The editor of Downcodes has learned that Intel will release its highly anticipated AI accelerator Gaudi3 next week. This chip, known as Intel's AI hero, is designed to handle large-scale AI training and inference tasks, has unlimited scalability, and has high hopes to challenge NVIDIA, which dominates the AI chip market. The release of Gaudi3 is an important strategic deployment by Intel in the AI field, and also indicates that AI hardware competition will become more intense in the future.
Intel's 2024 plan is progressing steadily. The company announced that its latest AI accelerator, Gaudi3, is expected to be officially released next week. Gaudi3 is known as Intel's AI hero and is designed to handle large-scale training and inference tasks, with unlimited scalability.
Intel CEO Pat Gelsinger first previewed the accelerator at last year's Intel AI Everywhere conference. At the subsequent "Intel Vision 2024" event, Intel reiterated its Gaudi3 release plan to address the growing demand for high-performance semiconductors that can train and deploy large-scale AI models.
Intel claims that Gaudi3 is more than twice as energy efficient as the previous generation and can run AI models 1.5 times faster than NVIDIA's H100 GPU. Compared to NVIDIA, Intel says its chips consume less power. It is worth noting that NVIDIA currently occupies about 80% of the AI chip market, and NVIDIA's GPUs have been the first choice for AI developers in the past year.
Although NVIDIA is already technologically ahead of Intel, and its Blackwell series products have been delayed, this is mainly due to mask issues encountered during chip production, rather than design flaws. The Blackwell series promises to provide significant AI performance improvements over competitors such as AMD's MI300X, which AMD also plans to launch soon.
Now, with the timing of its product releases, Intel is in a better competitive position against NVIDIA's Blackwell and AMD's MI300X. The Gaudi3AI accelerator is designed to dramatically improve the performance of AI applications, supporting the latest generative AI and large language models in the data center. Paired with Intel Xeon processors, Gaudi3 can provide enterprise-class performance and reliability in a variety of application scenarios such as 3D generation, language translation, sentiment analysis, and image generation.
Looking forward to 2025, Gaudi3's successor Falcon Shores will combine Gaudi's AI capabilities with powerful Intel GPUs to form a brand new product, which will add Intel's competitive advantage in the market.
Gaudi3 performed well in the test, supporting Meta's open source Llama and TII's Falcon model, demonstrating its potential in the training and deployment of Stable Diffusion and OpenAI's Whisper speech recognition model.
Highlight:
The Gaudi3 accelerator will be released next week and has high energy efficiency and fast computing capabilities.
⚡️ Intel claims it is twice as energy efficient as its predecessor, surpassing the NVIDIA H100.
Falcon Shores, launched in 2025, will further enhance Intel’s competitiveness in the AI field.
The release of Gaudi3 marks Intel's launch of a strong counterattack in the field of AI chips. Its high performance and high energy efficiency will bring new possibilities to the AI industry, and its future development is worth looking forward to. The editor of Downcodes will continue to pay attention to the follow-up progress of Gaudi3, so stay tuned!