Rúguǒ shuō jìndài shèhuì shì gōngyè dàidòngzhe shìjiè gèguó de jìnbù; nàme zài xiàndài shèhuì biàn shì kējì yùnzhuǎn shìjiè qiángguó páimíng. Shuō dào dāngxià zuì rèmén de kējì, réngōng zhìnéng juéduì kān chēng shì běn shìjì zuì jù diānfù xìng de jìshù. Jìnnián lái, réngōng zhìnéng yuè lái yuè jiējìn wǒmen de shēnghuó, yǔyīn zhùshǒu, rén liǎn shìbié, zhìnéng jiāotōng, wú rén chē děng děng dōu xiǎnshìzhe réngōng zhìnéng de cúnzài hé qiángdà. Rán’ér, suízhe quánqiú shùjù liàng de bùduàn tíshēng dǎozhì jìsuànjī de chǔlǐ nénglì xiāngduì zhìhòu, gāo xìngnéng de jìsuàn hé néng hào chéngwéile réngōng zhìnéng xiàng qián màijìn de zuìdà píngjǐng. Optalysys gōngsī chuàngshǐ rén jiān shǒuxí zhíxíng guān níkè·niǔ (Nick New) chénglì yú 2013 nián de Optalysys gōngsī shì yījiā cóng jiànqiáo dàxué zǒu chūlái de chūchuàng qǐyè, yěshì quánqiú shǒu jiā zhìlì yú kāifā guāngxué jìsuàn bìng huòdé zhuānlì rènzhèng de gōngsī.Optalysys de guāngxué xiétóng chǔlǐ qì yǐjīng shíxiànle jīyú túxiàng de shēndù xuéxí móshì, yě jiùshì wèi wǒmen shúzhī de juàn jī shénjīng wǎngluò (Convolutional Neural Networks). Zài diànzǐ chǔlǐ qì zhōng zhèxiē cāozuò shì chuàn xíng de, shǐyòng duō gè Tensor Core jìnxíng dī fēnbiàn lǜ de cāozuò. Tōngguò Optalysys de guāngxué jìshù néng shǐ zhèxiē cāozuò yǐ guāngsù bìngxíng, tíshēngle gāi móshì xià xùnliàn shùjù shēn céngcì wājué yǐjí tuīlǐ de sùdù. Zhè duìyú xūyào duì dàliàng túpiàn shùjù jìn háng gāosù chǔlǐ de yìngyòng lái shuō shì jué jiā de jiějué fāng’àn, zhè qízhōng shèjí de hángyè bāokuò wú rén jiàshǐ, yīxué túxiàng fēnxī yǐjí ānquán xìtǒng. Mùqián wéizhǐ, jíbiàn shì zuì kuài de diànzǐ chǔlǐ qì yě wúfǎ shíxiàn zhè yī gōngnéng. Optalysys gōngsī de chuàngshǐ rén jiān shǒuxí zhíxíng guān níkè·niǔ (Nick New) biǎoshì:“Réngōng zhìnéng lǐngyù jìnnián lái fāzhǎn xùnměng, qí miànlín de zuìdà tiǎozhàn jiùshì rúhé gāoxiào de tíshēng shùjù chǔlǐ nénglì yǐ mǎnzú gè hángyè de xūqiú. Zhōngguó zài réngōng zhìnéng jìshù lǐngyù chǔyú lǐngxiān dìwèi, ér zhè xiàng jìshù yě bèi pǔbiàn de yìngyòng yú rénmen de shēnghuó. Rán’ér, wèile chōngfèn fāhuī réngōng zhìnéng de qiánlì, cóngyè zhě xūyào duì mùqián shǐyòng de chǔlǐ jìshù jìnxíng zhòngdà de géxīn. Chuántǒng de guī jìshù chǔlǐ zài chǔlǐ sùdù yǐjí néng hào fāngmiàn dōu miànlínzhe yánjùn de tiǎozhàn. Yīncǐ,Optalysys tígōng de guāngxué réngōng zhìnéng chéngwéile bǎituō zhè yī kùnjìng de zuì jiā jiějué fāng’àn.” Níkè·niǔ jìxù bǔchōng dào:“Jīyú túxiàng de shēndù xuéxí móxíng duìyú jìsuàn de yāoqiú jí gāo, wǒmen cǎiyòng de shì guāngxué chǔlǐ jìshù ér fēi diànzǐ chǔlǐ jìshù. Tōngguò jiéhé dāngxià zuì xiānjìn de yèjīng wēi xiǎnshìqì,Optalysys de xìtǒng jùyǒu zhuóyuè de chǔlǐ nénglì, qí hào diàn liàng jǐn wèi diànzǐ chǔlǐ jìshù de yī xiǎo bùfèn. Tōngguò zhè zhǒng fāngshì, wǒmen nénggòu shēngchǎn dī gōnglǜ de guāng chǔlǐ qì, bìng nénggòu yìngyòng dào duō gè lǐngyù. Zhè xiàng jìshù jùyǒu jí qiáng de kě kuòzhǎn xìng. Wǒmen de dì yīgè guāngxué xiétóng chǔlǐ qì xìtǒng suǒ tígōng de chǔlǐ nénglì bǐ yīgè dǐngjí de túxíng kǎ hái yàoqiáng, qí gōng hào jǐn wèi túxíng kǎ de sì fēn zhī yī. Wǒmen jiē xiàlái tuīchū de xìtǒng chǔlǐ sùdù jiāng bǐ xiǎnkǎ kuài 300 bèi yǐshàng qiě gèng shěng diàn, ér zhè jǐnjǐn shìgè kāishǐ.” Hé zhòngduō chūchuàng gōngsī yīyàng,Optalysys zài chuànglì zhī chū yě yù dàole hěnduō tiǎozhàn. Mùqián réngōng zhìnéng lǐngyù háishì yīkào diànzǐ jìshù, duìyú Optalysys suǒ tígōng de diānfù xìng jìshù hái xūyào gǎibiàn rénmen gùyǒu de sīwéi fāngshì.Optalysys zhuānzhù yú wèi guāngxué chǔlǐ qì tígōng wú fèng xiánjiē, zhè jiù xūyào wèi cóngshì guāngxué gōngnéng de shǐyòng zhě tígōng zuì jiǎnbiàn de shǐyòng fāngshì, bìng shǐ qí nénggòu kuàisù biānchéng. Níkè·niǔ biǎoshì:“Chǎnpǐn de yì yòng xìng bùjǐn duì wǒmen lái shuō zhì guān zhòngyào, duìyú zhěnggè hángyè lǐngyù de zhuānyè rénshì yě fēicháng zhòngyào.” Zhōngguó shì guāng jìsuàn zuìdà de yìngyòng shìchǎng, tóngshí yěshì Optalysys zuì zhòngshì de shìchǎng zhī yī. Mùqián,Optalysys zhèng yǔ duō jiā zhōngguó tóuzī gōngsī hé kējì gōngsī jìnxíng cuōshāng yǐ jiànlì zhànlüè hézuò huǒbàn guānxì, yǐbiàn chǎnpǐn zài zhōngguó déyǐ shēngchǎn bìng kuàisù tóufàng shìchǎng. Zài tán dào duì zhōngguó shìchǎng de zhǎnwàng shí, níkè·niǔ biǎoshì:“Wǒmen de mùbiāo shì tōngguò jíchéng zhì chuángǎnqì de fāngshì, shǐ wǒmen de jìshù nénggòu wèi yún xìtǒng, fúwùqì xìtǒng, běndì fúwùqì yǐjí yídòng jiějué fāng’àn gōngyìng shāng tígōng zúgòu qiángdà de chǔlǐ nénglì, yǐ yìngduì wèilái suǒ miànlín de tiǎozhàn. Wèilái, wǒmen yě zhèngzài xúnzhǎo yī gè héshì de qìjī luòhù zhōngguó.”展开volume_up1443 / 5000

翻译结果

If modern society is the industry that drives the progress of all countries in the world; then in modern society, it is the ranking of the world’s powers of science and technology. Speaking of the hottest technology at the moment, artificial intelligence is definitely the most disruptive technology of this century. In recent years, artificial intelligence has become closer to our lives. Voice assistants, face recognition, intelligent transportation, unmanned vehicles, etc. all show the existence and power of artificial intelligence. However, as the global data volume continues to increase, the processing power of computers is relatively lagging, and high-performance computing and energy consumption have become the biggest bottleneck for artificial intelligence to move forward.

Nick New, founder and CEO of Optalysys

Founded in 2013, Optalysys is a start-up from Cambridge University and the first company in the world dedicated to developing optical computing and obtaining patent certification. Optalysys’ optical coprocessor has implemented an image-based deep learning mode, which is also known as Convolutional Neural Networks (Convolutional Neural Networks). In the electronic processor, these operations are serial, using multiple Tensor Cores for low-resolution operations. Through Optalysys’ optical technology, these operations can be paralleled at the speed of light, which improves the speed of deep mining and reasoning of training data in this mode. This is an excellent solution for applications that require high-speed processing of large amounts of image data. The industries involved include unmanned driving, medical image analysis, and security systems. So far, even the fastest electronic processors cannot achieve this function.

Nick New, the founder and CEO of Optalysys, said: “The field of artificial intelligence has developed rapidly in recent years. The biggest challenge it faces is how to efficiently improve data processing capabilities to meet the needs of various industries. The field of intelligent technology is in a leading position, and this technology is widely used in people’s lives. However, in order to fully utilize the potential of artificial intelligence, practitioners need to make major innovations in the processing technology currently used. Traditional silicon technology processing Both processing speed and energy consumption are facing severe challenges. Therefore, the optical artificial intelligence provided by Optalysys has become the best solution to get rid of this dilemma.”

Nick New continued to add: “The image-based deep learning model has extremely high requirements for calculations. We use optical processing technology instead of electronic processing technology. By combining the most advanced liquid crystal micro-display, Optalysys’ system has excellent Processing power, its power consumption is only a small part of the electronic processing technology. In this way, we can produce low-power optical processors, and can be applied to many fields. This technology has strong scalability The processing power provided by our first optical co-processor system is stronger than that of a top graphics card, and its power consumption is only a quarter of that of a graphics card. The processing speed of our next system will be faster than that of a graphics card More than 300 times and save more power, and this is just the beginning.”

Like many startups, Optalysys also encountered many challenges at the beginning of its establishment. At present, the field of artificial intelligence still relies on electronic technology, and the disruptive technology provided by Optalysys also needs to change people’s inherent ways of thinking. Optalysys focuses on providing seamless connections for optical processors, which requires users engaged in optical functions to provide the easiest way to use and enable rapid programming. Nick New said: “The ease of use of products is not only important to us, but also to professionals in the entire industry.”

已是最后文章

下一篇: Technology: Pair AI with optical scanning to achieve real product certification