Scientific Notation Calculator

Scientific notation — the complete guide

What is scientific notation?

Scientific notation writes a number as a coefficient multiplied by a power of ten, a × 10ⁿ, where the coefficient obeys the normalization rule 1 ≤ |a| < 10 and the exponent n is a whole number. The rule is what makes the form canonical: without it, the same value could be written 53 × 10⁵, 5.3 × 10⁶ or 0.53 × 10⁷, and no two people would agree on which digits count. With it, every non-zero number has exactly one scientific form.

The split does real work. The exponent carries the magnitude — how many orders of ten the value sits at — while the coefficient carries the digits, which is where precision lives. That is why 6.022 × 10²³ tells a reader two separate things in one glance: the number is enormous, and it is known to four significant figures. A plain decimal string with twenty-three zeros communicates neither quickly nor reliably.

The exponent is not obtained by taking a logarithm. It is a counting rule: write the value as a digit string with a power of ten attached, and the normalized exponent is digits.length − 1 + exponent10. For 0.00042 the digit string is 42 with a power of 10⁻⁵, so the exponent is 2 − 1 − 5 = −4, and the answer is 4.2 × 10⁻⁴. This calculator uses exactly that structural rule, which is why it never drifts on boundary cases such as 0.1 or 1000 the way a floating-point logarithm can.

How to convert a number to scientific notation

Move the decimal point until exactly one non-zero digit stands in front of it, then count how far it moved. Moving left — which happens for numbers of magnitude 10 or more — gives a positive exponent. Moving right — for numbers below 1 — gives a negative one. The count is the exponent; the digits you were left with are the coefficient.

Large example: 45,000. The point starts after the final zero. Moving it four places left gives 4.5, so 45,000 = 4.5 × 10⁴. Small example: 0.00032. Moving the point three places right reaches 0.32, which is still below 1; a fourth place gives 3.2, so 0.00032 = 3.2 × 10⁻⁴. A useful check on the sign: a small number must end up with a negative exponent, because a positive one would make it large.

Two habits prevent most errors. First, count the places moved, not the zeros you can see — 0.00032 has four zeros in view but the point moves four places to pass the 3, and those two counts only coincide by accident. Second, drop the leading zeros of the coefficient but keep every significant digit: 0.0001234 is 1.234 × 10⁻⁴, not 1.2 × 10⁻⁴, unless you have deliberately rounded. The Decimal movement panel on this page shows the point hopping one place at a time so you can check your own count against it.

How to convert scientific notation to standard form

Reverse the same movement. A positive exponent shifts the decimal point that many places to the right, padding with zeros as needed: 6.7 × 10⁵ becomes 670,000 (five places right, three zeros added after the 7). A negative exponent shifts it to the left: 9.81 × 10⁻³ becomes 0.00981 (three places left, two zeros added before the 9). E notation behaves identically, because it is the same statement in one-line form — 4.2e-4 is 4.2 × 10⁻⁴ is 0.00042.

The number of zeros you add is not always the exponent. For 6.7 × 10⁵ the exponent is 5 but only three zeros appear, because the digits 6 and 7 already occupy two of the five places. Counting places rather than zeros is the reliable rule in both directions. Paste either form into the single input above and the calculator detects which direction you need — you never have to pick "decimal to scientific" or "scientific to decimal" first.

Scientific notation examples

Every value in this table is produced by the same engine that answers the calculator above, and each row is also a regression vector in the test suite — so a typo in this table would fail the build rather than mislead a reader.

Scientific vs E vs engineering notation

The three forms describe the same value under different constraints on the exponent, and the difference matters more than it looks.

E notation is a typography convention, not a different mathematics: the e stands for exponent and has nothing to do with Euler's number 2.718…. Engineering notation is a genuine constraint change — allowing the coefficient up to 1000 buys an exponent that is always divisible by three, which is precisely the set of powers the SI prefixes name.

Calculating with scientific notation

Addition and subtraction need a shared power of ten before anything can be combined, exactly as fractions need a common denominator. Rewrite the smaller-exponent term over the larger exponent, then add or subtract the coefficients: 4.2 × 10⁵ + 7 × 10⁴ becomes 4.2 × 10⁵ + 0.7 × 10⁵ = (4.2 + 0.7) × 10⁵ = 4.9 × 10⁵. If the resulting coefficient reaches 10 or falls below 1, renormalize — 5.1 × 10⁵ + 11.2 × 10⁶ gives 11.71 × 10⁶, which normalizes to 1.171 × 10⁷.

Multiplication needs no alignment at all: multiply the coefficients and add the exponents. (2 × 10³)(3 × 10⁵) = (2 × 3) × 10³⁺⁵ = 6 × 10⁸. Division is the mirror image: divide the coefficients and subtract the exponents. (6 × 10⁸) ÷ (3 × 10²) = (6 ÷ 3) × 10⁸⁻² = 2 × 10⁶. Renormalization is more common here than people expect — 2 × 10³ ÷ 4 × 10⁵ gives 0.5 × 10⁻², which must become 5 × 10⁻³.

Powers distribute over the product: (2 × 10³)⁴ = 2⁴ × 10³ˣ⁴ = 16 × 10¹² = 1.6 × 10¹³. Roots run the same rule backwards and are easiest when the exponent divides evenly: √(9 × 10⁶) = √9 × 10⁶ᐟ² = 3 × 10³, exactly. When the exponent is odd, rewrite first — √(9 × 10⁷) is better handled as √(90 × 10⁶) = 9.4868… × 10³, which is irrational and is reported here as Approximate rather than dressed up as exact.

Calculate mode runs all of this in exact arbitrary-precision base-10 arithmetic, and shows the alignment, the coefficient step and the exponent step separately, because those are the two places students actually lose marks.

Significant figures and scientific notation

Significant figures count the digits that carry information: all non-zero digits, any zeros between them, and trailing zeros only when a decimal point makes them deliberate. Scientific notation is the cleanest way to state them, because the coefficient is the significant digits — 1.200 × 10³ says four, and 1.2 × 10³ says two, with no ambiguity in either direction.

Plain decimals cannot do that. Written as 1200, the same value could carry two, three or four significant figures, and nothing in the notation settles it. This calculator refuses to decide for you: enter 1200 and it reports 1.2 × 10³ with an explicit note that the trailing zeros are ambiguous, rather than silently claiming four. Choose a significant-figures setting and it will state the precision you asked for — 1200 at four significant figures is 1.200 × 10³.

Rounding here is half-up, away from zero, and applies only to the displayed value; the exact value stays intact underneath. The case worth testing on any calculator is a carry that crosses an exponent boundary: 9.9995 to four significant figures rounds to 10.00, which is not a legal coefficient, so it must renormalize to 1.000 × 10¹. A tool that shows you 10.00 × 10⁰ has failed the normalization rule. Setting a precision never invents digits either — asking for more figures than the value has pads with explicit zeros and says so.

Common powers of ten

These are the literal values behind queries like 1e-4 or 1e12. Every row is generated by the calculator's own engine, so the table and the tool can never disagree.

Large numbers in scientific notation

English number names below use the modern short scale, in which each new name is a thousand times the last — the convention used in the United States, in current British usage and in international finance. On the long scale still used in parts of continental Europe, a billion means 10¹², not 10⁹, so a figure quoted across conventions is worth checking.

Combining a name with a coefficient works exactly as multiplication does: 150 million is 150 × 10⁶ = 1.5 × 10⁸, and 10 billion is 10 × 10⁹ = 1 × 10¹⁰. Both of those initial forms are non-normalized, which is why the calculator accepts them and shows the renormalizing shift rather than rejecting them.

Engineering notation and SI prefixes

Engineering notation exists so the exponent can be spoken. Because it is always a multiple of three, it maps one-to-one onto the SI prefixes maintained by the BIPM — including the four added by the 27th CGPM in 2022 (ronna, quetta, ronto, quecto), which extend the range to 10³⁰ and 10⁻³⁰.

One caution the calculator enforces: a prefix names a power of ten for a unit, not for a bare number. 12.3 × 10⁶ metres is 12.3 megametres, but 12.3 × 10⁶ on its own is not "12.3 M" — there is nothing for the prefix to modify. That is why the result panel reads "SI prefix for the engineering power: mega (M)" rather than pinning a letter onto a dimensionless value.

Zero, negatives and edge cases

Zero is the honest edge case. It has no first non-zero digit, so no exponent can satisfy 1 ≤ |a| < 10 and zero has no unique normalized scientific form at all. This calculator reports the result as simply 0, and where a coefficient–power pair is structurally required it uses 0 × 10⁰ by convention, labelled as a convention rather than presented as a derived fact.

Negative numbers are straightforward: the sign belongs to the coefficient and must never touch the exponent. −0.00042 is −4.2 × 10⁻⁴. A negative exponent and a negative value are independent — −4.2 × 10⁴ is a large negative number, while 4.2 × 10⁻⁴ is a small positive one, and conflating the two is one of the most common slips in the topic.

Very long values are where most calculators quietly fail. Because everything here is stored as an exact digit string rather than a double, the thirty-nine-digit integer 123456789012345678901234567890123456789 converts to 1.23456789012345678901234567890123456789 × 10³⁸ with every digit intact, instead of collapsing to the seventeen digits a floating-point calculator can hold. Limits still exist and are declared: coefficients up to 1,000 digits, exponents to ±1,000,000, and a 10,000-digit exact-arithmetic budget. Past those the calculator returns a structured error instead of an approximation, and a decimal expansion too long to print is suppressed with a note while the scientific form stays exact.

Common mistakes

How this calculator is verified

Three independent layers stand behind every result. First, the engine itself is exact: values are base-10 digit strings with integer powers of ten, and arithmetic runs in arbitrary-precision integers, so no result is ever the residue of a floating-point rounding error. Second, a fixed set of reference vectors — conversions, roundings, arithmetic, inexact cases and every error code — carries expected values written down by hand from the definitions, never generated by the code under test; they run on every build, and the worked-examples table above is asserted against them so an editorial slip cannot ship.

Third, every single calculation is re-checked live by a second implementation that shares no code with the engine. It re-reads the published output strings with its own parser and its own rational arithmetic, then tests the invariants that must hold: that the coefficient really satisfies 1 ≤ |a| < 10, that coefficient × 10ⁿ reconstructs the value, that the E and engineering forms re-parse to the same number, that the engineering exponent is divisible by three, and — in Calculate mode — that the inverse operation returns the first operand. What was checked, and what could not be checked, is listed in the Receipt panel for the result on screen.

What that does not include is human review. There has been no independent third-party mathematical review of this page; correctness is enforced by the deterministic vectors, the exact-arithmetic invariants and the independent round-trip verification described above, and nothing here should be read as an expert endorsement.

Methodology, limits and review status