The former CEO of Intel, Pat Gelsingerstirred the debate about the future of technology by stating that a significant advance in quantum computing could be enough to burst the current artificial intelligence bubble.
In an interview with the Financial Times, he suggested that this turnaround could occur sooner than expected and that the dominance of GPUs, essential for training giant models, would be numbered.
Gelsinger’s bet on a computing “trinity”
Gelsinger described quantum computing as part of a kind of “trinity of modern processing”formed by classical systems, artificial intelligence and quantum architectures.
According to him, this combination will define the future of the industry. What caught attention was the deadline he projects: while figures like Jensen HuangCEO of NVIDIA, talk about two decades for the technology to reach maturity, Gelsinger talks about two years.
For the former CEO, this acceleration would be enough to change the industry’s axis. If quantum machines gain large-scale applications, current GPUs, which power much of the global AI infrastructure, would begin to be replaced by the end of this decade.
Reading about Microsoft and OpenAI
The interview also addressed the competitive landscape in AI. Gelsinger compared the relationship between Microsoft e OpenAI with the strategy adopted by Bill Gates in the 1990s.
At the time, Microsoft grew based on distribution partnerships with IBM. For him, something similar happens today: OpenAI would act as entrance doorwhile Microsoft maintains control of the operation through infrastructure and computational scale.
This analysis meets the argument that the AI race is directly connected to the hardware domain. In Gelsinger’s view, a quantum leap could break this logic and redistribute the weight of traditional companies that are currently responsible for the manufacture and use of GPUs.
Behind the scenes at Intel and the execution crisis
In addition to the predictions, Gelsinger spoke about his recent stint at Intel and the state he found the company in when he resumed his position. He stated that there was an environment of deep disorganizationmarked by constant delays, loss of efficiency and internal difficulties in delivering strategic products.
According to him, no product had been launched on time in the five years prior to his returnand the company appeared to have lost basic engineering processes.
The most symbolic case is that of technology Intel 18Adesigned to bring the company back to TSMC’s level in manufacturing capacity. Gelsinger had promised that the project would be completed in five years, but was fired before reaching the deadline. Your successor, Lip-Bu Tanofficially closed 18A within the same window.
During the interview, he summarizes his frustration:
It seemed like the company had unlearned how to engineer products
Pat Gelsingerex-CEO of Intel
Immersion in quantum computing at Playground Global
After leaving Intel, Gelsinger joined the fund Playground Globalwhich invests in deep tech startups. This direct contact with advanced hardware research would have been the trigger for his conviction that quantum systems are capable of making both classic architectures and AI chips obsolete.
Familiarity with experimental projects also shaped his view that the AI bubble is likely to continue for a few more years, driven by market interest and high training costs.
For him, the breaking point would come at the moment when qubits begin to solve problems beyond the reach of current chipscreating an inevitable shift in the industry.
What your predictions reveal about upcoming technology cycles
Gelsinger’s statements highlight an important narrative dispute: while some companies argue that GPUs still have room to evolve, others see quantum computing as the next structural leap.
NVIDIA itself has responded to this type of speculation by stating that its platform maintains advantage in performance and flexibility on specialized chips like Google’s TPUs.
Regardless of who is right, the debate points to something bigger: the hardware sector is entering a phase where old certainties are no longer stable.
The advancement of AI, the increasing cost of models and energy limitations create an environment in which new architectures gain space more easily.
Also read:
Horizon that could redefine the very concept of computing
Pat Gelsinger’s predictions may seem bold, but they carry a relevant point. If quantum computing really reaches the level he predicts, the dispute will not only be technicalbut strategic. Business models, partnerships, supply chains and research priorities can change direction in a matter of a few years.
The possibility opens up space for greater reflection: computing is entering a period in which disruptions are no longer the exception and are becoming part of the companies’ own planning. And, if this scenario is confirmed, the AI bubble may not burst due to saturation, but due to replacement.
Source: Financial Times
Source: https://www.adrenaline.com.br/tech/pat-gelsinger-computacao-quantica-fim-bolha-ia/
