JavaScript Navigator hardwareConcurrency Property

What You’ll Learn

navigator.hardwareConcurrency returns how many logical processors the browser reports. Learn physical vs logical cores, Worker pool sizing, safe caps, five examples, and try-it labs.

Introduction

Heavy JavaScript on the main thread freezes the UI. Web Workers move work off the main thread — and the next question is: how many workers should I start?

navigator.hardwareConcurrency answers with a number: the logical processors available to run threads. A four-core CPU might report 8 if each core runs two threads. That number is a practical hint for parallel Worker pools.

Understanding the hardwareConcurrency Property

Think of it as “how many threads can usefully run at once before they start fighting for the same cores?”

• Logical processors — often higher than physical cores (hyper-threading / SMT).
• Value — a number from 1 up to the reported logical count.
• Browser discretion — may under-report for a more realistic Worker budget.
• Main use — size a Worker pool (MDN’s classic example).
• Baseline — Widely available across modern browsers (MDN, since about March 2022).

📝 Syntax

General form of the property:

navigator.hardwareConcurrency

Value

• A number between 1 and the number of logical processors the user agent reports.

Common patterns

// Read the count:
console.log(navigator.hardwareConcurrency);

// Cap workers so a 32-core machine does not spawn 32 Workers:
const workers = Math.min(navigator.hardwareConcurrency || 4, 8);

// MDN-style pool sketch:
const workerList = [];
for (let i = 0; i < navigator.hardwareConcurrency; i++) {
  workerList.push({ id: i, inUse: false });
}

Examples Gallery

Examples follow MDN navigator.hardwareConcurrency patterns for reading cores and sizing Worker pools. Prefer Try It Yourself to see your machine’s reported count.

console.log(navigator.hardwareConcurrency);
console.log("type: " + typeof navigator.hardwareConcurrency);

8
type: number
(your number will vary — 4, 8, 12, 16…)

const cores = navigator.hardwareConcurrency;
console.log(
  "About " + cores + " logical processor" +
  (cores === 1 ? "" : "s") + " available"
);

About 8 logical processors available

const MAX = 8;
const poolSize = Math.min(
  Math.max(1, navigator.hardwareConcurrency || 4),
  MAX
);
console.log("pool size: " + poolSize);

pool size: 8
(on a 4-core report: pool size: 4)

const workerList = [];

for (let i = 0; i < navigator.hardwareConcurrency; i++) {
  workerList.push({
    id: i,
    // In a real app: worker: new Worker("cpu-worker.js"),
    inUse: false
  });
}

console.log("planned workers: " + workerList.length);
console.log(JSON.stringify(workerList.slice(0, 3)));

planned workers: 8
[{"id":0,"inUse":false},{"id":1,"inUse":false},{"id":2,"inUse":false}]

function concurrencyTier() {
  const n = navigator.hardwareConcurrency || 4;
  let tier = "medium";
  if (n <= 2) tier = "low";
  else if (n >= 8) tier = "high";
  return {
    cores: n,
    tier: tier,
    workers: tier === "low" ? 1 : tier === "medium" ? 2 : Math.min(n, 6)
  };
}

console.log(JSON.stringify(concurrencyTier()));

{"cores":8,"tier":"high","workers":6}

🚀 Common Use Cases

• Worker pools — size parallel workers for image / data processing (MDN pattern).
• Encoding / compression — split chunks across a capped worker count.
• Game / simulation loops — offload physics or AI to N workers.
• Adaptive quality — fewer parallel jobs on 2-core devices.
• Diagnostics — show reported concurrency in a “About this device” panel.

📝 Notes

• Baseline Widely available (MDN, since about March 2022).
• Not Deprecated, Experimental, or Non-standard — no status banner required.
• Reports logical processors — not always physical core count.
• Browsers may under-report; do not treat the value as absolute hardware truth.
• Related: hid, deviceMemory, JavaScript hub.

💡 Best Practices