4 ms·

That seems insane.

• dekhn 3y ago
  Why? In high throughput optics you are basically scanning one full frame of a 4K (3840x2160) per second (or higher!). If you store one bit per pixel, with 16 bit data, that's 10-20MB/sec. You would likely move the field of view to obtain 10FPS, so 200MB/sec per reader. However, it would be trivial to parallelize this (with more objectives, mirrors, higher resolution cameras or whole devices) and so it's not impossible to imagine an optimized, production version of this generating terabytes/second of raw data.
  • aziaziazi 3y ago
    Because positioning a 4K camera at nanometer scale is hard and moving it very precisely 10 times/sec is even harder. Reading each pixel at that scale and speed without substantial error correction seems crazy to me. You’ll probably want at least 3x3 pixels/bit and a higher resolution camera that moves less.