Artificial intelligence and the semiconductor industry
- Autores: Makushin M.1
-
Afiliações:
- НОБ «Военные науки и оборонная промышленность» БРЭ
- Edição: Nº 7 (2025)
- Páginas: 46-55
- Seção: Artificial Intelligence
- URL: https://journals.eco-vector.com/1992-4178/article/view/690208
- DOI: https://doi.org/10.22184/1992-4178.2025.249.7.46.55
- ID: 690208
Citar
Texto integral



Resumo
The development of artificial intelligence is impossible without the creation of a specialized element base and CAD tools for its design. At the same time, the introduction of AI systems affects the development of new CAD systems and their functioning.
Palavras-chave
Texto integral

Sobre autores
M. Makushin
НОБ «Военные науки и оборонная промышленность» БРЭ
Autor responsável pela correspondência
Email: kys@electronics.ru
ведущий научный редактор
RússiaBibliografia
- Макушин М. Искусственный интеллект: новые архитектуры ИИ процессоров и расширение роли в проектировании ИС // ЭЛЕКТРОНИКА: Наука, Технология, Бизнес. 2024. № 10 (00241). С. 138–147.
- Bailey B. Multi-Modal AI In EDA Development Flows // Semiconductor Engineering. July 31st, 2025.
- Moyer B. The Best DRAMs For Artificial Intelligence //Semiconductor Engineering. June 12th, 2025.
- Allan L. Mobile Chip Challenges In The AI Era // Semiconductor Engineering. June 5th, 2025.
- Allan L. AI Pushes High-End Mobile From SoCs To Multi-Die // Semiconductor Engineering. July 8th, 2025.
Arquivos suplementares
Arquivos suplementares
Ação
1.
JATS XML
2.
Fig. 1. Comparison of different synchronous DRAM families. Capacities are calculated on a per-die basis (not per-stack for HBM). No one family has advantages in all areas. DDR and LPDDR DRAM may have comparable bandwidth and capacity, but the main differentiator is cost. Source: Bryon Moyer/Semiconductor Engineering
Baixar (153KB)
3.
Fig. 2. A smartphone use case where a large language model (LLM) or AI engine must be available in storage. Source: Synopsys
Baixar (116KB)
Baixar (226KB)
