Nvidia has emerged as an indispensable company where AI is concerned, but the lead it currently maintains is now being threatened by rivals like Advanced Micro Devices Inc. (AMD) and Huawei.
Advanced Micro Devices Inc. Chief Executive Officer Lisa Su believes the company’s latest AI processors are good enough to give Nvidia Corp. a run for its money in a chip market she expects to soar past $500 billion in the next couple of years.
AMD is making a strong bid to capture market share in the AI chip sector. The latest installments in its MI350 chip series reportedly make them faster than Nvidia counterparts and represent major gains over earlier versions, according to what Su said at a company event Thursday in San Jose, California.
Su claimed the new MI355 chips, which started shipping earlier this month, are 35 times faster than predecessors. AMD hopes it can catch up with Nvidia or at least close the gap with its new products, and the stakes are higher than ever as the market revenue is predicted to exceed $500 billion by 2028.
In February, AMD shared a forecast for its data center business that showed the growth is happening at a slower pace than some analysts had predicted. The company believes this new update to its MI range will boost momentum and ultimately prove it can compete with a much bigger rival.
AMD says the MI355 outperforms Nvidia’s B200 and GB200 products when it comes to running AI software and equals or exceeds them when creating the code. It also claims customers can get them for significantly less than Nvidia products.
Response to AMD’s latest presentation from investors was tepid at best, with the shares falling as much as 1.9% despite the news.
For now, Nvidia and AMD are the leading providers of advanced computer graphics chips, which are the basis of components for developing AI and the demand for them has consistently outstripped supply.
AMD is optimistic about the future, convinced its products can eventually compete with Nvidia’s on the global stage but only time will tell.
AMD is not the only tech company chasing Nvidia’s excellence in the chip market; Huawei is too, particularly in China. Its Ascend 910D AI chip is reportedly being tested to rival Nvidia’s H100.
The 910D is an advancement over the 910C, which reportedly achieves performance close to the H100 by combining two 910B chips, though it lags in inference speed. Huawei is hoping to capitalize on Nvidia’s weakness in China.
Export controls in the US have made it difficult for Nvidia to sell advanced chips like the H100 and H20 in China. This has created a significant opportunity for Huawei, which has secured orders from major Chinese firms like ByteDance, Baidu, and China Mobile.
Huawei is a threat to Nvidia’s dominance because it has the backing of China’s government, which is pushing for tech self-sufficiency. However, the company faces challenges due to reliance on SMIC’s 7nm process, which is less advanced than TSMC’s 5nm or 4nm nodes used by Nvidia and AMD.
Furthermore, Huawei’s chips operate on older HBM2 memory, and sanctions from America limit its access to cutting-edge manufacturing tools like EUV lithography, impacting yield rates and scalability.
KEY Difference Wire: the secret tool crypto projects use to get guaranteed media coverage