What are exaFLOPS?
The computing power of modern computers is measured in FLOPS, which stands for floating point operations per second. Very powerful systems can reach the level of exaFLOPS, which is a quintillion (1018) operations per second. This enormous computing capacity is crucial for complex calculations.
Manage any workload with flexible GPU computing power, and only pay for the resources you use.
What are FLOPS and what are they used for?
FLOPS stands for floating point operations per second. Floating point arithmetic in computers is essential for scientific calculations, simulations, artificial intelligence and other computing-heavy applications. Unlike simple operations with integers, floating point numbers make it possible to calculate and represent decimal numbers very precisely.
A computer’s performance in FLOPS is measured using specialised benchmarks that test how many floating point operations can be performed per second. That often involves programs like LINPACK and HPCG, which perform complex mathematical calculations and measure computing speed under real-world conditions. Actual performance can vary based on hardware, software optimisation and the type of calculation.
How many FLOPS are in an exaFLOPS?
One exaFLOP corresponds to exactly 1018 FLOPS or one quintillion floating point operations per second. For comparison, one petaFLOPS (1015 FLOPS) is a thousand times smaller than an exaFLOPS, and one teraFLOPS (1012 FLOPS) is a million times smaller.
What are some other FLOPS units and how are they converted to exaFLOPS?
There are many FLOPS units, which differ in how many operations per second they refer to.
Unit | FLOPS value | Conversion into exaFLOPS |
---|---|---|
KiloFLOP | 103 FLOPS (1,000) | 10-15 ExaFLOPS |
MegaFLOP | 106 FLOPS (1 million) | 10-12 ExaFLOPS |
GigaFLOP | 109 FLOPS (1 billion) | 10-9 ExaFLOPS |
TeraFLOP | 1012 FLOPS (1 trillion) | 10-6 ExaFLOPS |
PetaFLOP | 1015 FLOPS (1 quadrillion) | 10-3 ExaFLOPS |
ExaFLOP | 1018 FLOPS (1 quintillion) | 1 ExaFLOP |
These units make it easier to compare the computing power of different systems to each other.
How many FLOPS do modern computers and GPUs reach?
Modern high-performance computers now reach the levels of exaFLOPS. The first exaFLOPS supercomputer was the Tianhe-3 in the Chinese National Supercomputer Center in Guangzhou, which delivers up to 1.7 exaFLOPS. Other supercomputers like the Frontier in the US now also operate at exaFLOPS levels.
High-end graphics cards like those used for gaming and machine learning also have high FLOPS values, even if they don’t come anywhere near exaFLOPS:
- NVIDIA RTX 4090: Around 82 teraFLOPS (0.000082 exaFLOPS) for FP32 calculations; with overclocking it can exceed 1,000 teraFLOPS
- NVIDIA H100 SXM: Around 989 teraFLOPS (0.000989 exaFLOPS) for FP32 Tensor Core calculations
- NVIDIA A30: Around 10.3 teraFLOPS (0.000010 exaFLOPS) for FP32 calculations
- AMD Radeon RX 7900 XTX: Around 61 teraFLOPS (0.000061 exaFLOPS) for FP32 calculations
While GPUs are primarily designed for parallel calculations in AI and graphics processing, supercomputers are used for very complex simulations in science and industry.
- Enterprise hardware
- Power and flexibility
- Latest security technology