{"slug": "amd-launches-helios-and-expands-ai-portfolio", "title": "AMD Launches Helios and Expands AI Portfolio", "summary": "AMD launched Helios, a rack-scale AI system combining 72 Instinct MI455X GPUs with 18 sixth-generation EPYC Venice CPUs, at its Advancing AI event on July 23. AMD says the system is in production, while Reuters reports shipments are expected near the end of the third quarter, and OpenAI expects to bring Helios systems online beginning in the fourth quarter. The launch expands AMD's AI portfolio with new EPYC CPUs, Instinct MI400 GPUs, Ryzen AI Embedded processors, and a Kria robotics platform, positioning it against Nvidia in integrated AI infrastructure.", "body_md": "# AMD Launches Helios and Expands AI Portfolio\n\nAMD launched Helios, a rack-scale AI system combining 72 Instinct MI455X GPUs with 18 sixth-generation EPYC “Venice” CPUs, at its Advancing AI event on July 23. AMD says the system is in production, while Reuters reports shipments are expected near the end of the third quarter. OpenAI expects to bring Helios systems online beginning in the fourth quarter.\n\nAMD launched **Helios**, its first rack-scale AI system, at Advancing AI 2026 in San Francisco on July 23. The company also introduced sixth-generation EPYC CPUs, the Instinct MI400 GPU series, Ryzen AI Embedded X100 processors, and a Kria robotics development platform.\n\nAMD says Helios is in production. Reuters, in reporting republished by Investing.com, said shipments are expected near the end of the third quarter. AMD's announcement separately says OpenAI expects to bring Helios systems online beginning in the fourth quarter of 2026, with deployments accelerating through 2027. Those dates describe different stages: general shipment timing and one customer's planned deployment.\n\n### What is inside Helios\n\nEach rack combines **72 Instinct MI455X GPUs** with **18 sixth-generation EPYC “Venice” CPUs**, Pensando networking, and AMD's ROCm software stack. The design moves AMD's competition with Nvidia beyond individual accelerators toward an integrated rack in which processors, memory, networking, and software are engineered together.\n\nAMD estimates that Helios can deliver up to 30% more inference tokens per dollar than a leading competing system. The company says that comparison uses a projected Kimi K2 Thinking workload and projected hourly GPU pricing. It is therefore a vendor estimate, not an independently reproduced benchmark. Data Center Dynamics also reported AMD's peak-performance and memory-capacity comparisons with Nvidia's Vera Rubin platform; those claims likewise need independent workload-level validation.\n\n### Deployment evidence matters next\n\nThe launch establishes the platform's architecture and planned availability, but it does not yet show production behavior across different models and serving stacks. For teams evaluating Helios, the most useful next evidence will include:\n\n- •throughput and latency on reproducible training and inference workloads;\n- •rack-level power, cooling, and networking requirements;\n- •ROCm compatibility across PyTorch, vLLM, SGLang, and custom kernels;\n- •deployment availability from the announced OEM and infrastructure partners.\n\n### LDS assessment\n\nHelios is significant because it packages AMD GPUs, CPUs, networking, and software as one deployable system. That can simplify procurement and tuning for large installations, but the business case will depend on measured performance, software maturity, power efficiency, and delivery at scale. OpenAI's planned deployment provides an important production signal; independently reported results will determine how broadly that signal generalizes.\n\n## Key Points\n\n- 1Helios combines 72 Instinct MI455X GPUs, 18 sixth-generation EPYC CPUs, Pensando networking, and ROCm in one rack-scale system.\n- 2AMD says Helios is in production; Reuters reports shipments near the end of the third quarter, and OpenAI expects to bring systems online beginning in the fourth quarter.\n- 3AMD's performance comparisons are vendor estimates, so reproducible workload benchmarks, software compatibility, and deployment evidence remain the critical next tests.\n\n## Scoring Rationale\n\nHelios is a major rack-scale infrastructure launch from Nvidia's principal accelerator competitor, with near-term shipment reporting and a planned OpenAI deployment. Its importance is the integrated GPU, CPU, networking and ROCm platform; vendor performance claims still require independent validation.\n\n## Sources\n\nPrimary source and supporting public references used for this report.\n\n## View 3 more sources\n\n[AMD says its newest AI server is in full production, will ship in monthsinvesting.com](https://www.investing.com/news/stock-market-news/amd-expected-to-launch-next-generation-of-ai-infrastructure-to-challenge-nvidia-4808203)[AMD officially launches Helios rackscale system, with MI455X GPUs, Venice Epyc CPUs, and Pensandodatacenterdynamics.com](https://www.datacenterdynamics.com/en/news/amd-officially-launches-helios-rackscale-system-with-mi455x-gpus-venice-epyc-cpus-and-pensando/)[AMD sees compute mkt zooming to $2 trillion by 2030, launches new productsthehindubusinessline.com](https://www.thehindubusinessline.com/info-tech/computers-and-laptops/amd-sees-compute-mkt-zooming-to-2-trillion-by-2030-launches-new-products/article71261175.ece)\n\nPractice interview problems based on real data\n\n1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.\n\n[Try 250 free problems](/problems)", "url": "https://wpnews.pro/news/amd-launches-helios-and-expands-ai-portfolio", "canonical_source": "https://letsdatascience.com/news/amd-launches-helios-and-expands-ai-portfolio-d7dab2d0", "published_at": "2026-07-24 08:10:42+00:00", "updated_at": "2026-07-24 16:28:43.556178+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-infrastructure", "ai-chips", "ai-products", "ai-startups"], "entities": ["AMD", "Helios", "Instinct MI455X", "EPYC Venice", "OpenAI", "Reuters", "Nvidia", "ROCm"], "alternates": {"html": "https://wpnews.pro/news/amd-launches-helios-and-expands-ai-portfolio", "markdown": "https://wpnews.pro/news/amd-launches-helios-and-expands-ai-portfolio.md", "text": "https://wpnews.pro/news/amd-launches-helios-and-expands-ai-portfolio.txt", "jsonld": "https://wpnews.pro/news/amd-launches-helios-and-expands-ai-portfolio.jsonld"}}