Hardware for your local AI
Local AI needs a solid foundation.
Our four models give you exactly the performance you need. Everything comes to you ready to use, your data remains in your network and the system grows with you as your requirements increase. This creates a stable foundation for AI that simply supports you in your everyday life.
meloki's hardware Family
Perfect cooperation between software and hardware
meloki is made up of two components that fit together perfectly. The software comes from SECUDOS, the hardware from Landitec. Both have been working together for years and build systems that are technically perfectly coordinated. The components are reliable, well tested and ideally suited to meloki. You buy meloki directly from Landitec and end up with a complete package that runs without any tinkering and immediately provides you with a stable local AI.
Go directly to the Landitec Shop and order now
- Order directly in the store
- Purchase on account
- Own customer area
Choose your Hardware
The four models give you the freedom to use meloki in exactly the way that suits your everyday life. Each variant has a clear technical structure and is perfectly matched to the software. You decide whether you want to start small, supply more users, need a central solution or want to tackle large tasks. The devices come prepared, run reliably and make it easy for you to use meloki wherever you need it.
scope7-4480
For small teams and a quick start. meloki runs cleanly, the system remains quietly in the background and handles light tasks reliably on a daily basis.
- 1 x NVIDIA L4 (24GB)
- AMD Ryzen 7 9700x (16 Core)
- 4 x 32GB DDR5 RAM
- 1 x 1TB 2.5" SSD
- 1 x 1TB M.2 NVMe
- 1U Appliance
scope7-5580
For multiple users working in parallel. meloki responds quickly, carries a noticeably higher load and remains stable and reliable even with frequent queries.
- 2 x NVIDIA L4 (48GB)
- AMD EPYC Genoa 9124 (32 Core)
- 8 x 32GB DDR5 RAM
- 1 x 2TB U2 NVMe
- 1 x 2TB M.2 NVMe
- 1U Appliance
scope7-6680
For centralized use with larger data volumes. meloki processes more extensive workflows and remains controlled with many simultaneous accesses.
- 1 x NVIDIA Tesla RTX Pro 6000 Blackwell (96GB)
- AMD EPYC 9535 (64 Core)
- 12 x 32GB DDR5 REG RAM
- 1 x 4TB U2 NVMe
- 1 x 4TB M.2 NVMe
- 2U Appliance
scope7-6680-MAX
For heavy loads and demanding work. meloki carries heavy models, handles long sessions and remains resilient even under intensive use.
- 4 x NVIDIA Tesla RTX Pro 6000 Blackwell (384GB)
- AMD EPYC 9745 (128 Core)
- 12 x 96GB DDR5 ECC REG RAM
- 2 x 4TB U2 NVMe
- 1 x 4TB M.2 NVMe
- 2U Appliance
Sophisticated hardware
Whether for SMEs or corporations, meloki is scalable as required. Do you need more GPU power? No problem. We offer ready-made AI appliances, but we can also meet your performance requirements if you have special needs. Talk to us!
GPU
8 - 94 GB VRAM
CPU
8 - 128 cores
SSD
256 GB - 40 TB
RAM
16 GB - 1 TB
👋 Write to us
We write back
Telephone
Address
Contact us
