据热点科技报道，人工智能领域再添新里程碑，Stability AI近日发布的一项报告显示，英特尔的Gaudi 2计算卡在AI图像生成训练方面显著优于英伟达的高端芯片。该报告详细对比了两家巨头的计算卡在处理大规模模型时的效率，结果显示，Gaudi 2在特定设置下表现出色。
具体而言，当使用2B参数的MMDiT模型，训练深度设定为24，并采用BFloat16的混合精度时，Gaudi 2计算卡的峰值性能达到了每秒训练1254张图片。在256容量下，这一速度降至每秒927张。相比之下，英伟达的H100-80GB计算卡每秒只能处理595张图片，而A100-80GB则为381张。这意味着在相同容量下，Gaudi 2的训练速度比H100提升了55%，性能是A100的2.43倍。
这一发现对AI研究和开发领域具有重大意义，英特尔的Gaudi 2芯片可能成为加速AI模型训练的新选择，尤其是在需要高效能计算的大型项目中。Stability AI作为Stable Diffusion模型的开发商，其报告的发布进一步证实了Gaudi 2在AI计算领域的领先地位，预示着未来AI技术的计算效率将有显著提升。
此次报告的发布，无疑在AI硬件市场中引发了一场新的竞争，各家企业将竞相提升自家产品的性能，以满足不断增长的AI计算需求。随着技术的不断进步，我们可以期待AI领域的更多创新和突破。
英语如下：
**News Title:** “Intel’s Gaudi 2 Chip Challenges NVIDIA: 55% Faster AI Image Generation Training”
**Keywords:** Intel Gaudi 2, AI Image Generation, Performance Boost
**News Content:**
**Title:** Intel’s Gaudi 2 Chip Outperforms NVIDIA’s H100 in AI Image Training, New Report Finds
According to Hot Tech News, a new milestone has been reached in the realm of artificial intelligence as a recent report by Stability AI reveals that Intel’s Gaudi 2 accelerator card excels in AI image generation training compared to NVIDIA’s high-end offerings. The report delves into the efficiency of both giants’ cards when dealing with large-scale models, showcasing Gaudi 2’s exceptional performance under specific conditions.
In specific terms, when utilizing the 2B-parameter MMDiT model with a depth of 24 and employing BFloat16 mixed precision, the Gaudi 2 card achieved a peak performance of training 1,254 images per second. At a capacity of 256, this rate dropped to 927 images per second. In contrast, NVIDIA’s H100-80GB card processed 595 images per second, while the A100-80GB managed 381. Consequently, Gaudi 2 demonstrates a 55% faster training speed than the H100 at equivalent capacity and performs 2.43 times better than the A100.
This finding holds significant implications for AI research and development, potentially positioning Intel’s Gaudi 2 chip as a new choice for accelerating AI model training, particularly in large projects requiring high-performance computing. As the developer of the Stable Diffusion model, Stability AI’s report further solidifies Gaudi 2’s leading position in AI computing, foreshadowing substantial improvements in computational efficiency for future AI technologies.
The release of this report has undoubtedly sparked a new competitive dynamic in the AI hardware market, with companies now racing to enhance their products’ performance to cater to the escalating demand for AI computing. With continuous technological advancements, more innovations and breakthroughs can be anticipated in the AI domain.
【来源】https://www.itheat.com/view/45945.html
Views: 5