Although some investors are trying to determine if Arm could duplicate Nvidia's success with a 200% climb in price this year, the two companies present noticeably different investments. Experts declared that Arm is probably not going to observe a significant increase in income from AI in the short-term, but may gain from AI applications being utilized on devices such as cell phones soon. Richard Windsor, creator of Radio Free Mobile, notified CNBC that it will take three-to-five years before Arm will receive any income gains from AI.
Arm is getting ready for a major IPO, amidst an atmosphere of investors that are especially keen on both semiconductors and artificial intelligence; this is supported by Nvidia's massive 200% surge this year. In this offering, Arm is aiming to generate about $5 billion, which would bring its value to over $50 billion. Additionally, Reuters have said that there is a strong likelihood that the pricing of the shares could even be higher than the top of the projected range, which could be due to Softbank's marketing of Arm as an artificial intelligence leader. The IPO prospectus reported that Arm would be "central" in the transition towards AI-driven computing. However, experts have told CNBC that Arm won't be capturing the benefits of the AI trend in the near future, unlike Nvidia.
Thanks to OpenAI's ChatGPT, AI has recently been brought to the forefront. This AI is generative in nature, meaning it is capable of forming answers to user inquiries based on an extensively trained model. For this process to be successful, a substantial amount of computing power must be utilized, and a type of semiconductor from Nvidia called a GPU is used to train and implement these AI models. This has resulted in a significant increase in Nvidia's profits. On the contrary, Arm designs the blueprints of certain semiconductors, including CPUs, which are found in 99% of the world's smartphones, including those produced by Apple. Although CPUs are sometimes used along with GPUs to train data, they are often not required. Arm makes the majority of its money from licensing and royalties, with over 50% of this income stemming from smartphones and consumer electronics. Despite this, the company has yet to experience much of a benefit from AI. Jamie Mills O’Brien, Investment Director at Abrdn, commented that “Growth in the near term for Arm is really not about AI, it's about mobile, it's about royalty increases. In the longer term, Arm is trying to focus investors' minds on the potential … AI in the edge, AI in the data center, but at the moment that's not a huge part of the company's exposure.”
It is unlikely that the future of Arm's AI will be based on the vast amounts of chips necessary for training expansive data models. Rather, it will likely be a major partner in "edge" AI, meaning computing processes are conducted on the device itself such as a smartphone, instead of in the cloud like ChatGPT. For this to be a possibility, devices need to have low-power yet high-performance chips that are able to handle the computations for AI. Arm is engineering the architecture for such chips. Commenting on this, Peter Richardson, research director at Counterpoint Research, shared with CNBC that, "You're not going to have the same level of compute power [on smartphones and cars]; hence, you need to tailor the model to run locally". He then concluded by saying that "Those processors will almost definitely be Arm-based".
Arm declared in its IPO document that its processors have already been employed to manage AI workloads "and every telephone available on the market is capable of operating AI inference applications like voice recognition and adding filters to digital images." Nevertheless, as informed by Richard Windsor, founder of Radio Free Mobile, to CNBC, Arm is unlikely to observe the advantage from AI filters in its income for at least three to five years.SwotBank has been obliged to publicize Arm as an AI firm like Nvidia," Windsor stated. "On the lengthy run, I'm a big supporter of running AI on end devices, it makes an enormous deal of economic commonalty to the provider of the service, as well as much more general in terms of the quality of the service, confidentiality and security and so forth. But those profits are not yielding to Arm at present."
Comments