AMD Newsroom

Latest News

View all

Artificial Intelligence

View all

AMD on LinkedIn

LinkedIn · AMD

Follow
  • Leading performance built to scale. MLPerf Inference 6.1 shows how AMD Instinct MI350 Series GPUs delivered leadership performance, production-scale inference and continued AMD ROCm software gains. • New MLPerf record with 5.75M tokens/sec at 512-GPU scale • GPT-OSS-120B leadership at 8 and 72 GPUs • Leading performance from the new AMD Instinct MI350P PCIe card • Partner results within 4% of AMD on average Dive into the full results of our broadest MLPerf Inference submission yet: https://bit.ly/4hdWnDr

  • At AMD, we're committed to helping startups: ✅Build and optimize demanding AI workloads ✅Connect with technical, cloud, and go-to-market ecosystems ✅Scale from prototype to production We believe great AI companies need more than compute. They need a platform for growth. Join us on the journey.

  • AMD is proud to be named one of the 100 Best Corporate Citizens of 2026, ranking 25th overall (up 14 positions from 2025). This recognition reflects the work of teams across AMD continuing to build responsible practices into how we operate, as well as our ongoing collaboration with customers and partners. Learn more about our approach and progress: https://bit.ly/3SLxd6Z

  • Better inference starts with the right architecture. At #AdvancingAI, Cerebras Founder and CEO Andrew Feldman shared how AMD and Cerebras are bringing together complementary strengths to optimize the two key stages of inference: pre-fill and decode. Together, we are delivering the higher throughput, faster token generation and greater efficiency needed to build what's next. Learn how this collaboration is advancing the future of AI inference.

  • In the first race at the Madring circuit, Kimi Antonelli set the standard. Congratulations to the entire Mercedes-AMG PETRONAS Formula One Team on a historic win and an incredible performance.

See more on LinkedIn

AMD on X

X · @AMD

Follow
  • Inference is moving fast, and AMD is providing the infrastructure needed to deliver at scale. Setting a new MLPerf record with 5.75M tokens/sec at 512-GPU scale, our broadest submission yet provides another proof point for performance, scale, software maturity and reproducibility. Read more about the results: https://t.co/8UUw33Z88b

  • AMD is honored to be recognized as one of the 100 Best Corporate Citizens of 2026, ranking 25th overall. Learn how this recognition reflects our continued progress in corporate responsibility: https://t.co/1gF9ZwmHjM

  • Better inference starts with the right architecture. @AndrewDFeldman, Founder and CEO at @Cerebras, highlights how we are combining pre-fill and decode to deliver high throughput, fast tokens and greater efficiency.

  • In the first race at the Madring circuit, Kimi Antonelli set the standard. Congratulations to the entire @MercedesAMGF1 team on a historic win and an incredible performance.

See more on X

Stay in the loop

Connect with AMD

Follow AMD for the latest news, product announcements and community updates.