Nvidia aside, the German supercomputing stock is up 95% this year | Wilnesh News
While chip darling Nvidia has been making headlines for its impressive stock performance driven by excitement about artificial intelligence, a low-profile German company has quietly seen its share price nearly double in 2024. Shares in supercomputing systems developer ParTec soared more than 95%. So far this year, investors have begun to take notice of its potential. Although the company has been listed on the Frankfurt Stock Exchange for less than a year, its roots date back to 1999, when it was spun out of Karlsruhe Institute of Technology in Germany. The small company, which had sales totaling 52 million euros ($56 million) in the first half of last year, secured a contract worth 300 million euros ($327 million) to build Jupiter, the continent’s first ultra-powerful supercomputer system , is expected to be able to perform at least a billion (or billion) calculations per second. What do investors think of JY0-DE 1Y line? Hendrik Leber, fund manager at Acatis Datini Valueflex Fonds, owns shares in ParTec and said he sees the company as a “very clear” investment opportunity. “Europe wants to spend billions of dollars on supercomputing over the next few years, and they want to invest locally in the EU. ParTec has the technology to do that and will be well-positioned in these public tenders,” Leber added . What exactly does ParTec do? ParTec has 60 to 70 employees and provides the software systems needed to build supercomputers, primarily for research. Miguel Lago Mascato, senior equity analyst at equity research firm Montega, told CNBC Pro: “Their main product is software ‘middleware,’ and in the future may be quantum computers.” Lago Mascato expects the company’s stock price to rise again in the next 12 months. 21% to €130. While rivals like Hewlett-Packard can also bid for supercomputing projects, Largo Muscato said ParTec’s middleware provides them with a “unique selling point.” Patents and royalties Earlier this year, independent auditors valued ParTec’s 150 patents for building and designing supercomputers at €767 million, which the company expected would “significantly strengthen” its equity base. However, some have expressed doubts about the company’s ability to monetize its patent portfolio. “Currently, the likelihood that ParTec will realize licensing revenue from its patent portfolio in the near future is very low,” Lago Mascato said bluntly. However, fund manager Leber dismissed concerns about a lack of royalties, pointing to the €300m Jupiter contract as an example of real cash flow. “The orders are very real and the economic profits are very real,” Leiber said. “I don’t care about the combined valuation of the patent portfolio. What I see is a very real computing center, and that’s where the revenue and profits are coming from.” Liquidity Another risk factor VCs should be aware of for ParTec is the liquidity of its stock limited. The company’s free-floating share capital is less than 13%. Leber noted that even a €2 million share placement would take about a month to complete due to thin transaction volumes. This lack of liquidity means that the share price may not always effectively reflect the underlying value of the company. Future Opportunities in AI ParTec sees significant opportunities beyond traditional supercomputing by providing systems specifically designed to train large AI models. ParTec CEO Bernhard Frohwitter told CNBC Pro that there is a huge demand for systems with context understanding capabilities in addition to language models. Large language models, such as the one behind ChatGPT, generate text, images, and videos by predicting the next word or pixel in a sequence. However, Frohwitter said the AI model required by the company requires “internal logic” from physics, mathematics and mechanics. Frohwitter added that ParTec has been “under tremendous pressure from institutions, even governments, to build these types of machines” to train these basic AI models. “We are in the process. In a few months, we will provide artificial intelligence machines that can perform all functions of training, data management, inference and so on at the same time,” he said.
