Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Beyond Bits: An Introduction to Computation over the Reals

Tillmann Miltzow

Abstract

We introduce a lightweight and accessible approach to computation over the real numbers, with the aim of clarifying both the underlying concepts and their relevance in modern research. The material is intended for a broad audience, including instructors who wish to incorporate real computation into algorithms courses, their students, and PhD students encountering the subject for the first time. Rather than striving for completeness, we focus on a carefully selected set of results that can be presented and proved in a classroom setting. This allows us to highlight core techniques and recurring ideas while maintaining an approachable exposition. In some places, the presentation is intentionally informal, prioritizing intuition and practical understanding over full technical precision. We position our exposition relative to existing literature, including Matousek's lecture notes on ER-completeness and the recent compendium of ER-complete problems by Schaefer, Cardinal, and Miltzow. While these works provide deep and comprehensive perspectives, our goal is to offer an accessible entry point with proofs and examples suitable for teaching. Our approach follows modern formulations of real computation that emphasize binary input, real-valued witnesses, and restricted use of constants, aligning more closely with contemporary complexity theory, while acknowledging the foundational contributions of the Blum--Shub--Smale model.

Beyond Bits: An Introduction to Computation over the Reals

Abstract

We introduce a lightweight and accessible approach to computation over the real numbers, with the aim of clarifying both the underlying concepts and their relevance in modern research. The material is intended for a broad audience, including instructors who wish to incorporate real computation into algorithms courses, their students, and PhD students encountering the subject for the first time. Rather than striving for completeness, we focus on a carefully selected set of results that can be presented and proved in a classroom setting. This allows us to highlight core techniques and recurring ideas while maintaining an approachable exposition. In some places, the presentation is intentionally informal, prioritizing intuition and practical understanding over full technical precision. We position our exposition relative to existing literature, including Matousek's lecture notes on ER-completeness and the recent compendium of ER-complete problems by Schaefer, Cardinal, and Miltzow. While these works provide deep and comprehensive perspectives, our goal is to offer an accessible entry point with proofs and examples suitable for teaching. Our approach follows modern formulations of real computation that emphasize binary input, real-valued witnesses, and restricted use of constants, aligning more closely with contemporary complexity theory, while acknowledging the foundational contributions of the Blum--Shub--Smale model.

Paper Structure

This paper contains 61 sections, 23 theorems, 50 equations, 24 figures, 1 table, 3 algorithms.

Table of Contents

  1. Models of Computation
  2. Motivation
  3. Word RAM
  4. Mathematical Analysis.
  5. Independent of Physical Computers.
  6. Limited Communication Value.
  7. Implicitness Assumption.
  8. Turing Machines
  9. Mathematically Simplistic and Formal.
  10. Equivalent to word RAM.
  11. Algorithm Design on Turing Machines.
  12. Real Models of Computation
  13. Hide Numerical Issues.
  14. Unrealistic Assumption.
  15. Naturalness.
  16. ...and 46 more sections

Key Result

Theorem 1

We can solve the Factoring Problem in polynomial time on a real RAM that has a rounding operation.

Figures (24)

  • Figure 1: An algorithm on the word RAM is a finite list of instructions. Input, output, and all working memory take place in an infinite array of registers.
  • Figure 2: Turing Machines originated from a simple thought experiment about a simple worker.
  • Figure 3: The real RAM has two infinite arrays of registers. One that can operate on words and one that is allowed to operate on real numbers.
  • Figure 4: Here, you can see the region in the plane that satisfies both of those constraints.
  • Figure 5: The Cook--Levin tableau idea: we encode the contents of memory cells (and the current instruction) at each time step and constrain consecutive time steps to follow the program.
  • ...and 19 more figures

Theorems & Definitions (38)

  • Theorem 1
  • proof
  • Lemma 1
  • Lemma 2
  • proof
  • Lemma 3
  • proof
  • Lemma 4
  • proof
  • Lemma 5
  • ...and 28 more