ASRock Rack expands AI infrastructure for agentic workloads - Engineering.com
Its new portfolio spans liquid-cooled AI factory systems, inference servers and edge platforms for robotics and medical use. ASRock Rack Inc. announced its latest portfolio of AI-native infrastructure designed for the era of agentic AI. Showcasing at COMPUTEX 2026 (Booth No.: R0514), ASRock Rack is introducing the new 2UXGM-VERA2 system powered by NVIDIA Vera CPU. The company is also presenting its next-generation AI platform portfolio for AI factories, inference cloud deployments, and edge AI applications, delivering a unified vision for the next wave of AI computing at every scale. Accelerating the Era of Agentic AI The AI industry is undergoing a fundamental transformation as scaling laws move beyond model size into the era of agentic AI. This shift from “human talking to AI” to “AI talking to AI” has placed CPU execution on the critical path. To enable this next wave, ASRock Rack is introducing the 2UXGM-VERA2, powered by NVIDIA Vera CPU, a processor purpose-built for the age of AI. Combining custom-designed NVIDIA Olympus cores with high-bandwidth LPDDR5X memory and NVIDIA Scalable Coherency Fabric, Vera delivers a unified compute architecture that completes complex agentic and reinforcement learning workloads 50% faster than traditional CPU infrastructure. Driving the AI computing wave at every scale At COMPUTEX 2026, ASRock Rack is showcasing AI server platforms integrated with advanced liquid-cooling solutions from ecosystem partners. The company’s exhibit highlights the NVIDIA Vera Rubin NVL72, cooled with a double-rack-width liquid-to-air coolant distribution unit (CDU) for next-generation AI factory environments. Also on display are liquid-cooled systems incorporating NVIDIA HGX Rubin NVL8, including the fully liquid-cooled 2U16X-GNR2/DLC and the 5U16X-GNR2/DLC, which combines liquid cooling for CPU and GPU with sufficient airflow for flexible integration of other key peripherals. The company is also demonstrating rack-scale liquid-cooling solutions with in-row CDU for high-density AI deployments. In addition to its rack-scale solutions, ASRock Rack will demonstrate a diverse lineup of MGX-based servers for enterprise and edge environments. This includes the 6UXGM-GNR2/DLC, supporting up to eight liquid-cooled NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs to accelerate enterprise AI and visual computing. Also featured is the 4UXGM-GNR2 CX8, a NVIDIA RTX PRO Server optimized for inference cloud environments. Bridging to the industrial edge, the 2UXGI-Thor, built on NVIDIA IGX Thor platform, showcases capabilities for Physical AI, enabling real-time sensor processing and functional safety for next-generation autonomous robotics and medical applications. For more information, visit asrockrack.com.