Intel and AMD Join Forces to Launch ACE Extensions for x86 with a Focus on AI
The new instruction set promises more power-efficient and denser matrix multiplication for the x86 architecture.
Intel and AMD have jointly announced the development of new CPU extensions called ACE (Advanced Computational Extensions), designed for the x86 architecture. The collaboration aims to introduce an instruction set optimized for artificial intelligence directly into traditional processors. The initiative seeks to improve machine learning performance without relying exclusively on dedicated accelerators.
The main highlight of the new technology is a redesign of how the x86 architecture handles matrix multiplication. This mathematical operation is fundamental to the functioning of AI models. With the ACE extensions, executing this process becomes significantly more efficient in both power consumption and transistor density.
Historically, the execution of heavy AI workloads has been dominated by GPUs and specialized NPUs. The introduction of native AI instructions into x86 processors represents a strategic shift for the market. This allows smaller inference and processing tasks to be performed faster and more cost-effectively directly on the CPU.
The collaboration between the two semiconductor industry giants is an important step toward standardizing the x86 ecosystem. By developing the ACE extensions together, Intel and AMD make it easier for software developers, who will be able to optimize their AI applications for a unified instruction set, benefiting a wide range of devices in the technology market.
What are the ACE extensions developed by Intel and AMD?
ACE (Advanced Computational Extensions) is a new instruction set for the x86 architecture jointly developed by Intel and AMD. It is designed to optimize artificial intelligence performance directly on traditional CPUs by improving matrix multiplication efficiency.
How do the ACE extensions improve AI performance on x86 processors?
The ACE extensions improve AI performance by redesigning how the x86 architecture handles matrix multiplication, a fundamental operation for AI models. This makes the process significantly more efficient in terms of both power consumption and transistor density.
Why did Intel and AMD collaborate on the ACE instruction set?
Intel and AMD collaborated to standardize the x86 ecosystem. By creating a unified instruction set, they make it easier for software developers to optimize AI applications across a wide range of devices, allowing smaller inference tasks to run faster and more cost-effectively on the CPU.