Floating-Point Calculation. According to Section 3.2, we denote a floating-point number x = [(-1).sup.s'] x s x [2.sup.e-k] (let b=2), which can be represented as a triple (s', s, e), and k is the constant precision.

After completing the

floating-point calculation, the result will be converted into fixed-point numbers for output.

According to Figure 7, it is understandable that the results are similar though the module has some

floating-point calculations. As the module used Q5.10 fixed-point formats so that the result is almost the same as decimal values.

Double-precision performance is more important for supercomputing applications as it carries higher precision for

floating-point calculations than single-precision calculations.

Number of

floating-point calculations that can be carried out per second.

Add complex

floating-point calculations into the mix, and the need for major advancements in processing power becomes obvious.

At a press event in San Francisco, USA, AMD demonstrated a 'Teraflop in a Box' system running a standard version of Microsoft Windows XP Professional that leveraged AMD Opteron dual-core processor technology and two next-generation AMD R600 Stream Processors capable of performing over 1tn

floating-point calculations per second, using a general 'multiply-add' (MADD) calculation.

Altogether, the 768 SGI 400 MHz processors from the two SGI Origin 3800 systems will help to quadruple ARL MSRC's computing power, making the facility one of the world's top 10 most powerful computer sites with a data rate of approximately 2 trillion

floating-point calculations per second.

Speedy

floating-point calculations are essential for applications that require sophisticated design, analysis and three-dimensional visualization capabilities, like mechanical crash simulation and computational fluid dynamics.

Floating-point calculations are critical to technical applications, such as seismic/petroleum exploration, aircraft design analysis, weather forecasting and computational fluid dynamics.

Solving the differential equations for a typical version of this model requires relatively large numbers of

floating-point calculations, which traditionally were performed on supercomputers.

The Spec-mark benchmark measures the ability of a workstation to handle

floating-point calculations. Most of the RISC architectures currently available, including IBM's Power for the RS/6000, Hewlett-Packard's PARISC for the HP Apollo Series 700, and the Clipper from Intergraph Corp.