For the recent NFD40 event, I had the idea of including and updating other delegates’ article postings on each vendor I posted about. I’ve gone back through and cross-checked against the Tech Field Da...

For the recent NFD40 event, I had the idea of including and updating other delegates’ article postings on each vendor I posted about. I’ve gone back through and cross-checked against the Tech Field Da...

For the recent NFD40 event, I had the idea of including and updating other delegates’ article postings on each vendor I posted about. I’ve gone back through and cross-checked against the Tech Field Da...

Upscale AI argues that traditional cloud and front-end networks, which are largely based on a client-server architecture, are fundamentally ill-suited for the unique demands of AI workloads. While…

Upscale AI argues that traditional cloud and front-end networks, which are largely based on a client-server architecture, are fundamentally ill-suited for the unique demands of AI workloads. While…

Upscale AI distinguishs between two critical domains: scale-up networking, which creates a large compute environment within a rack where multiple GPUs see a flat, unified memory, and scale-out…

Upscale AI is rethinking and optimizing AI networking with their SkyHammer architecture, due for 2026 release. Their solution architecture encompasses from ASIC chip up through the system, and rack la...

For the recent NFD40 event, I had the idea of including and updating other delegates’ article postings on each vendor I posted about. I’ve gone back through and cross-checked against the Tech Field Da...

최근 업계에서 AI 인프라 스타트업인 Upscale AI가 20억 달러 규모의 기업 가치를 인정받으며 대규모 투자 유치를 논의 중이라는 소식이 전해졌습니다. 창업 7개월 만에 제품 출시도 없이 이러한 평가를 받는 것은 현재 기술 시장이 무엇을 열망하고 있는지 보여주는 단면입니다.

Rohit Mittal Head of AI Products and Technologies - UpscaleAI, Chris Petersen Fellow of Technology and Solution Architecture - Astera Labs, Josh Collier Engineer - Intel, Priyank Shukla Director of Product Management - Synopsys   As AI compute demands continue to rise, Cloud Service Providers, System OEMs, and IP-Silicon Providers require a high-performance, efficient solution to meet the challenges of scale-up AI infrastructure. The UALink Specification delivers a low-latency, high-bandwidth interconnect designed to enable fast, efficient communication between accelerators and switches within AI computing pods, boosting performance while improving power and cost efficiency.  In this panel, UALink experts will explore real-world applications and use cases for the technology, highlighting how UALink supports the development of next-generation AI-ML scale-up systems. The discussion will also cover the importance of supply chain diversity and system interoperability enabled by UALink. Attendees will gain insight into the specification roadmap and upcoming enhancements to UALink technology.