Skip to content

FFT History Timeline

Open in full window

This interactive timeline visualization shows the 50 most important advances in Fast Fourier Transform (FFT) technology, from theoretical foundations to modern hardware implementations.

Timeline Coverage

The timeline spans from 1805 to 2024 and includes:

1. FFT Algorithms (Orange)

  • Gauss's original discovery (1805)
  • Cooley-Tukey algorithm (1965)
  • Radix-4, Split-radix, Winograd, and other algorithmic innovations
  • Modern sparse FFT algorithms

2. Software Libraries (Light Blue)

  • FFTW (Fastest Fourier Transform in the West)
  • CMSIS-DSP (ARM's standard library)
  • Intel MKL and IPP
  • NVIDIA cuFFT for GPU acceleration
  • KISS FFT

3. DSP Chips (Green)

  • Texas Instruments TMS320 family evolution
  • Motorola DSP56000 series
  • Analog Devices SHARC and Blackfin
  • Dedicated DSP processors that revolutionized signal processing

4. Microcontrollers (Pink)

  • ARM Cortex-M3, M4, M7 with integrated DSP
  • STM32 series
  • Raspberry Pi Pico and Pico 2
  • Modern MCUs with DSP extensions
  • ESP32 and other IoT processors

5. Hardware Features (Purple)

  • SIMD instructions (ARM Neon, Intel AVX)
  • GPU acceleration (CUDA)
  • ARM Helium vector extensions
  • RISC-V vector extensions

Interactive Features

  • Zoom Controls: Zoom in/out to see different time scales
  • Era Navigation: Jump to Early History, DSP Era, or Modern Era
  • Tooltips: Hover over events for detailed descriptions
  • Grouped Display: Events are organized by category with color coding

View the Timeline

Educational Value

This timeline is particularly useful for:

  • Understanding the historical context of FFT development
  • Seeing the relationship between algorithmic advances and hardware evolution
  • Appreciating how DSP moved from dedicated chips to integrated MCU features
  • Following the progression from floating-point DSPs to modern ARM Cortex-M processors
  • Understanding why the Raspberry Pi Pico 2 (RP2350) with Cortex-M33 is significant

Relevance to This Course

This course focuses on benchmarking FFT implementations on modern microcontrollers, particularly:

  • ARM Cortex-M4 and M33 processors (like those in Raspberry Pi Pico 2)
  • Understanding trade-offs between integer and floating-point FFT
  • Leveraging DSP instructions for performance
  • Using standard libraries (CMSIS-DSP) vs. custom implementations

The timeline shows how we arrived at today's powerful yet affordable microcontrollers that can perform real-time FFT operations that once required expensive dedicated DSP chips.

Data Format

The timeline data is stored in fft-timeline.json using the vis-timeline data format:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
{
  "items": [
    {
      "id": 1,
      "content": "Event name",
      "start": "YYYY-MM-DD",
      "type": "point",
      "group": "category",
      "title": "Detailed description"
    }
  ],
  "groups": [...]
}

References

  • Cooley, J. W., & Tukey, J. W. (1965). "An algorithm for the machine calculation of complex Fourier series"
  • Frigo, M., & Johnson, S. G. (1997). "FFTW: An adaptive software architecture for the FFT"
  • ARM CMSIS-DSP Software Library Documentation
  • Texas Instruments TMS320 DSP History