Cortex Starter AI Desktop Workstation – Ryzen 9 9950X , RTX 5090
£6,949.98 Inc.VAT
- Ryzen 9 9950X
- 32GB RAM DDR5
- 4TB SSD Storage
- GeForce RTX 5090 32GB
- Free Pick up from our London Store
Cortex
Elite Performance, Unmatched Power
The Cortex AI is designed for developers, students, researchers, content creators and small businesses looking for a powerful desktop PC for artificial intelligence (AI), machine learning and local AI applications.
Powered by an NVIDIA GeForce RTX 5090 323GB graphics card, 64GB DDR5 memory and a fast 4TB NVMe SSD, the Cortex AI provides the performance needed for modern AI workloads, software development, data analysis and AI-powered content creation.
Whether you're learning AI, building AI-powered applications, experimenting with local AI models or exploring private AI workflows for your business, the Cortex AI offers an excellent balance of performance, capability and value.
Why We Built the Cortex AI
Artificial Intelligence is changing the way people work, create and learn, but many users are unsure where to start.
The Cortex AI was created as an affordable entry point into local AI computing. Instead of relying entirely on cloud services, the Cortex AI gives developers, students, creators and small businesses a powerful desktop platform capable of running modern AI applications directly on their own machine.
Whether you're experimenting with local AI assistants, learning machine learning, generating AI images, building AI-powered software or exploring private AI workflows for your business, the Cortex AI provides the performance and memory needed to get started.
Powered by an NVIDIA GeForce RTX 5060 Ti 16GB graphics card and 64GB of DDR5 memory, the Cortex AI is designed to handle popular AI tools such as Ollama, LM Studio, Stable Diffusion, ComfyUI, Python and PyTorch.
The Cortex AI is not intended for large-scale enterprise model training. Instead, it focuses on what most users actually need: a capable, reliable and cost-effective workstation for learning, developing and experimenting with AI locally.
Built, tested and supported by House of Computers in London, the Cortex AI is designed to make AI computing more accessible for everyone.
These tools allow you to download and run open-source models (like Llama 3, Mistral, or DeepSeek) directly on your hardware.
Ollama: The industry standard for running LLMs locally via a CLI. It manages model weights and provides a local API.
LM Studio: A GUI-based powerhouse for Windows/Linux. It allows you to "shop" for models on Hugging Face and run them with a single click, providing a chat interface similar to ChatGPT.
vLLM / Text-Generation-WebUI (Oobabooga): Highly customizable engines used for high-throughput serving. With the 5090's massive bandwidth, these allow you to run models at 100+ tokens per second.
This is where the 32GB of VRAM and the 9950X’s 16 cores really shine.
Unsloth: Currently the fastest framework for fine-tuning LLMs. It uses significantly less VRAM, allowing you to fine-tune Llama 3 (8B or 70B) on a single 5090.
PyTorch / TensorFlow: The foundation of AI development. Your machine is built to handle heavy training loops and tensor operations locally.
Docker & NVIDIA Container Toolkit: Essential for isolating different AI environments (e.g., one container for a vector database, another for the LLM).
Beyond text, your workstation is a powerhouse for image and video synthesis.
ComfyUI / Automatic1111: The primary interfaces for Stable Diffusion (XL/3) and Flux.1. The RTX 5090 can generate high-resolution, complex images in under a second.
Forge: A high-speed variant of WebUI optimized specifically for the massive VRAM of 90-series cards.
Topaz Photo/Video AI: Uses local AI models to upscale, de-noise, and "fill in" frames for professional video production.
These apps use Retrieval-Augmented Generation to index your local files (PDFs, codebases, emails) so you can "ask" your computer questions about your own data privately.
AnythingLLM: A full-stack solution that turns your local documents into a private knowledge base.
PrivateGPT: An open-source project focused on absolute privacy—no data ever leaves the 5090.
Continue.dev: An open-source AI code assistant (like GitHub Copilot) that plugs into VS Code or PyCharm but uses your local Ollama instance to suggest code.
The Professional Edge
On a machine with these specs, you can run a Vector Database (like Pinecone or Milvus) in the background while simultaneously serving a 70B parameter model for your team, all while keeping your data within your own four walls.
Cortex AI Workstation
The Cortex AI Starter Workstation is an expertly tuned local node built to handle the most demanding AI frameworks. By pairing the massive 32-thread processing power of the Ryzen 9 9950X with the unmatched memory bandwidth of the RTX 5090, it delivers seamless performance across deep learning, local LLM fine-tuning, and complex data pipeline analysis. Every component within the Cortex Starter is meticulously curated to eliminate hardware bottlenecks, slash model training times, and manage large-scale datasets—giving AI professionals and researchers an enterprise-grade development experience right at their desk.
AMD Ryzen 9 Power
A flagship, high-performance workstation CPU featuring 16 cores, 32 threads, and blistering boost speeds up to 5.7 GHz. Built on AMD's advanced Zen 5 architecture, it is engineered for elite multi-threaded workflows, heavy content creation, and local AI development. Delivering industry-leading throughput for both complex compilation tasks and intensive single-threaded applications, it is the ultimate powerhouse for developers, creators, and power users alike.
420mm Liquid Cooling
An elite, ultra-performance all-in-one liquid CPU cooler featuring a massive 420mm radiator engineered for maximum heat dissipation. Explicitly designed to tame high-TDP flagship processors under relentless multi-threaded workloads, it pairs industrial-grade thermal regulation with whisper-quiet operation. Offering unrivaled reliability and cooling capacity, it is the definitive choice for intense AI model training, heavy content creation, and continuous production-class rendering pipelines.
AMD x870 Motherboard
A high-tier, next-generation foundation engineered for flagship AMD Ryzen processors, featuring an ultra-robust digital twin power delivery system and native PCIe 5.0 support across both graphics and storage lanes. Designed to anchor intense AI workloads, it offers advanced high-speed connectivity, including dual USB4 ports and cutting-edge Wi-Fi 7 networking. With multiple optimized M.2 slots for rapid dataset access and bulletproof structural stability, it provides the elite data throughput and reliability required for production-class computing, heavy development pipelines, and multi-GPU expansion.
High Speed Memory
A premium 64GB pool of ultra-fast DDR5 memory, engineered for next-generation speed, maximum data throughput, and flawless multi-threaded execution. Built to eliminate memory bottlenecks during intense local computing, model training, and heavy multitasking, it ensures your entire development pipeline remains blazingly fast and perfectly stable.
Lightning-Fast Storage
An ultra-capacity 4TB Samsung 990 Pro M.2 NVMe SSD delivering blistering PCIe 4.0 read/write speeds up to 7,450/6,900 MB/s for near-instantaneous boot times, fluid application execution, and rapid system responsiveness. Engineered to handle intense workloads, it provides the massive, high-throughput storage pipeline needed to cache heavy datasets, store expansive machine learning models, and access complex production files with absolute reliability and peak sustained performance.
NVIDIA GeForce RTX 5090
The definitive pinnacle of visual computing, the NVIDIA GeForce RTX 5090 features a massive 32GB of ultra-fast GDDR7 memory, a colossal 1,700+ GB/s of bandwidth, and next-generation Blackwell Tensor Cores. Engineered to completely bypass cloud latency, it runs dense, production-class open models at blistering token-generation speeds while delivering uncompromising, real-time cinematic ray tracing and high-frame-rate execution for the most intensive compute and creative pipelines.
1200W 80+ Titanium ATX 3.1 PSU
An ultra-premium, flagship power infrastructure delivering a massive 1200 watts of continuous, rock-solid power backed by an elite 80 Plus Titanium certification for the highest possible energy efficiency. Meticulously engineered to sustain high-draw, multi-GPU configurations and extreme transient loads, it features native ATX 3.1 compliance for flawless next-gen power delivery. Built with a fully modular design to optimize internal airflow, it guarantees near-silent operation and industrial-grade durability under relentless AI training loops and heavy compute workloads.
AI-Ready Desktop PCs in the UK Precision Built for Advanced AI Performance
Our AI-Ready Desktop PCs in the UK are crafted for professionals and researchers who need reliable, high-performance systems to tackle advanced AI software. Featuring powerful GPUs, multi-core processors, and ample high-speed memory, these desktops are optimized to run machine learning models, deep learning applications, and large-scale data analysis seamlessly. Each AI desktop is precision-built to ensure efficient processing, quick training times, and smooth handling of complex computations, making it the ideal choice for those pushing the boundaries of artificial intelligence.
Technical Specifications
Specifications
Processor (CPU)
16 Cores / 32 Threads
Up to 5.7 GHz Boost
Graphics Card
32GB GDDR7 VRAM
Next-Gen Blackwell Architecture
Memory (RAM)
High-Speed Performance
Optimized for AI Workloads
CPU Cooler
High-Performance AIO Liquid Cooler
Massive 420mm Radiator
Power Supply
Fully Modular | 80+ Titanium
ATX 3.1 Native / PCIe 5.1 Ready
Motherboard
AMD Socket AM5 Platform
PCIe 5.0 & Next-Gen Connectivity
1st M.2 SSD (System)
Flagship PCIe Gen4 NVMe SSD
Up to 7450MB/s Read Speeds
2nd M.2 SSD (Storage)
Operating System
64-Bit Edition
AI Features Ready
Case
Premium Mid-Tower ATX
High-Airflow Mesh Optimization
Connectivity
High-Speed Ethernet
USB4 / Type-A & Type-C I/O
Warranty
1 Year Collect & Return
1 Year Parts, 3 Year Labour
Frequently Asked Questions
Yes. The Cortex AI is designed to run many popular AI tools and local AI models directly on the system.
Yes. The RTX 5060 Ti 16GB and 64GB RAM make it suitable for running Ollama and a wide range of local AI models.
Yes. The Cortex AI is designed for AI image generation and works well with Stable Diffusion and ComfyUI workflows.
Yes, many business-grade Lenovo laptops offer better upgradeability than consumer ultrabooks. Models like the ThinkPad T14 (Gen 5 and later) have moved back toward modular designs, allowing for easier RAM and SSD upgrades. Always check the specific MPN (Manufacturer Part Number) on our product pages to confirm if the memory is "soldered" or "slotted."
The Cortex AI is designed for AI learning, development, research and small business AI projects. Large-scale enterprise model training may require more specialised workstation or server hardware.
