Imagine a counter receiving one signal every second. It starts at zero and never stops. It will collect a million signals in less than two weeks, a billion in about 32 years, and a trillion in about 32 thousand years. Why do three similar words describe such different scales?

A million is 1 000 000, a billion is 1 000 000 000, and a trillion is 1 000 000 000 000. Each number in this sequence is a thousand times the previous one. From a million to a trillion takes two multiplications by a thousand: an increase by a factor of a million. Let us see how the written digits show this and check the result using a familiar unit of time.

A thousand thousands makes a new group

If a box held a thousand tokens, a thousand such boxes would hold a million tokens. To make a billion, we would need a thousand sets of a million. For a trillion, a thousand sets of a billion. Each time, we keep the size of the preceding set and take a thousand copies of it.

In decimal notation, a digit’s place determines its value. In 111, the first 1 represents a hundred, the second a ten, and the third a single unit. Moving one place to the left increases the place value tenfold. Three places means 10 × 10 × 10: a thousandfold increase.

That is why it helps to split large whole numbers into groups of three digits, starting from the right. Consider 2 345 678: two million, 345 thousand, and 678 units. The spaces help us read the place values; they do not tell us to multiply the groups. Even an empty group needs its zeros to preserve the positions: 2 000 007 contains two million and seven units, with no thousands.

Adding three zeros on the right of a positive whole number multiplies it by a thousand. 1 000 000 becomes 1 000 000 000. Adding a thousand is different: a million plus a thousand is only 1 001 000.

Use a second as the unit of comparison

The boxes explain grouping, but thousands of boxes are hard to picture too. Return to one counter with an unchanging rate: one signal every second, without breaks. The count starts at zero, with the first signal after one second. A million signals then require exactly a million seconds.

We take a day as a unit of 86 400 seconds. To express longer intervals in years, we use a Julian year: exactly 365.25 days, or 31 557 600 seconds. This defined interval is used in astronomical calculations; it does not mean that every calendar year has the same length. JPL: the day and the Julian year.

Under these assumptions, a million signals take about 11.57 days, a billion 31.69 years, and a trillion 31 688 years. Changing to more convenient units leaves the ratio unchanged: a billion seconds is still a thousand times as long as a million seconds.

This is an imaginary instrument, not a promise that a person can count aloud to a billion in 32 years. Long number names take different times to say, and a person needs sleep and breaks. The exact arithmetic applies to the specified signal rate. The everyday comparison conveys the scale only under that condition.

Branch · CalculationOptional · you can skip this

Calculating the time in days and years

Divide the time in seconds by the number of seconds in the chosen unit:

(1 000 000 s)/(86 400 s/day); ≈ 11.57 days,

(1 000 000 000 s)/(31 557 600 s/year); ≈ 31.69 years.

The same rule gives 31 688.0878… years for a trillion. We divide before rounding, so the table shows 31 688 years. Rounding a billion seconds to 31.69 years first and then multiplying by a thousand would give 31 690: the error from early rounding also grows a thousandfold. Every year here is a Julian year of 365.25 days; we are not calculating a finishing date on a calendar.

Why an accurate chart can hide a million

Before switching the view below, try a prediction: if a full bar represents a trillion, what fraction will a billion fill? And a million?

On a linear scale, lengths have the same ratios as the numbers. A billion therefore fills one thousandth of the full bar, and a million one millionth. If the full bar were a metre long, these would be one millimetre and one thousandth of a millimetre. The smallest bar may be impossible to distinguish on a screen.

The grouped view serves another purpose: it shows successive multiplications by a thousand. Equal steps represent equal factors, rather than equal added amounts. Switch scales and compare the visible bars with the numbers in the table.

The same numbers, two different views
Choose a scale and compare: what does each view hide?

Grouped (logarithmic) scale: each step is ×1000

  1. Million1061 000 000
  2. Billion1091 000 000 000
  3. Trillion10121 000 000 000 000

Equal steps show equal multiplication. Card widths do not represent counts; zero is not shown on this scale.

Linear scale: full length = 1 trillion

Million0.0001%
Billion0.1%
Trillion100%

Length is proportional to the number. A billion fills 0.1%, a million 0.0001%: its bar may be invisible on this screen. It has not been enlarged.

Exact counts and hypothetical time at one signal per second
Signal countTime ≈Share of a trillion
Million1 000 00011.57 days1/1 000 000
Billion1 000 000 00031.69 years1/1000
Trillion1 000 000 000 00031688 years1

Start at 0, first signal after 1 s, no breaks. A day = 86 400 s; a year here is a Julian year of 365.25 days. This is a calculation, not a measurement of human counting speed.

Branch · A closer lookOptional · you can skip this

What the logarithmic view means

The difference between a million and a billion is 999 million. Between a billion and a trillion it is 999 billion. The ratio is a thousand in both cases, although the differences are very unequal.

A logarithmic scale gives equal distances to equal multiplication factors. Our ladder uses multiplication by a thousand as a step: million, billion, trillion. It lets us see all three levels without a bar a million times as long as another. However, those distances cannot be used to estimate differences. Zero does not appear on this scale: multiplying a positive number by a thousand will never turn it into zero.

The scale does not change the numbers. It changes which relationship is convenient to see: differences on a linear scale, or factors on a logarithmic one. The scale label is part of a chart’s meaning.

A power of ten shortens a long string of zeros

To avoid rewriting long strings of zeros, we use powers of ten. The notation (10)^(3) means 10 × 10 × 10 = 1000. The exponent above the 10 tells us how many tens are multiplied together. For a positive whole-number exponent, it also gives the number of zeros after the 1: (10)^(6) is a million, (10)^(9) a billion, and (10)^(12) a trillion.

In scientific notation for a positive number, the power is preceded by a number from 1 up to, but not including, 10. For example, 2.5 × (10)^(9) means 2.5 × 1 000 000 000, or 2 500 000 000: two and a half billion. The power sets the scale; the preceding number tells us how many of those scale units we take. Understanding this does not require solving equations first.

The same decimal factors appear in SI unit prefixes: mega- (M) means (10)^(6), giga- (G) means (10)^(9), and tera- (T) means (10)^(12). One megasecond, 1 Ms, is a million seconds; 1 Gs is a billion seconds; 1 Ts is a trillion seconds. A prefix without a unit does not yet tell us what was measured. A million seconds and a million metres have the same number but different physical meanings. BIPM: SI prefixes.

These prefixes represent powers of ten. They must not be redefined using multiplication by 1024; binary factors have separate prefixes. NIST: decimal prefix rules.

Similar words in different languages can mean different numbers

In current English, billion is 1 000 000 000, called мільярд in Ukrainian, while trillion is 1 000 000 000 000. These English meanings also appear in the US NIST prefix table. NIST: number names and prefixes.

In standard Spanish usage, billón means a million million: (10)^(12). For the English billion, the RAE recommends mil millones, literally “a thousand million”, or millardo. Regional variations also exist, making it especially useful to check the digits in a translation. Transferring a similar-looking word without checking can change a number by a factor of a thousand. RAE: billón and billion.

A reliable way to read large numbers in another language is to find the full numerical form or power of ten, check the unit, and only then choose a name. The counts in (10)^(9) seconds and (10)^(12) seconds have no such linguistic ambiguity, even with an unfamiliar word alongside.

How many millions fit in a trillion?

Try answering without counting each zero: does a trillion contain a thousand million or a million million?

A billion contains a thousand million. A trillion contains a thousand such billions, so 1000 × 1000 = a million million. This also explains why, at a constant rate, our counter takes a million times as long to reach a trillion as to reach a million.

To make a large number understandable, split it into groups of thousands, identify the factor, and name the unit. For a chart, check the scale as well. This also helps us read the numbers in a description of the Solar System, instead of treating every distant place as simply “enormously far away”. Next in the course, we will explore units that make those distances easier to measure.