在韩国产业通商资源部长官金正官宣布的这一愿景中，韩国计划到2047年总计投入700万亿韩元（约合 4760亿美元），新建10家半导体制造厂，使韩国成为全球最大的芯片生产中心。 当天，三星电子、SK海力士、FuriosaAI等主要半导体企业和AI初创企业参加了由韩国总统李在明主持 的政府间半导体产业培育会议，并分享了上述蓝图。 根据该计划，到2032年，韩国政府将投资2159亿韩元，开发神经处理单元（NPU）和内存处理（PIM） 等下一代存储芯片技术。 据韩国产业通商资源部称，这些技术有望实现比高带宽存储器（HBM）更先进的人工智能推理能力。 韩国政府将在未来5年内投入1.27万亿韩元，用于开发适合人工智能的半导体，到2031年进一步投入 2601亿韩元用于化合物半导体，到2031年投入3606亿韩元用于先进封装技术。 此外，在竞争力相对较弱的系统半导体领域，还将建立需求企业和无晶圆厂企业共同开发设备上的人工 智能技术和中间技术半导体的自主技术的合作生态系统，以提高韩国的能力。 中间技术半导体是指不像人工智能芯片那么先进，但仍需要先进工程技术的芯片，如汽车微控制器单元 和电源管理集成芯片。 为了帮助国内产业制造 ...

Core Viewpoint - FuriosaAI, a South Korean chip startup, aims to compete with Nvidia by leveraging its unique Tensor Contraction Processor (TCP) architecture to enhance AI performance and efficiency [2][3]. Group 1: Company Overview - FuriosaAI was founded in 2017 by June Paik, a former engineer at Samsung and AMD, with a vision for dedicated chips for deep learning workloads [2]. - The company launched its first-generation Neural Processing Unit (NPU) in 2021, manufactured by Samsung using a 14nm process, which performed well in MLPerf benchmarks [2]. Group 2: Product Development - The second-generation chip, RNGD (Renegade), is being developed over a three-year project initiated in 2021, focusing on generative AI and language models [3]. - RNGD is manufactured using TSMC's 5nm process, featuring 48GB of HBM3 memory, 1.5TB/s memory bandwidth, and 512 TFLOPS of FP8 performance with a maximum power consumption of 180W [3]. Group 3: System Integration - FuriosaAI is working on a complete system based on the RNGD card, the NXT RNGD server, which will include eight RNGD cards, totaling 384GB of HBM3 memory and 4 petaFLOPS of FP8 performance at a thermal design power (TDP) of 3kW [4]. - The NXT RNGD server aims to outperform traditional GPU-based systems, targeting the same market as Nvidia's H100 GPU [4]. Group 4: Performance Comparison - The Nvidia H100 GPU features 80GB of HBM2 memory, 2TB/s memory bandwidth, and 1513 TFLOPS peak performance, with a TDP of 350W for PCIe versions and up to 700W for SXM versions [5]. - FuriosaAI claims that RNGD's performance exceeds Nvidia's by three times when running large language models on a per-watt basis [5]. Group 5: Architectural Innovation - The TCP architecture is designed to minimize data movement, which is a significant energy consumer, by maximizing data reuse stored in on-chip memory [6]. - The architecture improves abstraction layers to overcome limitations of traditional GPU architectures, ensuring efficient data access and high throughput [7]. Group 6: Market Adoption and Client Engagement - FuriosaAI has gained traction with clients like LG AI Research, which reported that RNGD could deliver approximately 3.5 times the tokens per rack compared to previous GPU solutions [8]. - The company has attracted attention from major cloud computing firms, including Meta, which expressed interest in acquiring FuriosaAI [8]. Group 7: Future Plans and Funding - FuriosaAI completed a $125 million bridge financing round, bringing total funding to $246 million, and is focusing on ramping up RNGD production for global customer engagement by early 2026 [9].

Meta still spends billions on outside GPUs. Despite its internal work, Meta depends heavily on market leader Nvidia Corp. for chips. Deal terms are not public. But in August, Meta wants to cut Nvidia relianceMeta wants to cut Meta’s chip progress is too slow for CEO Mark Zuckerberg. People familiar with the matter said Zuckerberg is not happy with the pace, and leadership has been searching for outside help to boost the work. The Meta spokesperson disagreed. The company said, “our custom silicon work is pr ...

Group 1 - Rebellions Inc. raised $250 million in Series C funding, achieving a valuation of $1.4 billion, to mass-produce chips and accelerate product development in the competitive AI market [1][2] - The funding round included participation from strategic partner Arm Holdings, Samsung Ventures, Pegatron Corp. VC, and Lion X Ventures Pte, with continued support from existing investors like Korea Development Bank and Korelya Capital [1][2] - Rebellions aims to prepare for mass production of its flagship product "RebelQuark" and accelerate the development of new product lines, including "RebelIO," while also expanding recruitment and overseas operations [2] Group 2 - Groq Inc. completed a $750 million funding round, achieving a valuation of $6.9 billion, indicating strong investor interest in companies addressing AI computing and chip shortages [2] - Local competitor FuriosaAI is preparing for a potential Pre-IPO funding round exceeding $300 million [2] - Rebellions' CFO expressed confidence that this funding demonstrates the capability of the Korean capital market to nurture AI semiconductor companies globally, aiming to position Korea among the top three in the global AI landscape [2]

Core Viewpoint - FuriosaAI, a Seoul-based chip startup, is preparing for a potential IPO pre-funding round exceeding $300 million, aiming to challenge Nvidia's market position [1]. Group 1: Funding and IPO Plans - The company has selected several international investment banks to submit proposals for its D round financing, which may start as early as January. The funds will primarily be used for developing the company's third-generation AI chips, and this round could serve as a Pre-IPO round ahead of a potential listing in 2027 [2]. - Earlier this year, FuriosaAI successfully raised $125 million, surpassing its funding target of approximately $80 million. The company also rejected an $800 million acquisition offer from Meta Platforms [3]. Group 2: Company Background and Technology - FuriosaAI was co-founded in 2017 by June Paik, who previously worked at Samsung Electronics and AMD, focusing on the development of AI inference chips and related services [4]. - The company's AI chip, RNGD (pronounced "renegade"), boasts a reasoning performance per watt that is 2.25 times higher than that of graphics processors manufactured by companies like Nvidia [5].

Group 1: South Korea's Semiconductor Dominance - South Korea's semiconductor industry, led by SK Hynix and Samsung, holds a significant market share in DRAM and NAND, with a combined 70% share in DRAM and nearly half in NAND [1][3] - The country's semiconductor exports are projected to reach a record $142 billion in 2024, accounting for one-fifth of total exports [1] Group 2: AI Chip Development - The South Korean government announced a plan to invest 150 trillion KRW (approximately $108 billion) over the next five years to develop AI, semiconductors, biotechnology, defense, robotics, and green transportation [3] - Rebellions, a South Korean AI chip company, has raised $225 million since its inception in 2020 and is known for its energy-efficient AI computing chips [4] - Sapeon Korea, an AI chip startup under SK Telecom, became South Korea's first AI chip unicorn with a valuation of 1.3 trillion KRW (approximately $1 billion) [4] Group 3: SiC Semiconductor Focus - South Korea aims to increase its self-sufficiency in SiC (silicon carbide) power semiconductors from 10% to 20% by 2028, with a government investment of 902 billion KRW [9][10] - SK Siltron CSS holds a 6% share of the global SiC wafer market and has signed a long-term supply contract with Infineon Technologies [10] - The South Korean government is also focusing on developing core technologies for SiC chips and has plans to build demonstration infrastructure [10][11] Group 4: Industry Challenges and Opportunities - South Korea has approximately 160 fabless companies, significantly fewer than China's over 3,000, indicating a need for growth in this sector [16][17] - The local market's capacity to absorb semiconductor production remains a concern, especially in the context of geopolitical tensions affecting global supply chains [17]

FuriosaAI is preparing to raise more than $300 million in a potential pre-IPO funding round https://t.co/aogVFKw3Xt ...

South Korean AI Chip Startup FuriosaAI Raises $125 Million https://t.co/Da0zZLFO1c https://t.co/Da0zZLFO1c ...

South Korean AI Chip Startup FuriosaAI Raises $125 Millionhttps://t.co/gOUbL3eWnH https://t.co/bxfn125XJ4 ...

June Paik表示，FuriosaAI的使命是让AI计算变得"更便宜、更节能、更易获取"。他指出，即便是行业巨 头OpenAI，也因高昂的基础设施成本而难以实现盈利。 虽然FuriosaAI与市值4.2万亿美元、被视为AI芯片行业标准的英伟达相比仍显渺小，但June Paik表示有 信心凭借更先进、更高效的技术实现竞争优势，"我们的公司将在未来展现真正的价值"。 资料显示，June Paik于2017年与两位合伙人共同创办了FuriosaAI，他此前曾在三星电子和 AMD(AMD.US)工作。 这笔融资对这家已有八年历史的初创公司而言正值关键时刻。本月早些时候，FuriosaAI与LG集团签署 了一项重大合同。据此前的报道，该公司曾拒绝Meta(META.US)提出的8亿美元收购报价，选择继续独 立发展业务。 正寻求挑战芯片巨头英伟达(NVDA.US)的韩国芯片初创公司FuriosaAI在最新一轮融资中获得1.25亿美 元，以帮助其加快AI芯片的生产进度。 FuriosaAI在一份声明中表示，本轮融资金额超过原定的8000万美元目标，吸引了包括韩国开发银行 (KDB)、韩国产业银行(IBK)、Keisto ...