Cores – Unlocking the Full Potential of Your CPU

Cores is an experimental, CPU-focused programming language inspired by CUDA concepts. It is designed to leverage parallel execution, thread variables, and low-level concurrency directly on your CPU, with a Python-like minimal syntax.

---

Highlights

• Threaded Execution: Automatically run blocks in parallel using times:.

• Shared Variables: Use pub_var, pub_json, pub_array, pub_bool, pub_int, pub_float, pub_str and their uppercase PUB_* counterparts for global state.

• Lightweight Syntax: Minimal constructs like do:, times:, end:, and wait blocks.

• Thread Metadata Access: Inspect threads in runtime with variables such as:

  • THREAD-NAME
  • THREAD-NUMBER
  • THREAD-NATIVE-ID
  • THREAD-IS-ALIVE
  • THREAD-DAEMON
  • THREAD-IDENT
  • PUB_VAR

• Dynamic Variable Assignments: Use ASSIGN to evaluate and store expressions at runtime.

• Line Management: Use RM-LINE to remove lines after execution.

• CPU Power Utilization: Designed to maximize your CPU cores without the complexity of GPU programming.

---

Core Concepts

1. Blocks:

   • do: — Defines a code block that runs immediately.
   • times: (N) — Executes the block N times in separate threads.
   • end: — Concludes a code block.
   • wait — Ensures all threads finish before continuing.

2. Global Variables:

   • Variables starting with pub_ are automatically shared across threads.
   • The uppercase PUB_* versions can be used in expressions and prints.

3. Special Keywords:

   • ASSIGN — Marks a line to execute as a runtime assignment.
   • RM-LINE — Marks a line to be removed from the final code after execution.

4. Thread Info:

   • Access metadata for each thread, including name, ID, status, and daemon flag.

5. Execution Flow:

   • Replace PUB_* placeholders first.
   • Execute ASSIGN lines.
   • Extract and update pub_* variables.
   • Remove lines marked with RM-LINE.

---

Version

• Current Stable: v1.0

---

Tagline

Cores – Unlocking the Full Potential of Your CPU

DOCUMENTATION

Donate