SEMIFIVE Concluded Mass Production Contract for AI Chip with HyperAccel
SEMIFIVE, a leading design solution provider, has announced a contract with HyperAccel to develop and mass-produce a generative AI chip called Bertha. The chip will be developed using 4nm process technology, with mass production targeted for Q1 2026. HyperAccel's LLM Processing Unit (LPU) is designed specifically for transformer-based Large Language Models, offering up to 2 times performance improvement and a 19 times better price/performance ratio compared to typical supercomputers.
SEMIFIVE specializes in SoC platforms and ASIC design solutions, focusing on AI chip expert SoC design platforms. The collaboration aims to provide cost-effective and power-efficient LLM features, potentially reducing data center operational expenses and expanding into industries requiring LLMs.
SEMIFIVE, un fornitore leader di soluzioni di design, ha annunciato un contratto con HyperAccel per sviluppare e produrre in serie un chip di intelligenza artificiale generativa chiamato Bertha. Il chip sarà sviluppato utilizzando la tecnologia di processo a 4nm, con la produzione di massa prevista per il Q1 2026. L'Unità di Elaborazione LLM (LPU) di HyperAccel è progettata specificamente per modelli di linguaggio di grandi dimensioni basati su trasformatore, offrendo un miglioramento delle prestazioni fino a 2 volte e un rapporto prezzo/performance 19 volte migliore rispetto ai supercomputer tradizionali.
SEMIFIVE si specializza in piattaforme SoC e soluzioni di design ASIC, concentrandosi su piattaforme di design SoC per esperti in chip AI. La collaborazione mira a fornire caratteristiche LLM economiche ed efficienti in termini di energia, potenzialmente riducendo le spese operative dei centri dati ed espandendosi in settori che richiedono LLM.
SEMIFIVE, un proveedor líder de soluciones de diseño, ha anunciado un contrato con HyperAccel para desarrollar y producir en masa un chip de IA generativa llamado Bertha. El chip se desarrollará utilizando tecnología de proceso de 4nm, con la producción masiva prevista para el Q1 2026. La Unidad de Procesamiento LLM (LPU) de HyperAccel está diseñada específicamente para modelos de lenguaje de gran tamaño basados en transformadores, ofreciendo hasta 2 veces de mejora en rendimiento y una relación calidad/precio 19 veces mejor en comparación con los supercomputadores típicos.
SEMIFIVE se especializa en plataformas SoC y soluciones de diseño ASIC, centrándose en plataformas de diseño SoC de expertos en chips de IA. La colaboración tiene como objetivo proporcionar características LLM rentables y eficientes en energía, lo que podría reducir los costos operativos del centro de datos y expandirse a industrias que requieren LLM.
SEMIFIVE는 선도적인 설계 솔루션 제공업체로서 HyperAccel과 협력하여 Bertha라는 생성적 AI 칩을 개발 및 대량 생산하는 계약을 발표했습니다. 이 칩은 4nm 공정 기술을 이용하여 개발될 예정이며, 대량 생산은 2026년 1분기를 목표로 하고 있습니다. HyperAccel의 LLM 처리 장치(LPU)는 변환기 기반의 대형 언어 모델을 위해 특별히 설계되었으며, 최대 2배 성능 향상과 19배 더 나은 가격/성능 비율을 제공합니다.
SEMIFIVE는 AI 칩 전문 SoC 설계 플랫폼에 중점을 두고 SoC 플랫폼과 ASIC 설계 솔루션을 전문으로 합니다. 이 협업은 비용 효율적이고 전력 효율적인 LLM 기능을 제공하여 데이터 센터 운영비용을 절감하고 LLM이 필요한 산업으로 확장할 수 있는 가능성을 염두에 두고 있습니다.
SEMIFIVE, un fournisseur leader de solutions de conception, a annoncé un contrat avec HyperAccel pour développer et produire en masse une puce d'IA générative appelée Bertha. La puce sera développée en utilisant la technologie de processus 4nm, avec une production en série prévue pour le 1er trimestre 2026. L'Unité de Traitement LLM (LPU) de HyperAccel est spécifiquement conçue pour les grands modèles de langage basés sur des transformateurs, offrant une amélioration des performances allant jusqu'à 2 fois et un rapport qualité/prix 19 fois meilleur par rapport aux superordinateurs typiques.
SEMIFIVE se spécialise dans les plateformes SoC et les solutions de conception ASIC, en se concentrant sur les plateformes de conception SoC pour experts en puces AI. La collaboration vise à fournir des caractéristiques LLM rentables et écoénergétiques, réduisant potentiellement les frais d'exploitation des centres de données et s'étendant aux secteurs nécessitant des LLM.
SEMIFIVE, ein führender Anbieter von Designlösungen, hat einen Vertrag mit HyperAccel angekündigt, um einen generativen KI-Chip namens Bertha zu entwickeln und in Serie zu produzieren. Der Chip wird mit 4nm-Prozesstechnologie entwickelt, wobei die Massenproduktion für Q1 2026 geplant ist. Die LLM Verarbeitungseinheit (LPU) von HyperAccel ist speziell für transformerbasierte Large Language Models ausgelegt und bietet bis zu 2-mal bessere Leistung sowie ein 19-mal besseres Preis-/Leistungsverhältnis im Vergleich zu typischen Supercomputern.
SEMIFIVE ist auf SoC-Plattformen und ASIC-Designlösungen spezialisiert und konzentriert sich auf SoC-Designplattformen für KI-Chipexperten. Die Zusammenarbeit zielt darauf ab, kostenwirksame und energieeffiziente LLM-Funktionen bereitzustellen, wodurch die Betriebskosten von Rechenzentren potenziell gesenkt und in Branchen, die LLMs benötigen, expandiert werden kann.
- Contract secured for mass production of generative AI chip Bertha
- Bertha offers up to 2x performance improvement over existing GPUs
- Price/performance ratio 19 times better than typical supercomputers
- Potential to reduce data center operational expenses
- Expansion into industries requiring LLMs
- Mass production not expected until Q1 2026
Collaboration on design and development using 4nm process technology, with mass production targeted for Q1 2026
HyperAccel has developed a LLM Processing Unit (LPU), an AI chip specifically designed for transformer-based Large Language Model (LLM). This is the world's first semiconductor LPU tailored for LLM inference, offering low cost, low latency, and domain-specific features. It is expected to replace existing high-cost and low-efficiency Graphics Processing Units (GPUs) with performance improvements of up to 2 times and a price/performance ratio that is 19 times better than that of a typical supercomputer.
SEMIFIVE specializes in SoC platforms and ASIC design solutions. The company focuses on developing expert SoC design platforms for AI chips these days with plans to expand its roadmap to respond to customer demand for AI custom silicon. To achieve this goal, SEMIFIVE is actively developing its SoC chiplet platform in collaboration with industry-leading partners.
"We are delighted to work with SEMIFIVE, a leading provider of SoC platforms and comprehensive ASIC design solutions, for the development of Bertha to be mass-produced," said Joo-Young Kim, CEO of HyperAccel. "By collaborating with SEMIFIVE , we are excited to offer customers AI semiconductors that provide more cost-effective and power-efficient LLM features than GPU platforms. This advancement will significantly reduce the operational expenses of data centers and expand our business scope to other industries that require LLMs."
"HyperAccel is a company with the most efficient and scalable LPU technology for LLMs. As the demand for LLM computation is skyrocketing, HyperAccel has the potential to become a new powerhouse in the global processor infrastructure," said Brandon Cho, CEO and co-founder of SEMIFIVE. "We are extremely excited to be the production partner for HyperAccel's groundbreaking Bertha chip, and we are thrilled to contribute to yet another innovative success story built on the SEMIFIVE platform."
About SEMIFIVE
SEMIFIVE is the pioneer of platform based SoC design, working with customers to implement innovative ideas into custom silicon in the most efficient way. Our SoC platforms offer a powerful springboard for new chip designs and leverage configurable domain-specific architectures and pre-validated key IP pools. We offer comprehensive spec-to-system capabilities with end-to-end solutions so that custom SoCs can be realized faster, with reduced cost and risks for key applications such as data center or AI-enabled IoT. With a strong partnership with Samsung Foundry as a leading SAFETM DSP partner, as well as the larger ecosystem, SEMIFIVE provides a one-stop shop solution for any SoC design needs. For more information, please visit www.semifive.com.
View original content to download multimedia:https://www.prnewswire.com/news-releases/semifive-concluded-mass-production-contract-for-ai-chip-with-hyperaccel-302272655.html
SOURCE SEMIFIVE
FAQ
What is the name of the AI chip being developed by SEMIFIVE and HyperAccel?
When is the mass production of Bertha targeted to begin?
What process technology will be used to develop Bertha?
How does HyperAccel's LPU compare to existing GPUs in terms of performance?
What is the price/performance ratio of HyperAccel's LPU compared to a typical supercomputer?
