Microsoft is working on developing AI chips specifically for the next-generation HoloLens, enabling the device to recognize both speech and images directly. This advancement was announced at the recent CVPR conference in Hawaii. Unlike the current model, which sends data to the cloud for processing, the new HoloLens will be capable of performing these tasks locally, reducing latency and improving real-time interaction.
According to sources like Lei Feng Network, Microsoft's second-generation Holographic Processing Unit (HPU), known as HPU II, is currently under development for integration into future HoloLens models. Although there's no official timeline yet, this marks a significant leap forward in the capabilities of mixed reality devices.
The original HPU chip, launched in August 2016, featured a custom-designed 28nm digital signal processor (DSP) from TSMC with 24 Tensilica DSP cores, capable of processing up to 1 trillion instructions per second. It included 8MB of SRAM and 1GB of LPDDR3 memory, offering a performance boost of 200 times over traditional software solutions while consuming 10 watts less power.
Shen Xiangyang, Executive Vice President of Microsoft and Head of the Artificial Intelligence and Research Division, spoke at the CVPR conference about the commercial success of computer vision technologies. He highlighted Microsoft's commitment to advancing computer vision through research initiatives spanning 25 years. His presentation touched upon various applications, including Microsoft Pix, HoloLens, and Cognitive Services, showcasing tools like IRIS Interactive Vision Learning Service for developers.
Despite the excitement surrounding the HPU II announcement, some attendees noted that the focus seemed overly promotional, leading some to leave early. Nevertheless, Microsoft remains dedicated to integrating advanced AI capabilities into its products, reflecting broader industry trends where companies like Apple, Google, and Amazon are also investing heavily in proprietary AI hardware.
Intel, traditionally dominant in the chip market, now faces stiff competition as artificial intelligence demands specialized processors capable of handling complex neural networks efficiently. Microsoft has been proactive in leveraging FPGA technology alongside partnerships with firms like Altera (now part of Intel) to optimize AI workloads. Their efforts include translating vast amounts of Wikipedia content into Spanish within seconds using thousands of these chips, demonstrating unprecedented computational power.
As Satya Nadella, CEO of Microsoft, emphasizes, embedding AI directly onto devices represents a key strategic goal. By customizing chips tailored to specific applications, Microsoft aims to lead the way in delivering seamless AI experiences across diverse platforms. Kevin Scott, CTO of Microsoft, echoed this sentiment, stating, "We must tailor our chips to suit the scenarios and applications we’re building."
Overall, Microsoft’s investment in AI chips signals a pivotal moment in the evolution of wearable technology and intelligent systems, promising enhanced functionality and user engagement.
Perkins 0-20KW Diesel Generator
Perkins 0-20Kw Diesel Generator,Perkins Power Generator Set,Perkins Shanghai Power Generator Set,Perkins Power Generator
Shanghai Kosta Electric Co., Ltd. , https://www.generatorkosta.com