JAPAC Cloud Coverage
The current state of the public cloud in Asia-Pacific & Japan
South-East Asian economies are in the midst of absolutely staggering growth in the technology sector. Powerful factors like mobile penetration and e-commerce market potential are driving a massive influx of funding to startups in the region. Asia accounts for 44% of all VC funds invested globally in 2018 (KPMG Venture Pulse). Leading the pack is Singapore-based Grab raising $2.5 billion, Indonesia-based Go-Jek with $1.5 billion and India-based OYO rooms with another $1 billion from VCs. Just last week saw the news that Indonesia-based travel startup Traveloka is raising another round at a valuation of $4.1 billion up from around $2 billion last year.
Where there are high growth startups there is enormous demand for cutting-edge cloud infrastructure and clearly this is not lost on the major public cloud players. A few days ago, Google Cloud announced plans to build a region in Jakarta on the heels of previous announcements of new regions in Osaka & Hong Kong. This brings GCP’s tally to an astonishing 18 zones across 6 new regions announced or launched in just the last two years.
CloudCover sees this blistering pace of growth first-hand thanks to our amazing customers and since we support AWS, Azure and GCP, we’re in a fairly unique position. We figured we should take advantage of this and do a little research.
The Current Map
The scale of investments in public cloud infrastructure in JAPAC is quite impressive, as is the speed with which new regions have propagated. The current map shows the fantastic coverage we’re getting in our part of the world with strong representation (existing or announced) in vital geographies from all 3 clouds.
Competition in the region is fierce with each cloud vying for bragging rights and service differentiation in key markets around JAPAC. Last year AWS announced a new region coming to Hong Kong in 2018 and earlier this year brought its local region in Osaka online to give Japanese customers geographic redundancy. Google have committed to adding 3 new regions, at least 2 of which are expected to launch this year. Azure (rather conspicuously) haven’t announced any regional expansion so far this year.
By the numbers
At the end of 2017 regional coverage was a very close race with all the majors within spitting distance of each other. Taking this year’s announcements into account changes the landscape considerably.
Of particular note are the dotted lines showing the trend. Google’s acceleration is dramatic and shows the enormous investment they are making in the region.
However, it does depend on how you look at the numbers. Microsoft has historically had a different view on availability zones and what constitutes a “region”. They’ve recently changed their stand and have begun rolling out availability zones within existing regions, but this has yet to happen in any of their Asian data centers. When seen purely from the perspective of uniquely located regions (distinct cities, not counting zones), the same graph looks quite different.
You can see the data used to generate these graphs in this Google sheet. Please let us know if you find any errors or omissions.
While coverage is important, it doesn’t really tell the whole story. Features and new products often don’t launch in new regions until much later, usually starting with basic VM and storage services first. Amazon Web Services certainly boasts the highest installed capacity of computation and storage despite fewer regions in Asia than their competitors, thanks to both the platform’s maturity and its enormous customer base. Unfortunately, installed capacity is not publicly reported by any of the clouds making it impossible for us to compare them along that axis.
Irrespective of the method used to tally everything up, it’s very clear that the biggest winner here is the customer. Enterprises and startups in the region have unprecedented access to world-class cloud infrastructure without the usual compromises that we’ve grown accustomed to.
The trend of more local regions seems bound to continue as more governments and large enterprises in Asia begin to consider cloud technology for their internal workloads. In most cases, data sovereignty laws preclude key workloads like banking, insurance and telecom from using cloud services unless they are hosted in the country.
Asia is not an easy part of the world for a cloud provider to address. Our study doesn’t include China because it has different rules from the rest of the market and while the startup ecosystem is booming, it is hard for outsiders to play there. Google’s relationship with China in particular has been strained in the past (to say the least), though it seems to be thawing now. One can only wonder if Google Cloud is one of the reasons for the change of heart.
South-East Asia is no picnic either. The region is fragmented into many small markets across language, culture and religion. Many countries in the region have fairly complex laws governing foreign companies and direct investment. These issues coupled with the usual challenges of relative market size and willingness to switch to cloud infrastructure complicate strategy for the big clouds.
That said, the opportunity for public cloud players to make a splash in these markets seems too big to pass up. Perhaps ideas like AWS’s local region, Azure Stack or GKE on-prem will become more prevalent for smaller markets, granting data residency and latency benefits without the massive capital investment necessary for a full-blown cloud region.
We expect more of the same in the coming months as the big boys battle it out. Strap in, it’s going to be a wild ride.
