Liqid names former NVIDIA executive Kevin Connors to its board of directors

Liqid said on Tuesday that it has appointed Kevin Connors, a former NVIDIA executive, to its Board of Directors. The announcement, published Oct. 13, 2026, frames the move as a way to add board-level experience from the AI and accelerated computing industry, an area where data center buildouts increasingly depend on how efficiently systems move and manage data for GPUs.

The company, headquartered in Westminster, Colorado, describes itself as a provider of “software defined memory” and “GPU pooling infrastructure.” Software defined memory generally refers to approaches that manage memory resources through software rather than treating memory as a fixed, hardware-only component. GPU pooling typically means aggregating or coordinating GPU resources so that workloads can be scheduled and balanced across a shared pool rather than being tied to a single server or node.

In the statement, Liqid did not spell out Connors’ specific prior roles at NVIDIA or outline what committee responsibilities he will assume. It also did not provide details about whether the appointment is expected to coincide with new product milestones, customer wins, or strategic partnerships.

The timing of the board appointment comes as market attention remains focused on the “memory bottleneck” problem in AI training and inference. In broad terms, faster AI accelerators can be limited by how quickly data is supplied to the computing units and how efficiently intermediate results are staged in memory. Companies like Liqid position their technologies around that linkage, arguing that memory management and resource coordination can materially affect performance and scalability.

For NVIDIA, the relationship is notable given the company’s central role in GPU hardware and its broader data center platform ecosystem. While Liqid’s announcement does not claim any formal partnership with NVIDIA, the appointment highlights how AI infrastructure firms often seek leaders who understand GPU deployment realities, including the operational constraints data center operators face when scaling compute clusters.

Liqid’s description of “GPU pooling infrastructure” suggests it is aiming at the infrastructure layer that sits between raw accelerators and higher-level workloads. In such setups, the practical challenge is not only buying GPUs, but also orchestrating them to meet application demand patterns, latency requirements, and throughput goals. A director with prior experience inside the GPU supply chain and ecosystem may be viewed as relevant to that operational focus.

The company’s announcement also did not disclose any financial terms tied to Connors’ appointment, such as equity compensation or the expected term length. It did not specify whether Connors is an independent director, nor did it outline any immediate governance changes beyond adding him to the board.

Looking ahead, investors and customers will likely look for further information about how Liqid is translating its memory and pooling approach into deployments, partnerships, and measurable performance outcomes. Next steps that could clarify the impact of the appointment include additional detail on board committees, product roadmap updates, and whether the company’s go-to-market efforts extend into larger data center procurement cycles.