旧金山，10月1日：总部位于旧金山的半导体初创公司Volantis周四表示，已筹集8800万美元风险投资，旨在利用已存在于数亿部iPhone中的技术，解决人工智能芯片面临的一项关键挑战。
SAN FRANCISCO, Oct 1 : San Francisco-based semiconductor startup Volantis on Thursday said it has raised $88 million in venture capital in an effort to solve a key challenge for AI chips by tapping a technology that is already sitting in hundreds of millions of iPhones.
无论是英伟达还是超微半导体（AMD）等竞争对手，其人工智能芯片的根本限制在于芯片的计算部分与存储人工智能模型并向计算芯片提供数据的内存芯片之间的通信效率。英伟达和AMD通过在计算芯片周围环绕昂贵的高带宽内存芯片来解决这一问题。
The fundamental limit for AI chips from both Nvidia and rivals like Advanced Micro Devices is how well the computing parts of a chip can talk to the memory chips where an AI model lives and feeds data to the computing chip. Nvidia and AMD solve the problem by encircling their computing chips with expensive high-bandwidth memory chips.
但由于连接内存芯片的微小电线传输距离有限，即使是英伟达目前最好的产品，每个图形处理器（GPU）也只能容纳八个这样的内存芯片。Volantis表示，它正在开发一种利用激光束在计算芯片和内存芯片之间传输数据的方法，这消除了传输距离的限制，使其能够在GPU周围集成220个内存芯片。
But even Nvidia's best current offerings can hold only eight of those memory chips for each of its graphics processing units (GPUs) because of the limited reach of the tiny electrical wires that connect the memory chips. Volantis says it is developing a way to send data between computing chips and memory chips using beams of laser light, which eliminates the reach problem and would let it pack 220 memory chips around a GPU.
为此，Volantis采用了垂直腔面发射激光器（VCSEL）。虽然这种微型激光器听起来很新颖，但它们为苹果等众多设备的面部识别功能提供动力，而苹果在过去十年中大力投资以加强其供应链。
To do that, Volantis is tapping what are called vertical-cavity surface-emitting lasers, or VCSELs. While the tiny lasers sound exotic, they power the facial recognition features in many devices from Apple, which invested heavily in bolstering the supply chain over the past decade.
Volantis首席执行官兼联合创始人塔帕·戈什（Tapa Ghosh）在接受采访时表示，通过整合相对成熟的组件，Volantis希望避免开发人工智能芯片过程中的一些供应链瓶颈，并计划明年推出一款能够加速人工智能编码等任务的芯片。
By bringing together relatively well-known components, Volantis is hoping to avoid some of the supply chain bottlenecks in developing AI chips and deliver a chip next year that would speed up tasks like AI coding, said Tapa Ghosh, CEO and co-founder of Volantis, in an interview.
“先进封装总是值得尊重的——它绝非小事——但这未必是什么新鲜事，”Ghosh说道。“如果我们的项目成功了，没人会因此获得诺贝尔奖，但好消息是，他们也不需要。”本轮融资由Stripe资深人士Lachy Groom和Abstract Ventures领投。曾支持过谷歌和亚马逊早期发展的John Doerr也参与了投资，VXI Capital、Triatomic和Susa Ventures同样参与其中。本轮融资还包括来自人工智能播客主持人Dwarkesh Patel、人工智能芯片资深人士Naveen Rao以及Anthropic研究员Sholto Douglas的天使投资。
"Advanced packaging is always to be respected - it's never trivial - but it's not necessarily a new thing to do," Ghosh said. "No one is going to win a Nobel Prize if our project works, but the good news is, they won't need to." The funding round was led by Stripe veteran Lachy Groom and Abstract Ventures. John Doerr, who backed Google and Amazon.com in their early days, also participated, as did VXI Capital, Triatomic and Susa Ventures. The round also includes angel investments from AI podcaster Dwarkesh Patel, AI chip veteran Naveen Rao and Anthropic researcher Sholto Douglas.