Mukesh Ambani is trying to change one of the most deep-seated habits in personal computing: when a computer gets too old, instead of replacing it, you can move its computing power to the cloud. Reliance Jio has expanded JioPC, a service initially launched in 2025 for its broadband customers, making it available as a standalone subscription to any user with an internet connection in India.
The service turns the local computer into a terminal that displays a virtual PC running in Jio data centers. Configurations scale up to eight virtual CPUs, 16 GB of RAM, and 1 TB of storage. Jio claims that even eight-year-old machines can thus access modern software and artificial intelligence applications without installing new hardware. It is a promise that does not equate to turning an old laptop into an AI PC with a local NPU: the computing happens elsewhere. Yet it is precisely this distinction that makes JioPC interesting.
The cloud PC challenges the replacement cycle
According to IDC data reported by TechCrunch, over 65 million PCs were in use in India in 2025, and the installed base could approach 69 million by the end of 2026. Replacement cycles have lengthened: consumers and small businesses are keeping their machines longer than they did just a few years ago, while the refurbished market offers affordable alternatives.
Jio is attempting to step right into this gap between old hardware and new software. If the user has a display, keyboard, and machine functional enough to run the client, they can purchase computing power as a service instead of buying a new device.
Pricing is part of the strategy
The plans listed by Jio start at 1,000 rupees for two months. An annual subscription costs around 4,000 rupees for the configuration with 8 GB of RAM and 500 GB of storage, or around 5,000 rupees for 16 GB and 1 TB. Connectivity costs must naturally be added on top of these prices.
The comparison with a new PC is not purely arithmetic. A physical computer can be used offline, resold, and customized; a cloud PC, by contrast, depends on network quality and the ongoing availability of the service. But in a price-sensitive market, spreading out costs over time can be more accessible than bearing an upfront purchase.
“AI-ready” means something different
Over the past two years, Intel, AMD, Qualcomm, and Microsoft have tied the definition of an AI PC to the presence of local neural accelerators. Jio uses the term in a different sense: it does not necessarily promise on-device inference, but rather remote access to the power required to run AI applications.
This difference has implications for privacy and latency. A local model can run without sending every input to the cloud; a virtual workstation, by definition, depends on the data center. On the other hand, the cloud allows central hardware to be upgraded without replacing millions of endpoint devices.
The network becomes part of the computer
The most obvious limitation is connectivity. An outage or high latency can make the experience worse than that of a modest but local PC. However, Jio has an advantage that few hardware companies possess: it is also one of the country's leading telecommunications operators. It can therefore treat networking and computing as parts of the same product.
It is a strategy consistent with the history of Reliance Jio, which has leveraged aggressive pricing and infrastructure to lower the cost of access to digital services. The cloud PC extends that logic from data traffic to compute.
The model could be particularly appealing for schools and small businesses
A lab with dozens of aging PCs can be expensive to overhaul all at once. If compute capacity can be centralized, endpoint hardware becomes less critical. The same applies to small businesses that need to run modern applications without managing upgrades on every machine.
Questions of reliability and data management remain. Organizations need to know where files reside, how backups are handled, which applications are supported, and what level of service is guaranteed. The cloud reduces certain maintenance costs, but shifts part of the control to the provider.
The model can extend hardware lifespan
If a computer is replaced because it can no longer keep up with new workloads, shifting the burden to the cloud can avoid or delay that purchase. This has a potential environmental impact: manufacturing a new laptop requires materials, energy, and logistics. Extending the useful life of existing equipment can reduce the number of replaced devices.
The benefit does not come for free, because data centers consume energy. A proper assessment must compare the overall life cycle: cloud energy, network, avoided manufacturing, and the utilization intensity of shared machines. Well-utilized centralized capacity can be more efficient than millions of underutilized chips, but it depends on the architecture.
Competition also comes from refurbished PCs
TechCrunch cites analysts who point out that the refurbished market represents tangible competition. A used but reasonably recent PC provides hardware ownership and works even with imperfect connections. Jio must therefore prove that the cloud advantage outweighs network dependency.
The challenge is turning the service from an occasional workaround into an everyday utility. If users perceive it as a temporary stopgap, the model will remain niche. But if virtualization becomes the norm, it could change the way the Indian market measures the value of a computer.
The real product is access to compute
JioPC suggests that the future of the PC could bifurcate. In premium markets, laptops with powerful NPUs and on-device capabilities will continue to grow; in other segments, the physical machine could become increasingly simple while the value is delivered remotely.
It is an important distinction in the AI era. If running new models requires increasingly powerful hardware, the risk is widening the digital divide between those who can upgrade their computers frequently and those who cannot. The cloud does not eliminate the access problem, but it shifts the cost from hardware to connectivity and subscriptions.
The Indian experiment will therefore be watched beyond India as well. If tens of millions of aging PCs can remain useful by connecting to more powerful infrastructure, the very definition of an “obsolete computer” could change. The machine on the desk would be merely a window; the real PC would live in the data center.
