Weekend discussion in the ICP community turned sharply toward Cloud Engines, a product designed to let enterprises run private subnets on Internet Computer.
The immediate trigger was DFINITY’s decision to tie Cloud Engines to Mission 70, a deflation plan. Under that structure, node providers keep 80% of Cloud Engines revenue from enterprise sales, while the remaining 20% goes to the network to buy and burn ICP. Based on the figures cited in the article, if enterprise demand becomes large enough, ICP’s inflation rate could drop from 9.72% to 2.92%, a 70% reduction.
That creates a clean market narrative: enterprises move workloads to the network, revenue is generated, tokens are burned, supply tightens, and valuation can be revisited. With ICP down 99.6% from its all-time high and its circulating market capitalization compressed to $1.1 billion, any thesis that points to a fundamentals-driven reversal is likely to be amplified by the community.
The real question, though, is not whether the story is easy to follow. It is how far along that story actually is.
What Cloud Engines is meant to do
The article frames Cloud Engines as something close to a custom-built data center.
On the public ICP network, applications share the same base pool of nodes. Cloud Engines gives enterprises a dedicated lane instead. A customer can choose its own node provider, decide which country the machines sit in, and control replica counts. The underlying protocol, smart contracts and chain-key cryptography remain the same as on the public network, but hardware ownership and data sovereignty stay with the customer.
For enterprises, one of the main selling points is the lack of lock-in. If a company no longer wants to use a given node provider, it can switch providers without migrating code or rewriting system architecture. The article argues that a major pain point in traditional cloud computing, vendor lock-in, is being addressed here through protocol-level standardization.
Node providers list hardware on opencloud.org. Enterprise users can then filter by geography, hardware specifications and runtime duration, and assemble their own subnet. The minimum starting point is four nodes.
Key roadmap features are not live yet
The article draws a clear line between the current product and the more ambitious version being promoted.
One feature DFINITY has repeatedly highlighted is the use of SEV-encrypted nodes, which would prevent even node providers from seeing the data running on those machines. That matters most for customers in finance, healthcare and government.
But the report says those SEV nodes do not exist today. A developer in the community wrote on the forum that SEV encryption remains a roadmap item, with zero nodes available at present.
Another roadmap item is operator-controlled upgrades. That feature would allow enterprises to decide when their subnets are upgraded rather than being pushed along by protocol-driven upgrades. It has not been launched either.
What Cloud Engines can actually offer now, according to the article, is sovereignty isolation, single-tenancy, no lock-in, hardware choice and fixed costs. That may be enough to attract paying users, but mostly in situations where data sovereignty is a hard requirement.
Existing deployments show feasibility, not scale
The article points to several deployments that are already public: the Pakistan National Cloud Engine and a pilot project with the United Nations Development Programme, or UNDP. It also says Pakistan’s sovereign communication application ChatT has entered its promotion phase.
Those examples show the product can run in the field. They do not show that Cloud Engines has crossed into broad enterprise adoption at scale.
Mission 70 still depends heavily on the demand side
The article highlights a set of figures from the Mission 70 white paper.
ICP’s current cycle burn rate is roughly 0.0689 TCYCLES/s. To hit the stated 70% deflation target, that figure would need to reach 0.77 XDR/s, which the article describes as about 15 times the current level.
Mission 70 splits the inflation-reduction effort into two parts:
- supply-side cuts to node rewards, accounting for 44% of the reduction,
- demand-side growth in token burns, accounting for the remaining 26%.
The supply side looks more controllable. The article lists a 40% reduction in first-generation node rewards and a cut in minimum subnet size from 13 nodes to 7, both of which can be decided through governance votes.
The harder part is the demand side. For 20% of Cloud Engines revenue to create meaningful ICP burn volume, Cloud Engines itself first needs to generate materially larger revenue. Using the article’s current baseline of about $6,000 in daily burns, new burns would need to rise from near zero to tens of thousands of dollars per day for that leg of the thesis to stand on its own.
Community conviction is not uniform
Not everyone discussing Mission 70 is treating the timeline as a given. The article cites a Reddit analyst whose position captures that hesitation. He is said to be bullish on ICP over the long term, but skeptical that the 70% target will be achieved within 2026 without stronger evidence.
In his view, the market would need to see cycle burn rates rise consistently and hold in a range of 0.3 to 0.5 TCYCLES/s before the trend can be treated as established.
Near-term catalysts are different from the long-term bet
The report argues that Cloud Engines is aimed at a real problem. Enterprises want cloud infrastructure, but they do not always want to be tied to AWS or Azure, and some workloads carry strict data-sovereignty requirements. Using blockchain infrastructure to serve that demand is presented as a coherent idea.
From an investment standpoint, though, timing matters.
The narrative is running ahead of the product. The 20% burn mechanism is already being treated in the community as a central reason for a token re-rating, but Cloud Engines has only just entered Early Access, SEV-encrypted nodes remain on the roadmap, and total daily burn volume is still below $10,000. At the same time, ICP is down 99.6% from its high and its circulating market cap is $1.1 billion, which the article says leaves it looking inexpensive on valuation.
Each link in the chain takes time: Cloud Engines must move from Early Access to formal commercial use; enterprise buyers must move from pilots to scaled procurement; node providers must grow from a small set into a competitive market; and the 20% revenue share must shift from a paper mechanism into burn volume large enough to matter.
That is why the article treats the supply-side elements of Mission 70 as the closer catalyst. If governance proposals to cut node rewards and shrink subnets are implemented over the coming months, ICP’s inflation rate could step down before Cloud Engines has a measurable impact on burn demand. On that basis, the short-term effect on price would be more direct.
Cloud Engines, by contrast, is described as the longer-range wager: that ICP can carve out even a small piece of a trillion-dollar cloud market, and that this slice of revenue can make the 20% burn rule create real deflationary pressure.
Re-rating still waits for measurable burn figures
The article’s final judgment is that Cloud Engines is, for now, a trade where the narrative is ahead of the fundamentals. The Pakistan and UNDP cases show feasibility, but not scalability.
How far away is a token re-rating? The answer given in the piece is simple. The market likely needs to wait for the first batch of measurable burn data generated by Cloud Engines. Until that happens, calculations around future deflation remain an attempt to discount future revenue into today’s price.


