A SIM-based proxy gateway provides two different numbers that buyers often merge: the number of concurrent egress paths, which equals the healthy active SIMs, and the number of unique IP addresses, which depends on the carrier's addressing. The honest answer is that the SIM count sets the paths and the carrier sets the IP uniqueness, so the capacity planning has to measure both.
This guide explains the two numbers, the capacity math, pool sizing by use case, and what to verify on a proxy gateway before planning around a number.
The Two Numbers
The first number is the egress path count: each active SIM establishes a mobile data connection, and the gateway exposes each connection as a routable path. The path count is a hardware and SIM planning number, and it is the number the gateway controls.
The second number is the unique IP count: how many distinct public addresses the fleet presents, which is decided by the carrier's addressing architecture. Some carriers assign a unique address per connection, while others share addresses through carrier-grade NAT, and the difference is invisible until the IP behavior is tested.
The two numbers drive different decisions: the path count decides how much concurrent traffic the pool carries, and the unique IP count decides whether the pool looks diverse to the services it reaches. Both matter, and neither can be assumed from the other.
The path count also has a concurrency dimension: a single egress path can carry many sessions, so the path number is not the session limit, and the gateway's throughput decides how much traffic each path handles. The capacity plan therefore includes both the path count and the throughput per path.
The distinction matters for contracts and expectations: a supplier that promises IPs without naming the carrier's role is selling a number it does not control. The buyer's verification is the antidote, and the supplier's willingness to discuss carrier behavior is part of the evaluation.
The gateway's role in the unique IP number is rotation and selection, not creation: it cycles and assigns the addresses the carrier provides, and its value is the control over that process. Understanding the division of labor sets the right expectation for what the hardware can change.
The Capacity Math
The path capacity is simple: the number of active, healthy SIMs equals the number of concurrent egress paths, minus the SIMs that are idle, blocked, or off signal. The practical formula is healthy SIM count, not slot count, which is why fleet maintenance feeds capacity.
The unique IP math is carrier-dependent: on a carrier with per-connection addressing, each active SIM may present its own address, while on a CGNAT network, many connections can share a pool of addresses. The math must be measured per carrier, because the same gateway shows different numbers on different networks.
Worked example: a gateway with 32 SIMs on two carriers might present 32 concurrent paths, but the unique IP count could be 32 on one carrier and fewer on another, depending on addressing and rotation. The planning number is the measured one, not the slot count.
| Planning factor | What it sets | How to measure |
|---|---|---|
| Healthy SIMs | Concurrent paths | Status dashboard |
| Carrier addressing | Unique IPs | Egress test per carrier |
| Rotation | IP change rate | Session observation |
| Carrier-grade NAT | Shared addresses | Address repeat check |
The table is the capacity reality check: each factor sets a number, and each is measured rather than assumed.
The healthy SIM count changes over time: a card goes off signal, a plan expires, or a carrier throttles, and each event reduces the concurrent paths. The capacity math is therefore a live number, refreshed from the status dashboard rather than fixed at purchase.
The worked example should be run per market: a fleet of 32 SIMs across three countries presents different path and IP numbers in each, because the carriers differ. The per-market measurement is the only way to plan a multi-market pool.
Pool Sizing by Use Case
For a workload that needs concurrent paths, size by the healthy SIM count with headroom: data collection running many parallel sessions needs more paths, and the fleet is sized accordingly. The path count is the planning number for this case.
For a workload that needs IP diversity, size by the measured unique IPs per carrier, with rotation configured to cycle the addresses the carrier provides. The diversity number is the one to verify, because it is the property the workload depends on.
For a workload that needs stability, size a pool with static or session-based rotation so each session keeps one path, and confirm the carrier keeps the address stable for the session's duration. The three use cases lead to three different pool designs on the same gateway.
The use cases can share the gateway but need separate pools: a data-collection pool with rotation and a stability pool with static sessions coexist on the same unit with per-pool settings. The pool design is the interface that keeps the workloads from interfering.
The pool plan should be written down with its numbers: the workload, the required paths, the measured unique IPs, and the rotation mode, because the plan is what the verification results are compared against. The written plan turns the pool from a configuration into a spec.
What to Verify
The first verification is per carrier: place connections through each carrier's SIMs, record the egress addresses, and count the unique values and the repeats. The test answers the carrier's addressing question directly.
The second verification is over time: repeat the address check across a day and across sessions, because dynamic addressing and rotation change what the pool presents. The time test catches the behavior that a single snapshot misses.
The third verification is the gateway's reporting: the dashboard and connection logs should show the egress address per session, because the records are what make the capacity numbers repeatable. A proxy gateway that reports the address per connection turns the measurement from a project into a routine.
The verification should also include the session behavior: how long an address stays stable within a session, because a workload that needs persistence depends on that behavior. The stability test is as important as the uniqueness count for those workloads.
Record the verification results per carrier and date, because the numbers change and the record is the reference for the next review.
Telarvo Expert Views
The number buyers ask for is IPs, and the number that matters is paths and measured uniqueness. We advise running an egress test per carrier before planning the pool, because the same slot count produces different unique IP numbers on different networks, and the measured number is the only one worth planning around.
— Network Infrastructure Engineer, Telarvo Store
Validation note: IP uniqueness and persistence depend on carrier addressing, CGNAT, and rotation; verify in your operating markets.
Conclusion
The gateway's IP capacity is two numbers: concurrent paths from healthy SIMs, and unique IPs from the carrier's addressing, with pool sizing following the workload and verification per carrier making the numbers real.
Key Takeaways for B2B Buyers
Plan paths by healthy SIM count, verify unique IPs per carrier with an egress test, size pools by the workload's need, and use the gateway's per-session address reporting for the routine.
Questions to Ask Before Committing
Ask how the gateway reports the egress address per session, how rotation is configured per pool, and which carriers the supplier has tested for IP behavior.
Ask Telarvo Store which proxy gateway configuration matches your path and diversity needs before you plan the pool.
FAQs
Is one SIM equal to one IP?
Not always; the carrier's addressing decides IP uniqueness, so the SIM count sets paths and the carrier sets the unique IP number.
How many IPs can a 32-SIM gateway provide?
Up to 32 concurrent paths, and the unique IP count depends on the carrier's addressing and rotation; measure it per carrier.
What is carrier-grade NAT?
A carrier technique where many connections share public addresses; it reduces the unique IP count a pool presents and must be verified.
How do I get more unique IPs?
Add SIMs on carriers with per-connection addressing, configure rotation, and verify the measured uniqueness per market.
Why does my pool show repeated addresses?
Carrier-grade NAT or a carrier with limited ranges can share addresses; verify the carrier's addressing and add diversity where the workload needs it.
Can rotation create more unique IPs?
Rotation cycles the addresses the carrier provides, which can increase the number of distinct addresses seen over time, but it does not create addresses the carrier does not have.