Understanding eSIM, eUICC, SGP, iSIM and Multi-IMSI: How the Buzzwords Fit Together

SIM cards have come a long way from the small plastic chips we used to swap phone to phone. But along this journey, the industry collected a vast number of acronyms, some of which overlap.

This guide untangles the buzzwords and explains how they work together, from the tiny chip inside a device to the global standards that help users, innovators and network operators manage connections remotely.

‍

What is a SIM card?

We’re all “familiar” with the concept of a SIM card. Chances are you’ve handled one yourself. But to understand the complex terminology around cellular connected devices, it’s worth tightly defining what a SIM card is, and why it exists. 

Any device trying to connect to a cellular network needs to:

• know which network to connect to
• and possess the required credentials to be allowed to connect to that network

For example, your iPhone can’t just connect to any available network. Just because your device detects a Vodafone tower doesn’t mean you can connect to Vodafone, you can only connect if you’re a Vodafone subscriber.

That’s why devices need a subscriber identity module, or SIM card. 

A SIM card is a plastic card with simple electronic chips that stores a mobile network operator (MNO) profile. The profile is unique to a specific mobile operator, and contains the IMSI (international mobile subscriber identity) and the credentials you need to connect to that network.

‍

SIM Form Factors

SIM cards have been shrinking over time, reflecting the need to save space in increasingly compact devices. It started with the full‑size SIM (1FF) in the early 1990s, roughly the size of a credit card

As mobile phones got smaller, the industry introduced the Mini‑SIM (2FF), the classic plastic card most people remember sliding into feature phones, followed by the Micro‑SIM (3FF).

< insert an image comparing SIM sizes, with bullets of what the SIM does >

Today, most devices use Nano‑SIM (4FF).

But physical, removable SIM cards have obvious drawbacks: there’s a limit to shrinking a chip before it becomes difficult to handle, which leaves formats like 4FF too large to embed into tiny IoT devices. 

That led to the evolution of MFF2 (or machine form factor 2). Unlike earlier removable SIMs, the MFF2 is a tiny chip soldered directly onto a device’s circuit board.

It can’t be removed, but it takes up less space – making it ideal for industrial or IoT deployments device size and limited physical access make removable cards impractical.

The MFF2 standard is what we now call the first true eSIM form factor.

‍

What is an eSIM?

The term eSIM is often misunderstood. It’s common to assume the “e” stands for “electronic,” like “e‑mail” or “e‑banking”, and so in other words; marketing material often leaves users thinking that an eSIM is an “electronic SIM”, or in other words, a non-physical SIM.

There is some truth to that, but from a cellular industry standards perspective, eSIM has a very specific meaning. 

An eSIM is an embedded SIM card.

< Insert an image with two devices: one with a plastic card held in a hand as it is pushed into the device, the other of a device with the SIM inside the device, and a red cross on the hand outside the device>

So, instead of being a removable plastic card and chip that slides into a device, an eSIM is built directly into the device circuitry. That’s how eSIM and MFF2 are connected – eSIM became a reality with MFF2.

Functionally, an embedded SIM serves the same purpose as a physical SIM: it stores an operator profile and enables network authentication.

‍

MFF2 v. iSIM vs. SoftSIM

MFF2 was just the first way to embed a SIM card into device circuitry. As eSIM adoption grew, manufacturers experimented with ways to make SIMs even smaller and more integrated into device hardware. 

Two intermediate designs emerged after MFF2:

• WLCSP (wafer‑level chip scale package), packaged as part of the whole wafer so it is the same size as a silicon die itself
• MFF‑XS, an ultra-compact version of MFF2 soldered directly into the circuit board.

Both maintain the same electrical function as MFF2 but occupy less space, which makes it more practical for ultra‑compact IoT modules, wearables, and sensors.

iSIM (integrated SIM) moves a step further by embedding the SIM function directly into the device’s system‑on‑chip (SoC), alongside the processor and modem. This eliminates the need for a separate SIM component altogether, again saving space but also improving security because credentials are stored in a secure enclave within the chip itself.

As a next step, SoftSIM is purely software‑based. There’s no dedicated secure hardware element. Instead, SIM functions are emulated in software on the device. While appealing for flexibility and cost, SoftSIM as it stands has security, certification, and compatibility concerns, but does work well in some applications.

<Insert image of MFF2/iSIM/SoftSIM with iconography and three short bullets each>

Technically, MFF2, iSIM, and SoftSIM are all embedded SIMs, so you could call any of these formats an eSIM.

That said, as we mentioned, the way “vendor marketing” makes use of the eSIM really refers to a broader functional set.

‍

How eSIM relates to eUICC

Most vendors talking about eSIM refer to a combination of two things. First, the hardware standard called embedded SIM and the associated form factors. Second, the eUICC functionality. 

<Insert an image with an embedded SIM + eUICC software + SGP = “eSIM”>

So, what is eUICC? A traditional UICC, a Universal Integrated Circuit Card, is the reprogrammable platform in classic plastic SIMs. The SIM is the hard drive, while the UICC is the software on the hardware. UICC software is typically locked to one carrier profile at a time. 

eUICC (embedded UICC) takes this further by adding remote management to a UICC. With eUICC the device can securely download, store, and switch between operator profiles over the air, without physical card swaps or reboots.

The eSIM has the physical foundation, while eUICC software retrieves and stores the operator profile securely and adapts to new networks.

Note that using eUICC is not predicated on the device making use of an embedded SIM – you can integrate eUICC capabilities into a physical, 2FF/3FF/4FF SIM card that is not an eSIM.

‍

SGP Management Standards

There’s a third layer to the “eSIM technology stack”, the policies that help eUICC implementations manage the retrieval of mobile operator profiles.

Like many cellular networking standards, these policies are managed by the GSMA (the GSM Association), which uses the SGP naming scheme for its eSIM and Remote SIM Provisioning (RSP) specifications. 

SGP outlines how operators and devices should talk to each other when network profiles are downloaded, activated, or deleted over the air. In other words, SGP standards are the rulebook that keeps eSIM+eUICC implementations interoperable across different networks, SIM vendors, and device manufacturers.

‍

Understanding the M2M eSIM standard – SGP.02

M2M is a broadly used technology term that points to any form of machine to machine communication – whether it’s a wire connecting two devices, or a cellular standard that enables M2M communication. Think factories, smart meters and connected healthcare.

So, for example, vendors might refer to an M2M eSIM, and all they mean is that they have an eSIM solution that’s appropriate for machine-to-machine communications.

That eSIM solution requires a standard for downloading network profiles, and that’s where the GSMA’s SGP.01 and SGP.02 standards come in.

SGP.02 is a push-based model created in the early 2010s, driven by automotive and other industrial use cases where devices have no user interface and where devices are difficult or impossible to access physically:

• SGP.01 defines the high-level system components that make the standard work, including the subscription manager data preparation (SM-DP) for creating profiles and subscription manager secure routing (SM-SR) for managing secure channels to the eUICC.
  
  
• SGP.02 provides the technical implementation details for that same M2M architecture. It specifies the protocols, security procedures, APIs, and exact workflows for profile downloading, deletion, and error handling over SMS or IP connections. 

The standard does have drawbacks. With SGP.02 the devices cannot “request” a new profile; profiles can only be pushed. Integration can be complex, and it is difficult to switch connectivity providers later, without significant coordination or re‑manufacturing.

‍

Understanding the consumer eSIM profile, SGP.22

The consumer eSIM specifications SGP.21 and SGP.22 were created to solve a different problem from M2M: how to let end users add, change, or remove network subscriber profiles on their own devices, without swapping physical SIMs.

Where M2M assumes there is no user interface, and everything is controlled by the operator, the consumer model assumes the opposite – that an end user who initiates the profile download controls the device.

It’s now commonplace: you buy a new iPhone, and your network operator either provides a QR code or a link that enables your new device to download the operator profile.

<Insert image of smartphone + QR code + pull arrow from the cloud>

SGP.21 and SGP.22 are the consumer eSIM standard that enables the pull architecture that allows users to “retrieve” an eSIM profile from the network operator. SGP.21 defines the overall architecture and roles, while SGP.22 defines the technical details and protocols for Remote SIM Provisioning in the consumer world. 

The model supports multiple active or inactive profiles on a single device so users can easily switch operators or add temporary plans (for example, a travel eSIM) without any hardware changes.

That said, using SGP22 requires a “head” for the device – some or the other way to access the device software. That doesn’t work for headless devices with no screen or route to access the offline device, and doesn’t work where devices are in inaccessible locations.

‍

What about SGP.32?

SGP.32 is the GSMA's latest eSIM standard, released specifically to address IoT deployments that fall between traditional M2M (like automotive or industrial machines) and consumer devices (like smartphones).

Building on lessons from SGP.02 and SGP.22, it introduces a more streamlined architecture optimized for massive-scale, low-maintenance IoT—think smart meters, asset trackers, and sensors.

Unlike the rigid, push-based M2M model (SGP.02), SGP.32 adopts a pull-based provisioning flow similar to SGP.22, where devices can request profiles more flexibly without being locked to a single backend provider from manufacturing. 

<insert image of SGP02, SGP22, SGP32 with appropriate iconography and three bullets each>

Key benefits include easier multi-operator switching, lower integration costs for device makers, and better scalability for connectivity platforms managing millions of devices. While adoption is accelerating in 2026, it's not yet universal.

‍

Understanding multi-IMSI SIMs

Both traditional SIMs and eSIMs by default store only one subscriber identity at a time, even if eSIM has the flexibility to change the subscriber identity on the SIM.

A Multi-IMSI SIM (multiple international mobile subscriber identity SIM) is technology that stores several subscriber identities—or IMSIs—on a single SIM, whether physical SIM or eSIM.

It allows the device to automatically switch to the best available network based on signal strength, location, or cost—without any manual intervention or remote commands. 

It's like having multiple pre-loaded passports on one card, letting the device pick the right one for the country it's in.

<insert image with IMSI SIM, within the box multiple operator profiles, plus arrow to intelligent brain that switches these profiles>

When using an eUICC, the Multi-IMSI works as a dedicated applet that runs within the secure environment of the eUICC. 

This applet manages the stored IMSIs, monitors network conditions, and handles seamless switching logic (often triggered by rules like roaming limits or coverage thresholds). 

Unlike full operator profiles, IMSIs are lighter-weight, enabling faster transitions measured in seconds rather than minutes:

• Compared to eSIM/eUICC alone: eSIM/eUICC focuses on remote profile provisioning, which works for long-term flexibility but doesn’t enable instant network switching when a device crosses a border. Multi-IMSI is simpler and faster for short-term network selection.
  
  
• Compared to SGP.32: SGP.32 enables pull-based profile downloads to avoid vendor lock-in. But, just as with eSIM/eUICC, multi-IMSI complements it by providing instant switching.

So yes, you could have an SGP.32-enabled eSIM, with multi-IMSI capabilities – which combines the ability to fully control what’s on your SIM, over the air, with the ultimate network selection flexibility.

‍

Which standard should you choose?

It really does depend on your unique use case: is the choice of network operator up to you, as service provider, or is it down to your end user? How big is the device, and do you have physical access to the device?  Do you need network redundancy, and do you need to be able to update operator profiles once the device is in the field?

One point that can help you is to have a connectivity partner that offers both the physical hardware flexibility to provide connectivity for your use case, and access to a large number of networks.

‍