advertisement
Nobus Opens Nairobi Availability Zone
Nobus Cloud Services, the Nigerian public cloud provider, has gone live in Kenya with a Nairobi availability zone, designated Nobus-EA-AZ1, allowing Kenyan organisations to provision workloads on its platform locally.
The launch extends into Kenya the platform Nobus has built and established in Nigeria, where it hosts workloads for businesses, enterprises and government institutions. Alongside the launch, the company opened the Nobus Catalyst Programme, through which eligible African startups and enterprises can access free cloud credits and technical support, a customer-acquisition model familiar from the programmes global hyperscalers use to attract early-stage companies.
What An Availability Zone Is
advertisement
In hyperscale cloud architecture, an availability zone is a discrete cluster of compute, storage and networking with its own power, cooling and connectivity. It is isolated from other zones so that a failure in one does not necessarily take down workloads running in another. Providers offer multiple zones within a region so customers can architect systems for resilience across them.
Nobus has designated its Nairobi zone as its first in East Africa, reflected in the “EA-AZ1” designation, signalling the potential for additional zones as the platform expands.
The Proposition Being Exported
advertisement
Nobus belongs to a distinct cohort in the Nigerian cloud market. Operated by Nkponani Limited from Lagos, it is among a group of indigenous providers, including Galaxy, Suburban, Layer3 and Nebula, that have positioned themselves against AWS, Microsoft Azure and Google Cloud on three specific grounds: local-currency pricing, in-country data storage and lower latency for domestic users.
That proposition has gained traction at home. Nigerian startups have migrated to local providers over the past two years, while global platforms have responded directly to the currency argument. AWS began accepting naira payments in January 2025, saying local currencies are important to localising the payment experience.
The same three arguments apply in Kenya, where the Data Protection Act places conditions on transfers of personal data outside the country, sector regulators such as the Central Bank of Kenya maintain their own data residency expectations, and shilling-denominated billing can remove some of the foreign-exchange exposure associated with dollar-priced cloud contracts.
advertisement
Whether those advantages carry the same weight in Kenya remains an open question. The country’s enterprise cloud market has different incumbents, procurement practices and a significantly larger hyperscaler presence than Nigeria had when local providers began gaining ground.
Into A Market That Is Filling Up
The timing places Nobus in an increasingly contested Nairobi cloud market.
Digital Realty opened its 6.4MW Nairobi Two facility in Karen this week, completing iColo’s transition to the Digital Realty brand. Oracle is preparing a Nairobi cloud region to be hosted with iXAfrica. Airtel Kenya broke ground on a Nairobi data centre in August, targeting commissioning in July 2027. Microsoft, Google, Africa Data Centres and other providers already have a presence in the market.
Nobus enters that field competing on factors the larger platforms may find harder to match, particularly currency, locality and price, rather than breadth of services.
It is also notable as an African provider expanding into another African market rather than a global platform entering from outside. That puts Nobus on the side of an argument that has featured prominently in discussions around Africa’s digital infrastructure: that local workloads and local demand are critical to making the continent’s infrastructure investments commercially viable.
The claim that will ultimately determine whether the strategy works is simple: workloads.
A zone that is live is not the same as a zone that is loaded. The real measure of Nobus’ Kenyan launch will be the regulated institutions, enterprises and startups that actually provision workloads on the platform.