| 会社名 | Group | 親会社 | 子会社 | 得意先 | 仕入先 | 業務提携 | 銀行等 | 株主等 |
|---|---|---|---|---|---|---|---|---|
| Pensando Systems Inc. | ||||||||
| Silo AI Oy | ||||||||
| Xilinx, Inc. |
Sixth-generation AMD EPYC processors will power Oracle Cloud Infrastructure (OCI) Compute E7 instances. AMD Pensando technology will underpin Oracle Acceleron converged NICs for AMD compute and accelerator instances on OCI.
AMD says it provides innovation, high performance, and scalability across its entire product line of adaptive SoCs and FPGAs.
Advanced Micro Devices (AMD) and Meta have announced a strategic partnership to power Meta’s superintelligence infrastructure. The collaboration will leverage AMD’s data center GPUs and AI solutions to accelerate Meta’s AI workloads. The partnership aims to enhance Meta’s AI capabilities and support the development of advanced AI models.
AMD and Intel are partnering to merge their x86 CPU optimizations for PyTorch in oneDNN, creating a unified software path for developers across both CPUs.
AMD’s latest FSR Upscaling 4.1, a machine‑learning based upscaling technology, now supports AMD Radeon RX 6000 Series GPUs, Ryzen AI 9 and Ryzen AI Max CPUs, select Ryzen Z2 processors, and Xbox consoles, delivering higher frame rates and improved image quality.
The article argues that as AI moves from answering questions to taking action in the physical world, performance must be defined by the capabilities of the entire system rather than peak compute alone.
AMD CTO Mark Papermaster joins EE Times journalist Nitin Dahad at World Summit AI 2026 for an interview on the future of AI and semiconductor innovation.
AMD's CEO and Chairman Lisa Su released a replay of the OCP 2026 keynote, where she introduced a suite of open full‑stack computing products that support agent‑based AI at all scales.
In the keynote address at OCP 2026, AMD CEO Lisa Su explained that the company's broad range of AI products will support the construction of next‑generation AI infrastructure. She also introduced the sixth‑generation EPYC "Verano", which will be the first to adopt an LPDDR memory interface, and said it will be available in 2027.
AMD Helion is an open blueprint that provides flexibility for agentic AI infrastructure. It supports CPU‑led racks for agent execution at hyperscale and all‑in‑one configurations that combine CPU execution and GPU inference for enterprise deployments.