과제정보
이 논문은 2023년도 정부(과학기술정보통신부)의 재원으로 정보통신기획평가원의 지원을 받아 수행된 연구임[No.2022-0-00018, 거대 인공지능 학습을 위한 K-인공두뇌 반도체 개발].
참고문헌
- Epoch, Parameter, Compute and Data Trends in Machine Learning, 2023. 8. 21., Retrieved from https://epochai.org/mlinputs/visualization
- Yahoo!finance, "ChatGPT on track to surpass 100 million users faster than TikTok or Instagram: UBS," 2023. 2. 3.
- A. Vaswani et al., "Attention is all you need," in Proc. Int. Conf. Neural Inform. Process. Syst., (Long Beach, CA, USA), Dec. 2017, pp. 6000-6010.
- https://chat.openai.com/
- A. Dosovitskiy et al., "An image is worth 16×16 words: Transformers for image recognition at scale," arXiv preprint, CoRR, 2020, arXiv: 2010.11929.
- OpenAI, "GPT-4 technical report," arXiv preprint, CoRR, 2023, arXiv: 2303.08774.
- S. Bubeck et al., "Sparks of artificial general intelligence: Early experiments with gpt-4," arXiv preprint, CoRR, 2023, arXiv: 2303.12712.
- Wired, "OpenAI's CEO Says the Age of Giant AI Models Is Already Over," 2023. 4. 17.
- https://www.sapeon.com/
- Korea IT News, "Sapeon 'Enhancing omputation with next-generation memory artificial brain 'CIM''," 2022. 11. 17.
- N. Jouppi et al., "Tpu v4: An optically reconfigurable supercomputer for machine learning with hardware support for embeddings," in Proc, ISCA 2023, (Orlando, FL, USA), June 2023, pp. 1-14.
- Graphcore, "Graphcore documents," accessed online at https://docs.graphcore.ai/en/latest/.
- E. Talpes et al., "The microarchitecture of dojo, tesla's exa-scale computer," IEEE Micro, vol. 43, no. 3, 2023, pp. 31-39.
- FuriosaAI, FuriosaAI WARBOY: High performance inference chip for the most advanced vision applications.
- FuriosaAI, FuriosaAI NPU & SDK 0.10.0 Documents, Accessed online at https://furiosa-ai.github.io/docs/latest/en/
- https://kakaoicloud.co.kr/service/detail/1-44
- NVIDIA, NVIDIA H100 Tensor Core GPU Architecture: EXCEPTIONAL PERFORMANCE, SCALABILITY, AND SECURITY FOR THE DATA CENTER v1.04, 2023.
- Head Topics, "Nvidia's AI Chips Are Pulling Ahead in the Cloud," 2023. 8. 18.
- Rebellions, "ATOM: 5nm Versatile Inference SoC," 2023.
- https://enterprise.kt.com/pd/P_PD_NE_00_316.do
- A. Firoozshahian et al., "MTIA: First generation silicon targeting Meta's recommendation systems," in Proc. ISCA 2023, (Orlando, FL, USA), June 2023, pp. 1-13.
- CNN, "The big bottleneck for AI: A shortage of powerful chips," 2023. 8. 6.