The Eclipse S-CORE (Safe Open Vehicle Core) initiative aims
to rethink the base software for software-defined vehicles, following a strict
“code-first” philosophy rather than relying on endless specification documents.
Presented publicly by leading OEMs, suppliers and the
Eclipse Foundation at the Automotive Electronics Congress (AEK) 2025,
the project has already been described as an “Airbus moment” for the automotive
industry. With version 0.5 Alpha, a tangible implementation milestone is now
available, offering core modules such as communication, orchestration, logging
and persistency.
But theory alone is not enough. At ELIV 2025, Qorix – the
joint venture of KPIT Technologies and ZF – showcased its Performance Stack
based on S-CORE components running on target hardware. Against this backdrop,
we spoke with Dr Nico Hartmann, CTO of Qorix, to understand how far S-CORE and
its industrial deployment have progressed, what role Qorix plays within the
ecosystem, and where the potential stumbling blocks lie when bringing an
open-source middleware to true production readiness.
Dr Hartmann, what exactly did Qorix present at ELIV 2025
– and how was it received?
One of our two ELIV demonstrators showcased a series
application for Adaptive Cruise Control on a ZF production ECU, running our
Qorix Adaptive Stack. In addition, we presented our Performance Middleware
Stack, which is also used within the S-CORE project of the Eclipse Software
Defined Vehicle Working Group. This demonstrator was integrated with an
Adaptive Cruise Control function on Qualcomm’s Snapdragon Digital Chassis
platform. The response was extremely positive: for many visitors it was the
first time they could see S-CORE components actually running on target hardware
rather than merely as a concept.
What is the technical distinction of the Qorix
Performance Stack?
The Qorix Performance Stack is a
highly capable middleware designed specifically for high-performance computers
in modern vehicle architectures. It enables efficient orchestration of
applications with different criticality levels and provides high-performance
data exchange without redundant copying thanks to zero-copy communication.
The result is low latency, deterministic runtime behaviour and a reliable basis
for safety-critical systems. Building on this foundation, Qorix provides a
production-ready commercial distribution based on the S-CORE approach,
including safety artefacts, certification support and long-term maintenance.
How does Qorix Performance differ from the open S-CORE
base?
Eclipse S-CORE provides the open-source foundation – Qorix
Performance turns it into a production-grade, safety-certifiable middleware
product. Within S-CORE, we contribute features such as zero-copy communication,
deterministic orchestration and DevOps-capable toolchains. Beyond this, we
provide safety artefacts, certification documentation and long-term support –
everything OEMs and Tier-1 suppliers need to bring open-source middleware into
series production. Put simply: S-CORE is the blueprint – Qorix Performance is
the completed building.
Why does zero-copy communication play such a central role
in the Performance Stack?
Zero-copy is not a buzzword but an enabler for the next
level of efficiency in high-performance ECUs. In traditional middleware
architectures, data must be copied between processes, which costs latency and
memory bandwidth. Our technology removes this overhead entirely by exchanging
data between tasks via references in a thread-safe manner. The outcome is
constant latency, deterministic behaviour and up to 40 per cent lower CPU load.
Especially for ADAS and sensor-fusion functions, this marks the difference
between “technically possible” and “fit for market”.
How far along is the implementation of S-CORE components
on real hardware?
The demonstration shown at ELIV is based on the current
S-CORE implementations. This version includes the core modules Communication,
Orchestration, Logging and Persistency – all already running on real target
hardware. Of course, we are not yet talking about production maturity:
certification and safety artefacts are still missing, which aligns with the
development roadmap. Work on version 0.5 will be completed in November. Version
1.0 is planned for 2026.
What is the overarching goal of S-CORE – and why was 2025
the right moment to launch it?
S-CORE aims to provide a publicly available reference
implementation for essential, non-differentiating functions within the
software-defined vehicle. Open source enables the necessary speed and breadth
of collaboration: we develop together rather than solving the same problems
repeatedly. The timing was ideal – the first SDV architectures are in the
field, the industry understands the challenges, and virtually all upcoming E/E
architectures follow this paradigm.
How does S-CORE fundamentally differ from AUTOSAR or
earlier standardisation initiatives?
The crucial difference is the approach: code first. We
produce functioning software – visible, testable and improved iteratively –
rather than writing endless specification documents. Instead of competing
implementations that remain incompatible despite sharing a specification, there
is a single shared base which is extended through distributions. User benefit,
not differentiation, is the focus. The fundamental functionality is developed
once and remains consistent.
How is governance organised: who makes strategic
decisions and how open is access for new partners?
Governance is divided into two parts. Strategically, the
X-CORE Platform Council under the Eclipse Foundation oversees topics such as
public communication, roadmaps and the feedback loop from the automotive
industry. Technically, S-CORE is led by the Project Leads Circle and its
feature teams, also following Eclipse rules. All meetings are public and access
is handled transparently via the Eclipse Foundation’s calendars. New partners
are explicitly welcome – what matters is contribution, whether code, safety or
security artefacts, or process input. Voting rights are assigned on merit,
based on actual contributions.
Qorix is one of the most active members of the S-CORE
project. What role do you play specifically?
Qorix is one of the driving forces behind Eclipse S-CORE. We
are responsible for key architecture and runtime components and contribute a
substantial portion of the code for orchestration, logging and persistency. As
a founding member, we have shaped the project’s technical direction from the
outset – from early feature definitions to implementation on real hardware. We
not only contribute, we demonstrably take a leading role in S-CORE’s evolution.
Why are the founding members almost exclusively German or
European companies – where is the international participation?
Germany, as a major automotive hub, naturally had the
interest and the right starting position to initiate the effort. Under the
motto “less talk, more code”, our aim was to deliver results rather than get
lost in preliminary debates. S-CORE is not a VDA initiative, but the
association views it as an important impulse for the industry. Additional
European partners, such as Valeo, have joined, and the first international
contributions from the commercial-vehicle sector are imminent. Many potential
partners are observing the project closely – and they can see that what is
being developed here is real, usable code.
How realistic is it to bring a safety-certified
open-source middleware to production level?
Completely realistic. S-CORE is not a typical open-source
project that is industrialised afterwards. Every accepted contribution must be
close to production quality, or at least aim for it. We do not accept “proof of
concept” code. From the outset, the project has been designed for safety and
production readiness, following ISO 26262-compliant processes and clear quality
metrics.
When do you expect production-ready applications based on
S-CORE?
We expect the first production-near applications in 2026,
meaning real ECUs in vehicles using S-CORE modules.
How do you ensure that S-CORE does not remain stuck in
presentations – but actually reaches the vehicle?
By focusing on what matters: code. We invest less in slides
and more in functioning software. Our partnerships with chip manufacturers, OS
providers and Tier-1 suppliers ensure that S-CORE runs on real hardware early
on. The ELIV demonstration was a major milestone – in 2026 S-CORE will be
visible in vehicles.
And if you look back at the project three years from now
– what would count as personal success for you?
If vehicle manufacturers say: “Our software runs on Qorix
Performance – and it just works.” That would be proof that, together, we have
created something through open source and industrial discipline that genuinely
transforms the industry.
