals for transmission to other servers in distant locations.
To oversimplify, as I type this post on my computer, each keystroke creates and electrical signal. After these signals are processed by my computer and web browser, and I hit the Return key, the electrical signals are converted to optical signals by the transceiver to be sent to LinkedIn's servers. The signal is then converted back into electrical signals for processing.
Eopticon's Linear-drive Pluggable Optics ("LPO") transceivers are better suited to data centers because they perform better for transmission over short distances in a dense cluster. Most transceivers are optimized for performance between distant locations, e.g. various corporate locations.
Data centers are continually upgrading their bottlenecks where overall performance can be most improved. The technology leader, in this case Eoptolink, can help open these bottlenecks.
The switch to Eoptolink is expensive, but worthwhile because the difference in performance is several times better. Even after Eoptolink's technology is no longer the most cutting edge, customers will be unlikely to feel the urgency to replace it when it does not create a bottleneck.
Eoptolink is a founding member of the Linear Pluggable Optics Multi-Source Agreement ("LPO MSA"), along with NVIDIA, Arista, etc. This enables Eoptolink to influence compatibility standards.
The competitive moat for optical transceivers is less strong than that of Riverstone's clean room gloves since the industry has not yet reached the technological ceiling for transceivers. As long as there is room for exponentially better transceivers, the old can be replaced by the new.
Just as Eoptolink's transceivers offered several times better performance, making the costs of switching worthwhile, it is more likely than not that another such disruption will occur.
Eoptolink's market cap has nearly tripled in the past year despite a decline in net income. Its valuation is almost wholly based on a supposedly desperate need for its technology. It was founded 16 years ago, and has suddenly become a "hot stock" with the unveiling of its new product and the explosion in data centers.
Next up, pet medicines.