Nov 6, 2026 · 19 min · 9 segments
What comes after the bricks and mortar of traditional data centres?In this episode of Data Centre News Asia "In Conversation", we step into the hot seat with Johnny Blackfield, President of EMEA &…
James MurphyHostSo do you feel then that neoclouds are filling, from your description it sounds like they're filling a gap that was left in a category of infrastructure that hyperscalers didn't already have a plan for.
Does that inevitably mean that the hyperscales will close, because that's habitually what they've done to gain the position in the monopoly, allegedly, they have in their own markets, has been to crush those kind of competitors, or to absorb and acquire those kind of competitors.
So it's an interesting market because there are a few different directions this could go.
because the demand was so high, supply wasn't there, they created this new asset class specifically designed for AI.
But now you see all the hyperscalers, the ones that were the biggest clients to Neoclouds, are spending billions of dollars investing in data center bills that are designed for AI.
But at the same time, the hyposcalers are also the ones that provide data lakes, you know, other kind of storage facilities and all of that, so they can't all be just AI.
They still have to provide enterprise data lakes, enterprise data storage, all of that, and so it's not just AI.
So I believe what's going to happen is that there will always be a place for NeoClouds, because the one thing we haven't talked about here, why NeoClouds still exist and thrive, A typical AWS sells compute, GPU compute, anywhere from $7 to $14 a GPU hour.
Whereas for the same type of compute, a Neo Cloud, because they're so specialized and they can really architect it, optimize the architectures to make it as cheap as possible, sell anywhere from $2.50 to $4 a GPU hour.
They have the price advantage that an AWS cannot compete with because an AWS has to average out that price across all of their assets.
So do you feel then that neoclouds are filling, from your description it sounds like they're filling a gap that was left in a category of infrastructure that hyperscalers didn't already have a plan for.
Does that inevitably mean that the hyperscales will close, because that's habitually what they've done to gain the position in the monopoly, allegedly, they have in their own markets, has been to crush those kind of competitors, or to absorb and acquire those kind of competitors.
So it's an interesting market because there are a few different directions this could go.
because the demand was so high, supply wasn't there, they created this new asset class specifically designed for AI.
But now you see all the hyperscalers, the ones that were the biggest clients to Neoclouds, are spending billions of dollars investing in data center bills that are designed for AI.
But at the same time, the hyposcalers are also the ones that provide data lakes, you know, other kind of storage facilities and all of that, so they can't all be just AI.
They still have to provide enterprise data lakes, enterprise data storage, all of that, and so it's not just AI.
So I believe what's going to happen is that there will always be a place for NeoClouds, because the one thing we haven't talked about here, why NeoClouds still exist and thrive, A typical AWS sells compute, GPU compute, anywhere from $7 to $14 a GPU hour.
Whereas for the same type of compute, a Neo Cloud, because they're so specialized and they can really architect it, optimize the architectures to make it as cheap as possible, sell anywhere from $2.50 to $4 a GPU hour.
They have the price advantage that an AWS cannot compete with because an AWS has to average out that price across all of their assets.
The rest of this transcript — segmented and speaker-labeled, so you land on the exact moment something was said
Search every transcript — by keyword, by phrase, or by meaning, across every show Radar indexes
Trends — what is surging across podcasts, measured against its own baseline
Alerts — when a name you follow appears in a newly indexed episode
No account is needed to search Radar.
