Dnotitia Begins Testing VDPU ASIC Samples
Dnotitia Inc. announced in a press release that its first Vector Data Processing Unit ASIC samples have returned from fabrication and are undergoing chip characterization. The company also presented a server architecture using VDPU cards for vector retrieval in retrieval augmented generation and agentic AI at AI Infra Summit 2026.
A server with four VDPU cards delivered up to 5.77 times the vector search throughput of a dual socket CPU only server on Dnotitia's FPGA evaluation platform. In a 4,096 dimensional multimodal workload, it reduced host CPU use during index building by 92% and host memory use by 73%. Dnotitia said these results do not represent final ASIC performance.
The FPGA platform supports FAISS, Milvus and hnswlib across several index types. Dnotitia plans to begin evaluations using VDPU ASICs in the fourth quarter of 2026 and is targeting up to 10 times the vector search performance of a CPU based server.
We hope you enjoyed this article
Consider subscribing to one of our newsletters like Silicon Brief or Daily AI Brief.
Also, consider following us on social media:
More from Data Centers
Nov 14 Telus pulls out of two Vancouver AI data centre projects with Westbank Nov 14 Petition to reinstate Jensen Huang as Nvidia CEO passes 115,000 signatures, with Elon Musk among signers Nov 14 Asian governments respond to Jensen Huang's exit from Nvidia Nov 14 Malaysia plans law making data centres disclose AI chip imports and power use to curb smuggling Nov 14 More than 13,000 object to Chapelcross plan for world's largest AI data centre campusSilicon Brief
Weekly coverage of AI hardware developments including chips, GPUs, cloud platforms, and data center technology.
Industry analysis
2025 Global Business Services Agenda: Gen AI Takes Center Stage
This industry analysis by The Hackett Group explores the transformative impact of generative artificial intelligence (Gen AI) on global business services (GBS) in 2025. The study highlights the shift from exploration to acceleration of Gen AI initiatives, with 89% of executives advancing these projects to improve customer satisfaction, innovate products, and reduce costs. The report also discusses the challenges and strategies for successful Gen AI adoption, emphasizing the need for a technology-enabled operating model and the importance of reskilling the workforce.
Read moreYou may also like
Bentaus Cuts GPU Power 30% With DriveNets AI Fabric on AMD Cluster
Innodisk adds data-centre SSDs and server memory for AI workloads
AMD Makes 6th Gen EPYC Zen 6 Server CPUs Generally Available, Targets Agentic AI
Q.ANT and IONOS open a pilot for running AI models on photonic processors in a German cloud
AEWIN to show liquid-cooled AMD AI servers at SC26
Daily AI Brief: the AI news that matters, in your inbox.