40G photodetector: The other end of the link

As you may have read in Ansheng’s earlier post, our photonics labs recently disclosed a 40 gigabit per second laser modulator. Optical modulators, which encode high speed data onto an optical beam, are something we have been working on in silicon for a while. I actually worked in the lab when we first hit 1Gbps and then 10Gbps, and seeing them finally hit 40G is truly an accomplishment. But in all this time we haven’t said much about the other end of the fiber — how you detect optical data at 40 Gbps and convert the information back into electrical signals which a computer can read.

 

 

 

The challenge is simple when you think about it. The whole reason we can use silicon to make integrated data pipes (called waveguides) is that silicon is clear to the wavelengths of light used for optical communication. For them, it’s as clear as glass. And that fact means that silicon can’t detect the light. For detection, the material needs to absorb light, as silicon does for the visible light we see with the naked eye, making it opaque.

 

40G-photodetector-chip.jpg

 

I’m sitting in the Westin hotel at our IDF “Day0” press briefing, watching my old boss Mario announce an achievement in Silicon Germanium photodetectors (pictured here). Germanium is a material that can efficiently detect light at comms wavelengths (roughly 1.3-1.6 microns). In order to convert our waveguides into detectors, the researchers have added a layer of germanium to the device as shown here.

 

SigeSlide.jpg

 

Germanium is a CMOS compatible material, often used in semiconductor processing. However, building a device like this that performs well is a challenge. Photodetectors need to be fast, efficient, and low-noise, and the one Mario is announcing now arguably has the best overall performance of any waveguide detector in silicon. The challenge, it turns out, is in the difference in the crystal lattices of Germanium and Silicon. I got an update on this from researcher Yin Tao a few days ago which you can see here.

 

 

This is a great achievement, since between the 40G detector, the 40G modulator, and the hybrid silicon laser we announced earlier, we have all the basic building blocks for a high speed link. To learn more flip though Mario’s slides.

 

3 Responses to 40G photodetector: The other end of the link

  1. Erik Sorensen says:

    Hello Sean,
    I am a 4th year Engineering Physics undergraduate student at McMaster University in Canada. I am specializing in optics and I have taken many courses in silicon VLSI (including a chip manufacturing lab). I feel as though I am very well suited towards a job in the optics research lab at Intel. Is there someone I can talk to to make this happen?

  2. Joe Warner says:

    Could someone contact me about your optical systems. I am interested in them for space communication.
    Joe Warner