Booth: J02
회사 정보
HyperAccel is a Korean AI semiconductor company specializing in LPU (LLM Processing Unit) technology, purpose-built for large language model inference. Founded by KAIST researchers, we design domain-specific processors that deliver faster token generation with significantly lower power consumption and cost than GPUs. Our product lineup ranges from data-center inference servers to edge and workstation solutions, enabling enterprises to deploy LLM services on-premises with full data security. HyperAccel works with global cloud providers, enterprises, and public institutions to make generative AI infrastructure affordable and energy-efficient. We are seeking partners and customers across Southeast Asia’s fast-growing AI market.
KO 
EN 
VI 
CN 
JP