When measuring and testing computer applications, scientists and engineers collect huge amounts of data every second of the day. For instance, the world’s largest particle holder collider known as Large Hadron Collider generates approximately 40 terabytes of data per second. The jet engine of a Boeing creates approximately ten terabytes of data every thirty minutes. When a Jumbo jet takes a trip across the Atlantic ocean, the four engines on the jet can produce approximately 640 terabytes of data. If you multiply that kind of data with an average of 2,500 daily flights, the amount of data produced per day is staggering; this is what is called Big Data.
It is a difficult task to draw conclusions and get actionable data from the large sums of data, and Big Data encompasses this issue. Big data has brought about new ways of processing data; we have deep data analysis tools, data integration tools, search tools, reporting tools and maintenance tools that help in processes big data to derive value from it.
The International Data Corporation (IDC) performed a study in which music, video files, and other data files were analyzed. The study indicated that the amount of data being produced by systems is doubling each year. This is the general concept of Moore’s law.
How Moore’s law might change
We might be experiencing the last breadths of Moore’s law when it comes to microprocessor power. If there is an increase in processing power, other computing fields will have to be examined. Looking at the cloud computing capabilities, computing is going to advance since the cloud provides shareable resources, processing capacities which will improve innovation and more effectiveness in business engagements.
To increase the processing power of microprocessors, there is a new technology called Photonics that is being researched and tested. Intel is testing photonics in Texas. Photonics uses light to transfer data much faster without signal loss. This lowers the generation of electricity and enables data to travel at the speed of light. This experiment will help Moore’s law to increase its process streams and capacities, starting a new cycle again.
How can AI pick up after Moore’s Law?
AI has become the next tech paradigm to go mainstream, and this makes AI require new power because Moore’s law and Dennard scaling are not strong enough to keep up. Moore’s law states that the number of transistors in specific regions of a chip will double after two years. In Dennard scaling, the amount of power required to keep up transistors is shrinking.
Intel over the last few years has reduced its pace of generating new chips which have denser and smaller transistors. The gains in efficiency of small transistors also came to a halt a few years ago, and this made power consumption a problem.
There is a need in how AI which will handle even much more loads of data will require more powerful chips.
Scientists and Big data
Big data sources are very many. For instance, data collected in the physical world is shockingly diverse and in huge loads. Measurements of RF signals, vibrations, pressure, magnetism, sound, temperature, light, voltage and so on are all recorded in different forms and high velocities.
Where is Moore’s law headed?
A transistor’s physical length and other important dimensions of key logic would progressively shrink until 2028, but 3D concepts have taken center stage. The industry concerned with memory have accepted 3D architectures to boost NAND flash capacity and ease the pressure of miniaturization. This does not mean that there is an end to Moore’s law.
Conclusion
Moore’s law is still effective in the handling of big data, but there is a more economic sense in using the 3D architecture. AI is going to bring increasing demand for processing power in the years to come, and the chip manufacturing companies have to produce really fast processors to handle the workload.
