AMD Adds UCIe Support to Select Versal Adaptive SoCs
AMD announced that select Versal RF Series adaptive SoCs will support Universal Chiplet Interconnect Express 1.1 interfaces. The support is designed to let Versal devices communicate directly with chiplets in the same package for workloads such as RF, AI, CPU, GPU, security, communications, and ASIC functions.
Select Versal RF Series devices will support up to four UCIe-SP interfaces and up to two UCIe-AP interfaces, with aggregate bandwidth at multiple terabits per second inside the package. The devices combine high resolution RF data converters, dedicated DSP IP, AI Engines, and programmable logic, and provide up to 80 TOPS of DSP compute.
The UCIe connections will support use with RF analog front ends, AI acceleration chiplets, CPUs, GPUs, custom security engines, communications processors, and application specific ASICs. Production chiplets are expected to be available with select Versal RF Series devices in the fourth quarter of 2027.
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
Oct 16 Hitachi Energy and Ansaldo Energia Partner on Data Center Power Systems Oct 16 Clivet Unveils Magnetic Bearing Compressor for AI Data Center Cooling Oct 15 Rising rates strain AI data center fundraising Oct 15 Ukrainian Drones Hit Sberbank Data Center Planned for Supercomputer Oct 15 Brightray Signs Letter of Intent for 10 MW Indiana AI Data CenterSilicon Brief
Weekly coverage of AI hardware developments including chips, GPUs, cloud platforms, and data center technology.
Whitepaper
Tensordyne Napier: What If One Rack Could Do the Work of Nine?
Tensordyne
This Tensordyne whitepaper presents Napier, an inference-focused AI processor and rack-scale system based on the company’s TDN Math logarithmic number system. It examines infrastructure requirements for large mixture-of-experts and agentic models, compares major inference architecture approaches, and details the TDN AIP processor, TDN72 pod, TDN Link fabric, and Napier Ultra configuration. The paper reports simulation-based performance, cost, and accuracy-validation results, including Tensordyne’s projected comparison of one Napier rack with a nine-rack Nvidia Rubin plus Groq deployment; the chip is reported as taped out and in fabrication.
Read moreYou may also like
Advantech Expands WEDA Edge AI Platform With Intel, Qualcomm and AMD
Xsight Labs Introduces 25 Watt E1L DPU for Edge Systems
AEWIN Introduces Liquid Cooled AMD EPYC 9006 AI Servers
Vuzix to Show Causeway Optical Interconnect Platform for AI Data Centers
Synopsys and TSMC Expand AI Chip Design Partnership
Daily AI Brief: the AI news that matters, in your inbox.