{"version":"1.0","provider_name":"da.news","provider_url":"https:\/\/da.news\/en\/","title":"Performance above all else - da.news","type":"rich","width":600,"height":338,"html":"<blockquote class=\"wp-embedded-content\" data-secret=\"i3vXerMSDk\"><a href=\"https:\/\/da.news\/en\/performance-above-all-else\/\">Performance above all<\/a><\/blockquote><iframe sandbox=\"allow-scripts\" security=\"restricted\" src=\"https:\/\/da.news\/en\/performance-above-all-else\/embed\/#?secret=i3vXerMSDk\" width=\"600\" height=\"338\" title=\"&amp;quot;Performance above all else&amp;quot; &#x2013; da.news\" data-secret=\"i3vXerMSDk\" frameborder=\"0\" marginwidth=\"0\" marginheight=\"0\" scrolling=\"no\" class=\"wp-embedded-content\"><\/iframe><script type=\"text\/javascript\">\n\/* <![CDATA[ *\/\n\/*! This file is auto-generated *\/\n!function(d,l){\"use strict\";l.querySelector&&d.addEventListener&&\"undefined\"!=typeof URL&&(d.wp=d.wp||{},d.wp.receiveEmbedMessage||(d.wp.receiveEmbedMessage=function(e){var t=e.data;if((t||t.secret||t.message||t.value)&&!\/[^a-zA-Z0-9]\/.test(t.secret)){for(var s,r,n,a=l.querySelectorAll('iframe[data-secret=\"'+t.secret+'\"]'),o=l.querySelectorAll('blockquote[data-secret=\"'+t.secret+'\"]'),c=new RegExp(\"^https?:$\",\"i\"),i=0;i<o.length;i++)o[i].style.display=\"none\";for(i=0;i<a.length;i++)s=a[i],e.source===s.contentWindow&&(s.removeAttribute(\"style\"),\"height\"===t.message?(1e3<(r=parseInt(t.value,10))?r=1e3:~~r<200&&(r=200),s.height=r):\"link\"===t.message&&(r=new URL(s.getAttribute(\"src\")),n=new URL(t.value),c.test(n.protocol))&&n.host===r.host&&l.activeElement===s&&(d.top.location.href=t.value))}},d.addEventListener(\"message\",d.wp.receiveEmbedMessage,!1),l.addEventListener(\"DOMContentLoaded\",function(){for(var e,t,s=l.querySelectorAll(\"iframe.wp-embedded-content\"),r=0;r<s.length;r++)(t=(e=s[r]).getAttribute(\"data-secret\"))||(t=Math.random().toString(36).substring(2,12),e.src+=\"#?secret=\"+t,e.setAttribute(\"data-secret\",t)),e.contentWindow.postMessage({message:\"ready\",secret:t},\"*\")},!1)))}(window,document);\n\/\/# sourceURL=https:\/\/da.news\/wp-includes\/js\/wp-embed.min.js\n\/* ]]> *\/\n<\/script>\n","thumbnail_width":1024,"thumbnail_height":683,"description":"Chip innovations are becoming ever smaller and more powerful. Computer chips are the foundation of our digital world. No smartphone, no computer, no car, and no remote control can function without them. And one day, not so far off, they will likely be implanted in our heads as well. Artificial intelligence (AI), in particular, is driving the demand for more computing power in the smallest possible space. Chip manufacturers are reaching the limits of physical technology. European startups and universities are researching chip innovations to satisfy the growing hunger for computing power. In 1971, the chip company Intel launched the first commercially successful microchip, the \"4004.\" At the time, it was a sensation; today, however, the three-by-four-millimeter computing chip seems clunky. And, above all, underpowered: it contained a mere 2,300 transistors\u2014an unimaginable number back then, ridiculously few by today's standards. Transistors remain the heart of every computer chip. Tiny switches that toggle between on and off, between zero and one. These switches lay the foundation of our digital world. The more transistors, the more powerful the chip. And the global demand for computing power and data storage is exploding. For years, the chip industry has been preoccupied with one question above all: How many more of these transistors can fit on a single chip? Over the decades, transistors have become ever smaller, so that today they consist of only a handful of atoms. Today, these switches are thinner than a human hair, smaller than a red blood cell, and equipped with kilometers of wiring. A tiny switch\u2014500,000 times smaller than a millimeter\u2014that has become absolutely indispensable for our daily lives. 200 million transistors can fit on one square millimeter, and even tens of billions on a single chip. But in the near future, this shrinking experiment in semiconductor science will reach its physical limit. Currently, we use chip technologies such as CPUs\u2014that is, Central Processing Units\u2014for computers and smartphones. The alternative technology of GPUs \u2013 Graphics Processing Units \u2013 also known as graphics cards, was originally developed for images, video content, and 3D graphics for computer screens. Chip manufacturer Nvidia made a name for itself with these chips and is currently riding the wave of demand for AI chips. These graphics cards have the advantage of being able to perform parallel tasks and handle many tasks simultaneously \u2013 exactly what AI needs to work efficiently. For AI, graphics chips have proven to be the best solution at present; GPUs are the fallback option for AI algorithms. Currently, there are simply no new approaches for AI chips. What there is, however, is a great deal of research and innovation. Although every nanometer on your chip is already packed with switches, there is still unused space, specifically in terms of height. This is what Semron is researching. The Dresden-based startup has developed chips that bring AI directly to end devices such as smartphones and headphones. This allows data to be processed locally on the devices, which is particularly advantageous for sensitive information. For the chips to be powerful enough, they must be as small, compact, cost-effective, and energy-efficient as possible. Co-founder Aron Kirschen knows that simply stacking three, four, or even five chips in a package isn't enough. Five times the performance combined with the cost of five chips is useless when a 1000-fold increase in performance is needed. Instead, Semron plans to apply multiple chip layers on top of each other during the manufacturing process. The patented semiconductor technology \"CapRAM\" allows AI models to be processed locally. This already works with memory chips, like those found in smartphones. Our phones contain up to 200 layers of memory. This technique proves more difficult for processors\u2014the computer components of a chip. Firstly, transistors aren't so easy to stack. And secondly, this design requires more energy, and at higher energy densities, the chip risks overheating. Semron boasts that it has solved this problem. Now the team faces the major challenge of getting chip manufacturers to implement the patented manufacturing process and produce the chips in large batches. The idea alone, however, is not enough to satisfy the ever-increasing demand for computing power. The chips also need to be manufactured \u2013 and that doesn't happen in Germany or even in Europe. It's comparable to the industrial revolution 150 years ago."}