Microsoft is introducing two of its own chips, the Maya 100 and the Cobalt 100. AI accelerators and CPUs designed for Azure cloud infrastructure and scheduled to launch in 2024 will make AI technology more efficient and powerful.
Microsoft recently announced two of its own chips: the Maya 100 AI accelerator and the Cobalt 100 CPU. These chips are special Azure Cloud Infrastructure It marks a significant step forward in the company’s and Microsoft’s efforts to innovate in artificial intelligence (AI) and cloud computing. These chips are introduced Planned for 2024.
Azure Maia 100 as an AI accelerator
Maia 100 is designed as an AI accelerator and specifically… Training and estimating large language models (LLMs) can be deployed in the cloud. Microsoft intends to use Maia to run some of its largest AI workloads on Azure, including parts of its multi-billion dollar partnership with OpenAI, for which Microsoft provides all workloads. Maya will be on one TSMC’s five-nanometer process Produced and has 105 billion transistors.
Azure Cobalt 100 as a 128-core CPU
The Azure Cobalt CPU offers 128 cores and is based on an Arm Neoverse CSS theme customized specifically for Microsoft. This chip is designed To support common cloud services in Azure. Microsoft is currently testing Cobalt CPUs in workloads such as Microsoft Teams and SQL Server with the goal of providing customers with virtual machines for various workloads.
These new chips are part of a major overhaul of Microsoft’s entire cloud server architecture to optimize performance, power consumption and cost. The contacts of the two chips are also the same A response to the growing demand for powerful AI platforms, as embodied, for example, by Nvidia’s H100 GPUs. Microsoft’s move toward in-house chip development could help reduce reliance on outside vendors while providing a more diverse and efficient infrastructure for cloud services.
Goal: Unlock new AI possibilities
Microsoft emphasizes that the development of the Maia 100 and Cobalt 100 chips is not only intended to improve the performance and efficiency of the Azure platform, but also to better meet the needs and demands of customers. Software adaptation and use of our own data center should be integrated Added value compared to strong competition While opening up new opportunities for the use of AI in various fields.
Exact system specifications or benchmark results for the new chips are not officially available. Microsoft’s push to develop and deploy its own chip technology reflects a growing trend in the technology industry, Tailor-made hardware solutions For specific requirements and challenges in the cloud computing and AI areas.
formula
For reasons of better readability, masculine, feminine and different language forms (m/f/d) are not used at the same time. All personal names apply equally to all genders.
