Key Takeaways
- Pliops accelerates GenAI infrastructure performance, enabling data access up to 50 times faster with reduced computational load.
- Pliops XDP LightningAI enhances GPU performance and efficiency, significantly improving resource utilization and reducing carbon footprint.
- Pliops offers a comprehensive toolbox for GenAI applications, facilitating easy adoption of their silicon, hardware, and software solutions.
Prior to becoming CEO of Pliops in 2023, Ido Bukspan was the senior vice president of the Chip Design Group at NVIDIA and one of the leaders at Mellanox before it was acquired by NVIDIA for nearly $7 billion.
Tell us about your company.
Pliops accelerates and amplifies the performance and scalability of global GenAI infrastructure, driving unparalleled efficiency and innovation.
Pliops was founded in 2017 by storage industry veterans from Samsung, M-Systems, and XtremIO. Pliops is pioneering a new category of product that enables cloud and enterprise data centers to access data up to 50 times faster with one-tenth of the computational load and power consumption. Its technology consolidates multiple inefficient layers into one ultra-fast device based on a groundbreaking approach. Pliops’ solution addresses the scalability challenges posed by the cloud data explosion and the increasing data demands of AI and ML applications.
What problems are you solving?
Pliops XDP LightningAI, our revolutionary Accelerated Key-Value distributed smart node, introduces a new tier of memory that surpasses HBM for GPU compute applications. Our product can double end-to-end performance and enhance efficiency for vLLM, a leading inferencing solution. By leveraging Pliops’ state-of-the-art technology, we deliver advanced GenAI and AI solutions, significantly improving GPU utilization, reducing total cost of ownership, and cutting power consumption and carbon emissions.
What application areas are your strongest?
Our strongest application areas focus on accelerating GenAI applications, including LLM inferencing, DLRM, RAG/VectorDB, and SQL and Document DB acceleration. Our users benefit from over 3X better utilization of compute resources, more than 50% savings in capital expenditures, and a 50% reduction in carbon footprint.
What keeps your customers up at night?
Our customers are deeply concerned about power consumption in data centers, especially as AI infrastructure and emerging AI applications significantly increase power footprints and strain cooling budgets. They also worry about maintaining margins as they expand their AI infrastructure, adding GPU tiers of compute. The growing power and cooling demands, combined with substantial capital expenditures on GPUs, are consuming margins and keeping our customers up at night.
What does the competitive landscape look like and how do you differentiate?
The competitive landscape in GenAI and applications requiring new infrastructure, including additional CapEx, cooling budgets, and increased power, is still emerging as we stand at the dawn of the AI revolution. Innovation is essential in this space, and we embrace all solutions, not just our own. Specifically, our focus is on enabling GPUs to have IO access to large amounts of data for local consumption by applications. There are various approaches to solving this, and our strategy is among them.
What new features/technology are you working on?
We are continually advancing our technology to meet the evolving demands of AI applications. Our latest developments include enhancements to our XDP LightningAI, which now provides unprecedented levels of memory and compute efficiency for GenAI workloads. We’re focusing on solutions that allow GPUs to access vast amounts of data for local processing, significantly improving performance and reducing energy consumption. Additionally, we are working on innovative methods to further decrease the power footprint and cooling requirements of data centers, ensuring our solutions remain at the forefront of sustainable AI infrastructure.
How do customers normally engage with your company?
We provide a comprehensive toolbox for critical GenAI applications, making it exceptionally easy for our customers to adopt our products and technology. Our solutions encompass silicon, hardware, and software components, and we offer reference designs that ISVs, CSPs, and operators can utilize with minimal effort to realize significant benefits. Our products are available as cloud-ready solutions, and we collaborate with compute/storage OEMs and ISVs to support HPC and on-prem data center operators.
Talk to a Pliops product expert.
Also Read:
CEO Interview with Roger Cummings of PEAK:AIO
CEO Interview with Richard Hegberg of Caspia Technologies
Share this post via:
Intel’s Path to Technological Leadership: Transforming Foundry Services and Embracing AI