{"slug": "completing-the-gpu-performance-picture-understanding-taf-alongside-peak-flops", "title": "Completing the GPU Performance Picture: Understanding TAF Alongside Peak FLOPs and MAF", "summary": "AMD introduced Typical Attained FLOPs (TAF), a third GPU performance metric that measures FLOP-weighted sustained throughput across a fixed representative GEMM set selected before testing, in the third blog post of its series on AMD Instinct GPUs. TAF joins Peak FLOPs, which describes theoretical compute capability, and Max-Achievable FLOPs (MAF), which measures the highest sustained throughput after testing a range of GEMM sizes. AMD states the goal is not to replace Peak FLOPs or MAF but to answer a different question alongside them for a more complete view of GPU compute performance.", "body_md": "Modern AI accelerators can execute an extraordinary number of floating-point operations every second. But a single FLOPs number rarely tells the full performance story. In the first two blogs in this series, we discussed Peak FLOPs and Max-Achievable FLOPs (MAF) for AMD Instinct™ GPUs., we discussed Peak FLOPs and Max-Achievable FLOPs (MAF) for AMD Instinct™ GPUs. Peak FLOPs describe the theoretical compute capability of the hardware. MAF goes a step further by measuring the highest sustained throughput the GPU can achieve after testing a range of GEMM sizes. This blog introduces a third metric: Typical Attained FLOPs (TAF). TAF measures FLOP-weighted sustained throughput across a fixed representative GEMM set selected before testing. The goal is not to replace Peak FLOPs or MAF. Each metric answers a different question, and together they provide a more complete view of GPU compute performance.", "url": "https://wpnews.pro/news/completing-the-gpu-performance-picture-understanding-taf-alongside-peak-flops", "canonical_source": "https://rocm.blogs.amd.com/artificial-intelligence/taf-blog/README.html", "published_at": "2026-10-06 00:00:00+00:00", "updated_at": "2026-10-06 17:17:16.748328+00:00", "lang": "en", "topics": ["ai-chips", "ai-infrastructure", "machine-learning"], "entities": ["AMD", "AMD Instinct", "Typical Attained FLOPs", "Peak FLOPs", "Max-Achievable FLOPs", "GEMM"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/completing-the-gpu-performance-picture-understanding-taf-alongside-peak-flops", "markdown": "https://wpnews.pro/news/completing-the-gpu-performance-picture-understanding-taf-alongside-peak-flops.md", "text": "https://wpnews.pro/news/completing-the-gpu-performance-picture-understanding-taf-alongside-peak-flops.txt", "jsonld": "https://wpnews.pro/news/completing-the-gpu-performance-picture-understanding-taf-alongside-peak-flops.jsonld"}}