gigaflops


Also found in: Dictionary, Wikipedia.

gigaflops

[′gig·ə‚fläps]
(computer science)
A unit of computer speed, equal to 109 flops.

gigaflops

(unit)
(GFLOPS) One thousand million (10^9) floating point operations per second.

One of them is strictly "one gigaflops" in the same way that one mile per hour isn't 1 MP.

See prefix.

gigaFLOPS

(GIGAFLoating point OPerations per Second) One billion floating point operations per second. See FLOPS.
Mentioned in ?
References in periodicals archive ?
The Myriad 2 VPU housed inside the Movidius Neural Compute Stick provides powerful, yet efficient performance - more than 100 gigaflops of performance within a 1W power envelope - to run real-time deep neural networks directly from the device.
Intel also announced the Phi SE10X and SE10P processors with 61 cores and 1,073 gigaflops of computing power.
3 GHz PowerPC G5 processors with over 35 gigaflops of processing power per system.
It is based on the Linpack benchmark, which measures a computer's floating-point rate of executing linear equations and is expressed in Gigaflops (Gflop/s) or billions of floating point operations per second.
According to Cupertino, California-based Apple, Xserve G5 is about 60% more powerful than its PowerPC G4-based Xserve predecessor, and delivers 30 gigaflops of processing power per 1U rack-mount server.
Intel is keen on lowering the cost of a gigaflops of computing power, and increasing the performance and lowering the price of the chip because it has smaller cache should do this.
A laboratory for the earth and space sciences, computing managers there threw down a gauntlet to develop a $50,000 gigaFLOPS workstation for processing satellite data sets.
They took hundreds of PCs and linked them into a single machine that achieves and sustains scientific computing performance of more than 10 Gigaflops - over ten billion operations per second.
2 gigaflops of processing power to performing geometric transformations and accelerating image processing.
Designed from the ground up for HPC, computational science and supercomputing, NVIDIA's Tesla M2090 GPUs provide 512 CUDA cores with 665 gigaflops of peak performance to accelerate even the most demanding computationally intensive workloads.
Tesla M2090, which has 512 CUDA parallel processing cores, delivers 665 gigaflops of peak double-precision performance, enabling application acceleration by up to 10 times, compared to using a processing unit (CPU) alone, Nvidia said.
As the world's highest-performance general-purpose CPU, the processor offers both performance and energy efficiency, achieving a computational speed of 128 gigaflops per CPU.