LabForty at the ASBIS conference 2026
This year, part of the LabForty team attended a specialised conference organised by our long-standing technology partner ASBIS. The event focused on B2B solutions for building modern AI and data platforms and brought together major names in the industry: NVIDIA, Supermicro and Everpure.
For us, this was not just another event. It was a chance to see what AI infrastructure looks like in a real environment. Strong teams keep learning, and the most valuable knowledge comes directly from the people who build the technology, not only from documentation.
NVIDIA's approach to modern AI infrastructure
NVIDIA speaker Alex Pollack opened the conference with a highly technical session, which is rare in this kind of format. He explained how modern AI models work, where the real limits appear in training and inference, and how real systems solve them.
The section on NVIDIA's open-source ecosystem was especially useful: a library of tools freely available to developers, which makes working with GPU infrastructure much more accessible.
The strongest impression, however, came from the range of applications, from the automotive industry and healthcare to robotics and industrial automation. NVIDIA is no longer only a hardware company. It is becoming the infrastructure layer for entire industries.
What Supermicro presented
Supermicro presented a complete view of the AI infrastructure layer and showed how quickly the field is moving. Its portfolio includes GPU servers based on NVIDIA HGX and PCIe architectures, flexible MGX platforms and multi-node systems such as BigTwin and SuperBlade.
The edge and IoT solutions were of particular interest to us: systems designed to run AI where the data is created, without depending on central infrastructure. Together, NVIDIA and Supermicro offer an integrated model with high performance, scalability and operational flexibility, both in the data centre and at the edge.
What Everpure showed
The closing presentation by Everpure (formerly Pure Storage) was probably the one that started the most discussion after the event. Powerful GPUs alone are not enough. Data has to move quickly, reliably and without unnecessary complexity. Everpure showed how its all-flash storage and Evergreen architecture support an entire AI pipeline, from collecting and preparing data, through training and analytics, to fast and reliable access.
The platform brings on-premise infrastructure and the cloud into one environment and works with NVIDIA and Supermicro solutions. The result is a complete stack for enterprises: not separate products, but one integrated system.
What we take away at LabForty
The conference confirmed what we already see in our projects: enterprise-grade AI infrastructure is moving quickly into everyday business. The combination of NVIDIA, Supermicro and Everpure offers a realistic, scalable approach that will shape technology across industries in the coming years.
For us, this is more than inspiration. These are directions we already apply, for example when we run private AI on a client's own servers. The most valuable thing we can offer our clients is a team that keeps growing. See our AI solutions for what we build.
Questions
- Can LabForty run AI on our own servers instead of a public cloud?
- Yes. We already run private AI on clients' own servers, which keeps data inside your environment.
- Which hardware and storage vendors does LabForty follow?
- We work with ASBIS as a long-standing technology partner and follow the NVIDIA, Supermicro and Everpure stack for AI infrastructure.
- Why does storage matter for an AI project?
- Powerful GPUs alone are not enough. Data has to move quickly and reliably through the whole pipeline, from preparation to training and analytics.
If the question is where a model should run, we can look at your servers, a private cloud, or a public one, and what must stay inside.
