Both ANTAREX and EVEREST cite DSL development and compiler work as central contributions, making this the most consistent signal across their H2020 portfolio.
SYGIC AS
Slovak technology SME building compilers, DSLs, and programmability tools for heterogeneous HPC and big data platforms.
Their core work
Sygic AS is a Slovak technology SME that contributes compiler engineering, domain-specific language (DSL) toolchains, and runtime management software to research consortia working on high-performance and heterogeneous computing. In ANTAREX, they helped build the programming infrastructure that allows HPC applications to auto-tune themselves for energy efficiency across heterogeneous processor architectures. In EVEREST, they extended this expertise toward big data analytics platforms, working on programmability abstractions and high-level synthesis tools that let developers target diverse hardware without rewriting code. Their core value to a consortium is practical software tooling that bridges the gap between algorithm design and efficient execution on non-standard hardware.
What they specialise in
ANTAREX targeted heterogeneous HPC systems for energy efficiency; EVEREST explicitly addresses heterogeneous platforms for extreme-scale big data analytics.
ANTAREX (2015–2018) focused on autotuning and runtime adaptivity for exascale systems, reflecting deep expertise in dynamic resource management.
EVEREST (2020–2024) introduced high-level synthesis and programmability as keywords, indicating a newer capability being applied to big data workloads.
Artificial intelligence appears as a keyword in EVEREST, suggesting early-stage engagement with AI-driven optimisation within distributed computing pipelines.
How they've shifted over time
In their first project (ANTAREX, 2015–2018), Sygic focused on the software stack needed to make HPC applications self-adaptive: compilers, DSLs, and runtime systems that could automatically tune energy use across heterogeneous processors. By their second project (EVEREST, 2020–2024), the emphasis shifted from runtime adaptivity toward programmability frameworks and high-level synthesis — tools that help developers describe computations at a high level and map them efficiently onto diverse hardware for big data workloads. The thread connecting both phases is the same: making heterogeneous hardware accessible through better software abstractions, but the application domain has broadened from HPC energy efficiency to extreme-scale data analytics with AI components.
Sygic is moving from low-level HPC runtime management toward higher-level programming abstractions for heterogeneous platforms, positioning them increasingly at the intersection of big data infrastructure and AI-aware computing tools.
How they like to work
Sygic participates exclusively as a consortium partner — they have never led an H2020 project — which suggests they prefer to contribute a focused technical component rather than manage multi-partner coordination. With 15 unique partners across just 2 projects, they operate in medium-sized consortia rather than bilateral arrangements, implying comfort with multi-institutional research environments. Their consistent participant role points to a company that brings a well-defined technical specialisation and integrates it into larger systems built by others.
Sygic has collaborated with 15 distinct partner organisations across 6 countries through their two H2020 projects, suggesting a moderately international but not deeply global footprint. Their network is concentrated in European research and technology organisations active in advanced computing, with no evident geographic anchor beyond their Slovak base.
What sets them apart
Sygic brings industrial-grade software engineering — specifically compiler construction and DSL tooling — into academic research consortia that often lack this practical implementation capability. As an SME rather than a university or large system integrator, they can move quickly and focus narrowly on the software toolchain layer without the overhead of larger partners. For consortium builders targeting heterogeneous computing or big data infrastructure projects, they fill the specific niche of "who will actually build the programming tools that make the hardware usable."
Highlights from their portfolio
- ANTAREXTheir highest-funded project (EUR 296,250), addressing the technically demanding challenge of autotuning exascale HPC applications for energy efficiency — a priority area for EU supercomputing infrastructure.
- EVERESTTheir most recent project (ending 2024) shows a deliberate pivot toward big data and AI-integrated platforms, signalling ongoing relevance in post-HPC computing research.