𝗔𝗠𝗗 𝗮𝗻𝗱 𝗜𝗻𝘁𝗲𝗹 𝗨𝗻𝗶𝘁𝗲 𝗳𝗼𝗿 𝟭𝟲𝘅 𝗔𝗜 𝗕𝗼𝗼𝘀𝘁

𝗔𝗠𝗗 𝗮𝗻𝗱 𝗜𝗻𝘁𝗲𝗹 𝗨𝗻𝗶𝘁𝗲 𝗳𝗼𝗿 𝟭𝟲𝘅 𝗔𝗜 𝗕𝗼𝗼𝘀𝘁

AMD and Intel are working together on a new standard called ACE.

These two rivals joined the x86 Ecosystem Advisory Group to fix a major problem. For years, x86 chips lacked a unified language for AI. This made it hard for developers to write code that works on both Intel and AMD chips.

ACE changes this. It introduces AI Compute Extensions to speed up math operations.

Here is why this matters for you:

• 16x Faster AI: ACE provides 16x the compute density of current instructions.
• One Codebase: You can run the same AI code on any x86 CPU. You do not need to rewrite code for different brands.
• Better Efficiency: ACE uses 2D Tile Registers to handle matrix math in one cycle.
• Lower Power: CPUs can now handle AI tasks that usually require a GPU. This helps in laptops and edge devices.

The goal is not to replace GPUs. Instead, ACE helps CPUs handle AI tasks when a GPU is too big, too expensive, or uses too much power. This is perfect for thin laptops and real-time tasks.

Software support is already starting. Major libraries like NumPy, PyTorch, and TensorFlow are working to add ACE support.

You will not see this hardware today. It usually takes two to three years for new instructions to reach products. Expect to see this in future chip architectures.

This partnership helps x86 stay competitive against ARM. By working together, Intel and AMD ensure their platforms remain relevant in the AI era.

Source: https://dev.to/techmag/amd-and-intel-join-forces-on-ace-16x-ai-boost-for-x86-cpus-32ll

Optional learning community: https://t.me/GyaanSetuAi