Description
Sirius Pro AI Ultra 6096 RTX PRO 6000 96GB AI Workstation
96GB Professional Blackwell Graphics. 24-Core Intel Processing. Built for Local AI.
The RTX PRO 6000 96GB AI Workstation is designed for professionals whose workloads exceed the practical limits of conventional gaming PCs and consumer graphics cards.
The defining component is the NVIDIA RTX PRO 6000 Blackwell Workstation Edition with 96GB ECC GDDR7 memory.
That enormous local GPU-memory capacity gives AI engineers, developers, researchers, 3D artists, VFX professionals, architects, engineers, and technical users substantially more room for:
- Large language models
- Local AI inference
- Fine-tuning workflows
- Computer vision
- Generative AI
- Large 3D scenes
- Professional visualization
- Unreal Engine
- High-resolution video
- CUDA workloads
- Large GPU-resident datasets
CPU processing comes from the Intel Core Ultra 9 285K, providing 24 total cores across 8 Performance-cores and 16 Efficient-cores.
The rest of the workstation is built around professional connectivity and sustained operation with 64GB DDR5-6000 memory, a Crucial T705 2TB PCIe 5.0 NVMe SSD, ASUS ProArt Z890-CREATOR WIFI motherboard, Sirius 360mm AIO cooling, nine case fans, 1200W power, 10Gb networking, Wi-Fi 7, and Thunderbolt 5.
This is not a gaming PC with a workstation graphics card installed.
It is a professional single-GPU workstation designed around large local GPU memory, sustained compute, fast storage, high-speed networking, and professional expansion.
96GB ECC graphics. Local Blackwell compute. Thunderbolt 5. 10Gb networking. Sirius Built.
Sirius Pro AI Ultra 6096 Specifications
| Component | Specification | Professional Role |
|---|---|---|
| CPU | Intel Core Ultra 9 285K – 24 Cores / 24 Threads | Compilation, preprocessing, encoding, application logic, development, and heavy multitasking |
| CPU Cooler | Sirius 360mm AIO Liquid Cooler | Cooling capacity for sustained high-power CPU workloads |
| Motherboard | ASUS ProArt Z890-CREATOR WIFI | Thunderbolt 5, 10GbE, 2.5GbE, Wi-Fi 7, PCIe 5.0, DDR5, and extensive NVMe expansion |
| Memory | 64GB DDR5-6000 CL36 | AI development, editing, 3D work, virtual machines, development, and multitasking |
| Storage | Crucial T705 2TB PCIe 5.0 NVMe SSD | High-speed storage for applications, active models, cache, datasets, and projects |
| GPU | NVIDIA RTX PRO 6000 Blackwell Workstation Edition – 96GB ECC GDDR7 | AI, CUDA, rendering, visualization, simulation, professional graphics, and large GPU-memory workloads |
| Networking | 10GbE + 2.5GbE + Wi-Fi 7 | High-speed NAS, server, dataset, media, and project connectivity |
| External Connectivity | Thunderbolt 5 / USB4 | High-speed storage, docks, displays, capture hardware, and professional peripherals |
| Case | Black Knight Satellite PC Case | Premium serviceable Sirius workstation chassis |
| Case Cooling | 9 Case Fans | Sustained airflow around the GPU, radiator, storage, motherboard, and power hardware |
| Power Supply | 1200W PSU | High-capacity power for the professional GPU, CPU, cooling, storage, and expansion |
| Support | 3-Year Labor Warranty + Lifetime Technical Support | Long-term Sirius workstation support |
| Shipping | Insured Shipping + Wooden Crate Protection | Added protection for high-value professional hardware |
96GB ECC VRAM Is the Main Reason This Workstation Exists
The most important specification in the Sirius Pro AI Ultra 6096 is not RGB lighting, CPU frequency, or even benchmark FPS.
It is 96GB of dedicated GPU memory.
Professional GPU workloads are often limited by whether the active:
- Model
- Dataset
- Scene
- Texture set
- Frame buffer
- KV cache
- Activations
- Geometry
- Simulation data
can remain inside GPU memory.
When they cannot, software may be forced to:
- Quantize more aggressively
- Reduce batch size
- Reduce context length
- Split scenes
- Lower texture resolution
- Offload data to system RAM
- Process workloads in stages
- Fall back to slower CPU-side processing
The RTX PRO 6000 substantially raises that ceiling.
The result is not simply a faster GPU.
It is access to a different class of workload.
Built for Local AI
The Sirius Pro AI Ultra 6096 is particularly suited to professionals who want substantial AI capability directly on their desk.
Potential workloads include:
- Local LLM inference
- Retrieval-augmented generation
- Agentic AI
- Embedding generation
- Model evaluation
- Fine-tuning
- LoRA and QLoRA
- Computer vision
- Generative image models
- Generative video
- Vision-language models
- Speech processing
- CUDA development
- AI application development
Local processing can also be useful when working with:
- Proprietary source code
- Internal documents
- Confidential business data
- Research datasets
- Client assets
- Unreleased creative material
- Intellectual property
- Large files that are inconvenient to move repeatedly to cloud services
Local execution does not automatically satisfy every security or compliance requirement.
It does give organizations control over where their compute is performed.
Large Language Model Workloads
The 96GB graphics-memory pool makes this workstation particularly attractive for local LLM development and inference.
Potential workflows include:
- Local inference
- Quantized models
- RAG systems
- Code models
- Long-context experimentation
- AI agents
- Embedding models
- Model evaluation
- Multiple models in one pipeline
- Adapter-based fine-tuning
The model that fits cannot be described accurately with one universal parameter count.
Memory requirements depend on:
- Parameter count
- Numerical precision
- Quantization
- Context length
- KV cache
- Batch size
- Framework overhead
- Additional loaded models
The appropriate claim is therefore:
96GB of local VRAM gives significantly more room before graphics memory becomes the limiting factor.
AI Development & CUDA
The RTX PRO 6000 provides a professional NVIDIA platform for GPU-accelerated development.
Potential environments include:
- PyTorch
- TensorFlow
- JAX
- ONNX Runtime
- CUDA
- Computer-vision frameworks
- Generative-AI pipelines
- Data-science applications
- GPU research tools
The Core Ultra processor handles the rest of the development environment while GPU-intensive compute runs on the RTX PRO card.
That can include:
- IDEs
- Browsers
- Documentation
- Containers
- Databases
- Preprocessing
- Compilation
- Local services
- Monitoring tools
This makes the Sirius Pro AI Ultra 6096 useful not only for running models, but for building the applications around them.
Fine-Tuning & Model Adaptation
The 96GB GPU can also support model-adaptation workflows including:
- LoRA
- QLoRA
- Adapter training
- Domain-specific fine-tuning
- Evaluation
- Dataset experimentation
- Embedding-model development
Training generally requires substantially more memory than inference because memory may also be needed for:
- Activations
- Gradients
- Optimizer states
- Training batches
The 96GB framebuffer dramatically expands what can be attempted on one desktop GPU.
It does not turn every training job into a single-GPU workload.
Very large training projects can still require multi-GPU or server infrastructure.
Generative Image & Video AI
Large VRAM is also useful for modern generative-media pipelines.
The workstation can be used for:
- High-resolution image generation
- Batch generation
- Image-to-image workflows
- Inpainting
- Control models
- Multiple-model pipelines
- AI upscaling
- Video generation
- Image-to-video
- Frame interpolation
- AI-assisted video effects
Complex workflows may load several models or pipeline components simultaneously.
The 96GB memory pool provides substantially more freedom before those components have to be moved between GPU and system memory.
Computer Vision
The Sirius Pro AI Ultra 6096 can also serve as a powerful computer-vision workstation.
Potential applications include:
- Object detection
- Segmentation
- Tracking
- Classification
- OCR
- Video analytics
- Industrial inspection
- Medical-imaging research
- Geospatial processing
- Multi-camera analysis
- Vision-language systems
Large VRAM can support:
- Higher-resolution inputs
- Larger batches
- Larger models
- Multiple video streams
- More complex pipelines
ECC GPU Memory
The RTX PRO 6000 uses ECC-capable GDDR7 memory.
ECC can detect and correct certain memory errors.
That capability becomes particularly relevant when a GPU is running:
- Overnight renders
- Long AI inference sessions
- Research workloads
- Engineering calculations
- Scientific processing
- Production exports
- Large visualization jobs
ECC does not prevent every software failure or system error.
It provides an additional layer of memory integrity appropriate for professional workloads.
Professional Blackwell GPU Platform
The RTX PRO 6000 is built around NVIDIA’s professional Blackwell architecture.
Its capabilities are aimed at demanding workloads involving:
- Artificial intelligence
- GPU compute
- Professional rendering
- Visualization
- Ray tracing
- Virtual production
- Video processing
- Scientific workloads
- Engineering applications
The workstation edition also provides a professional software and driver ecosystem rather than being designed primarily around gaming.
That difference matters when software stability and professional application support are part of the purchasing decision.
Core Ultra 9 285K – 24 CPU Cores Behind the GPU
The RTX PRO 6000 defines the workstation, but the CPU still matters.
The Intel Core Ultra 9 285K provides:
- 24 total cores
- 8 Performance-cores
- 16 Efficient-cores
- 24 threads
- High boost performance
- Large CPU cache
- Modern DDR5 platform support
The CPU handles workloads such as:
- Data preparation
- Compilation
- Encoding
- Decompression
- File processing
- Asset conversion
- Application logic
- Development tools
- Background services
- Multitasking
The GPU accelerates supported compute.
The CPU keeps everything around that compute moving.
Hybrid Intel Architecture for Heavy Multitasking
The Core Ultra 9 combines Performance-cores with Efficient-cores.
That gives the workstation more resources when applications such as:
- Python
- Visual Studio
- Docker
- Browsers
- Databases
- File preprocessing
- Monitoring software
- Communication tools
are running alongside GPU-intensive workloads.
A professional AI system rarely runs one application in isolation.
The CPU configuration reflects that reality.
ASUS ProArt Z890-CREATOR WIFI
The ASUS ProArt Z890-CREATOR WIFI is one of the most important supporting components in this build.
The workstation platform provides professional connectivity for:
- High-speed storage
- Network storage
- Project servers
- External arrays
- Displays
- Capture hardware
- Production peripherals
- Internal expansion
Key platform capabilities include:
- Thunderbolt 5
- 10Gb Ethernet
- 2.5Gb Ethernet
- Wi-Fi 7
- PCIe 5.0
- DDR5
- Five M.2 storage positions
This is the type of motherboard that makes sense when connectivity is part of the workflow rather than an afterthought.
Thunderbolt 5
Thunderbolt 5 provides high-bandwidth connectivity for compatible:
- External NVMe storage
- Professional docks
- Storage arrays
- Displays
- Capture systems
- Audio hardware
- Creator peripherals
For an ordinary gaming system, Thunderbolt may be optional.
For a professional workstation with large datasets and external production hardware, it can be a meaningful feature.
10Gb Ethernet
The integrated 10Gb Ethernet connection makes the system suitable for professional network environments.
Potential uses include:
- NAS storage
- Central datasets
- Editing servers
- Model repositories
- Shared project storage
- Render assets
- High-speed backups
Large AI, video, and 3D projects can involve hundreds of gigabytes or multiple terabytes of data.
Moving those files across conventional Gigabit Ethernet can become a workflow bottleneck.
The workstation already provides the interface required for a 10GbE environment.
2.5Gb Ethernet
A second 2.5Gb Ethernet connection adds networking flexibility.
It can be useful for:
- Dedicated internet access
- Secondary networks
- Network segmentation
- Management networks
- Backup connectivity
Professional networking is one of the key differences between this workstation and a conventional high-end gaming PC.
Wi-Fi 7
Built-in Wi-Fi 7 provides modern wireless connectivity for environments where wired networking is not practical.
For large professional file transfers, the 10GbE connection remains the more predictable solution when compatible network infrastructure is available.
64GB DDR5-6000 System Memory
The workstation starts with 64GB DDR5-6000 system memory.
That is a strong baseline for:
- AI development
- Dataset preparation
- Blender
- Unreal Engine
- CAD
- Video production
- Virtual machines
- Software development
- Heavy multitasking
GPU VRAM and system RAM perform different jobs.
A GPU having 96GB of VRAM does not mean the workstation no longer needs substantial system memory.
Applications still use RAM for:
- Preprocessing
- CPU-side data
- Application state
- Development environments
- File caching
- Virtual machines
- Background services
When 128GB Makes Sense
Consider increasing system memory for:
- Large datasets
- Multiple virtual machines
- Heavy AI preprocessing
- Unreal Engine
- After Effects
- Large CAD assemblies
- Photogrammetry
- Scientific applications
- Multiple professional applications
- Large development environments
More RAM does not automatically make the GPU faster.
It prevents system-memory pressure when the applications surrounding the GPU require more capacity.
Crucial T705 2TB PCIe 5.0 NVMe SSD
The Crucial T705 2TB PCIe 5.0 NVMe SSD provides high-speed primary local storage.
Its role extends well beyond fast Windows startup.
The Gen5 drive can hold:
- Operating system
- AI frameworks
- Applications
- Active models
- Current datasets
- Development environments
- Project files
- Cache
- Temporary renders
- Build files
High-speed storage becomes valuable when professional workloads are actually moving large amounts of data.
Five M.2 Slots for Storage Expansion
The ProArt platform provides extensive NVMe expansion.
That means this workstation does not need to remain a single-drive system.
A professional storage layout could use:
Drive 1 – Operating System & Applications
Windows or Linux, drivers, frameworks, development tools, and professional software.
Drive 2 – AI Models & Active Data
Models, datasets, embeddings, project data, and active assets.
Drive 3 – Scratch & Cache
Video cache, render cache, temporary datasets, intermediate files, and simulation cache.
Drive 4 – Project Storage
Unreal Engine, Blender assets, video projects, CAD files, or software repositories.
Network Storage
Completed projects, team datasets, archives, and backups can move to 10GbE-connected NAS or server storage.
That is a storage architecture designed around professional workflows rather than simply installing the largest possible C: drive.
Blender & Professional 3D Rendering
The Sirius Pro AI Ultra 6096 is also a serious professional rendering system.
Potential applications include:
- Blender
- Maya
- 3ds Max
- Cinema 4D
- V-Ray
- OctaneRender
- Redshift
- Omniverse
- Unreal Engine
The 96GB framebuffer provides room for:
- Large geometry
- High-resolution textures
- Complex environments
- Large visualization scenes
- Extensive lighting data
- High-resolution assets
When a scene exceeds available GPU memory, artists may otherwise have to simplify it or break it apart.
The RTX PRO 6000 substantially increases the point at which that becomes necessary.
Unreal Engine & Virtual Production
The workstation is exceptionally well suited to Unreal Engine and real-time visualization.
The platform combines:
- 24 CPU cores
- 96GB GPU memory
- 64GB system memory
- PCIe 5.0 storage
- Thunderbolt 5
- 10Gb networking
That makes it useful for:
- Large Unreal environments
- Nanite-heavy projects
- High-resolution textures
- Ray tracing
- Virtual production
- Digital twins
- Shader compilation
- Asset preparation
- Large project files
- Real-time visualization
Very large projects may justify additional system memory and dedicated project storage.
CAD, Architecture & Engineering
The professional RTX platform also makes the Sirius Pro AI Ultra 6096 appropriate for demanding design and engineering work.
Potential applications include:
- AutoCAD
- Revit
- SolidWorks
- CATIA
- Creo
- Siemens NX
- Rhino
- BIM
- Digital twins
- Large assemblies
- Point-cloud visualization
Exact certification should always be checked for the application and driver version being deployed.
The workstation is designed to provide the memory capacity, connectivity, and professional GPU environment these applications often require.
Video Production & VFX
The system is also built for demanding media workflows.
Potential applications include:
- DaVinci Resolve
- Adobe Premiere Pro
- After Effects
- Fusion
- Nuke
- Media Encoder
The 96GB professional GPU provides enormous graphics-memory headroom for:
- High-resolution timelines
- Large frame buffers
- GPU effects
- Noise reduction
- AI-assisted video features
- Compositing
- Multiple high-resolution streams
- Complex production workloads
The 64GB system-memory configuration provides a strong starting point, while heavier VFX and After Effects environments may justify more RAM.
Four Hardware Encoders
The RTX PRO 6000 provides multiple dedicated hardware video encoders.
That can be useful for:
- Transcoding
- Video processing
- Streaming
- Multi-output production
- High-throughput media workflows
This is another example of why the GPU should be evaluated as a professional compute and media device rather than simply by gaming performance.
Data Science & Scientific Computing
The Sirius Pro AI Ultra 6096 can support GPU-accelerated research and analytical workflows including:
- Data science
- RAPIDS
- Machine learning
- Scientific visualization
- Medical imaging
- Geospatial processing
- Engineering analysis
- Signal processing
- Custom CUDA applications
Not every scientific workload is primarily GPU-bound.
Some depend more heavily on:
- CPU core count
- Double-precision performance
- System RAM
- Memory bandwidth
- Specialized libraries
- Distributed infrastructure
The correct workstation always depends on the actual application.
1200W Power Supply
A 1200W PSU makes sense in this workstation because the hardware itself can consume substantial power.
The power supply must support:
- RTX PRO 6000
- Core Ultra 9 285K
- Sirius 360mm AIO
- Nine case fans
- ProArt motherboard
- DDR5 memory
- Gen5 storage
- Expansion hardware
- External peripherals
For this class of system, PSU quality matters as much as rated wattage.
The final production unit should provide:
- Appropriate efficiency
- Native high-current GPU connectivity
- Strong transient handling
- Quality internal components
- Correct connector capacity
UPS Recommended
A properly sized UPS is strongly recommended for a workstation at this level.
It can provide:
- Temporary battery power
- Time to save active work
- Controlled shutdown
- Protection against brief outages
- Additional protection from unstable power
That matters during:
- Long AI runs
- Overnight rendering
- Video exports
- Compilation
- Simulation
- Large file transfers
UPS sizing should be based on real system load plus monitors, network equipment, and other protected devices.
Sirius 360mm AIO Cooling
The Core Ultra 9 285K can sustain substantial CPU power under demanding workloads.
The Sirius 360mm AIO provides cooling capacity for:
- Compilation
- Data preparation
- Encoding
- CPU rendering
- Simulation preparation
- Development
- Heavy multitasking
Professional cooling must be evaluated over long workloads, not a short benchmark burst.
Nine-Fan Black Knight Satellite Airflow
The Black Knight Satellite chassis includes nine case fans to support the complete thermal environment.
Airflow helps manage heat from:
- RTX PRO 6000
- CPU radiator
- Motherboard
- DDR5 memory
- Gen5 SSD
- Power-delivery components
- Additional drives
Professional workloads may keep the CPU and GPU under load for hours.
The cooling system is built around sustained operation.
Windows or Linux?
The base product should list an operating system only if one is actually included in the final configuration.
Windows 11 Pro
A strong fit for:
- Adobe applications
- CAD
- Professional creator software
- Business environments
- General workstation use
- Thunderbolt-based workflows
Linux
A strong fit for:
- CUDA development
- AI frameworks
- Data science
- Scientific applications
- Containers
- Research environments
- Software development
Hybrid Environments
Developers may also use:
- WSL
- Virtual machines
- Containers
- Dual boot
- Remote Linux systems
The software stack should determine the operating system.
Who Should Buy Sirius Pro AI Ultra 6096?
This workstation is built for:
AI Engineers
Developers who need substantial local VRAM for inference, evaluation, fine-tuning, computer vision, and AI experimentation.
Machine Learning Engineers
Professionals building and testing CUDA-accelerated AI applications.
Researchers
Users working with local GPU-accelerated scientific, analytical, and research workloads.
3D Artists
Professionals whose scenes and textures exceed consumer-GPU memory.
VFX & Video Professionals
Editors, compositors, colorists, and production teams working with high-resolution media.
Architects & Engineers
Users running CAD, BIM, large assemblies, professional visualization, and digital twins.
Unreal Engine Developers
Professionals working with large environments, virtual production, real-time visualization, and game development.
Organizations Requiring Local AI
Teams that want substantial GPU capability on local infrastructure rather than depending entirely on cloud GPUs.
Who Should Choose Another Sirius System?
Primarily Gaming
Choose a GeForce RTX-based Sirius gaming PC.
The majority of this workstation’s cost and capability would be wasted on a gaming-only workload.
Need More CPU Cores
Choose a Threadripper or Threadripper PRO workstation for highly parallel CPU workloads.
Need Very Large System-Memory Capacity
A Threadripper PRO or server-class platform may be more appropriate.
Need Multiple Professional GPUs
A platform with more CPU-connected PCIe lanes and greater chassis capacity should be considered.
Need Rack Deployment
Choose a dedicated rack workstation or server.
Need One 96GB Professional GPU in a Desktop
That is exactly where the Sirius Pro AI Ultra 6096 belongs.
Sirius Workstation Validation
A professional workstation at this level should not be tested like an ordinary office PC.
Every Sirius Pro AI Ultra 6096 includes validation of the complete platform.
That includes:
- Professional Sirius assembly
- BIOS configuration
- Firmware setup
- Memory configuration
- NVMe installation
- Driver configuration
- Cable management
- Cooling configuration
- CPU stress testing
- GPU stress testing
- Combined CPU/GPU load testing
- Memory stability testing
- Storage validation
- Thermal verification
- Network verification
- System stability testing
- 3-Year Labor Warranty
- Applicable manufacturer warranties
- Lifetime Technical Support
- Insured shipping
- Wooden crate protection
The goal is not merely to verify that the system boots.
The goal is to verify that the completed workstation can sustain the kind of workloads its hardware was purchased to run.
Frequently Asked Questions
What is Sirius Pro AI Ultra 6096 built for?
It is designed primarily for local AI, machine learning, professional rendering, Unreal Engine, CAD, VFX, video production, CUDA development, data science, visualization, and other GPU-intensive professional workloads.
What GPU does Sirius Pro AI Ultra 6096 use?
It uses the NVIDIA RTX PRO 6000 Blackwell Workstation Edition.
How much VRAM does it have?
The GPU includes 96GB of ECC GDDR7 memory.
Why is 96GB VRAM important?
Large GPU-memory capacity allows larger models, scenes, textures, datasets, frame buffers, and compute workloads to remain directly accessible to the GPU.
Does the RTX PRO 6000 use ECC memory?
Yes. It uses ECC-capable GDDR7 memory.
Is Sirius Pro AI Ultra 6096 good for local LLMs?
Yes. Large local AI and LLM workloads are among its primary use cases. Exact model compatibility depends on precision, quantization, context length, KV cache, batch size, and software overhead.
Can it fine-tune AI models?
Yes. It can support fine-tuning and adapter-based workflows. Very large training jobs may still require multi-GPU infrastructure.
Is it good for generative AI?
Yes. The large GPU-memory pool is particularly useful for image generation, video generation, multimodal pipelines, and other memory-intensive generative workflows.
Is it good for machine learning?
Yes. The professional NVIDIA GPU is built for CUDA and GPU-accelerated AI workloads.
What CPU does it use?
The workstation uses the Intel Core Ultra 9 285K.
How many cores does the Core Ultra 9 285K have?
It has 24 CPU cores: 8 Performance-cores and 16 Efficient-cores.
How much system RAM comes standard?
The configuration includes 64GB DDR5-6000 memory.
Should I upgrade to 128GB?
128GB can make sense for large datasets, virtual machines, Unreal Engine, VFX, heavy AI preprocessing, and other memory-intensive professional workloads.
What SSD does it include?
It includes a Crucial T705 2TB PCIe 5.0 NVMe SSD.
Can I add more NVMe storage?
Yes. The ProArt motherboard provides extensive M.2 expansion.
Does it have Thunderbolt 5?
Yes. The ProArt Z890-CREATOR WIFI provides Thunderbolt 5 connectivity.
Does it have 10Gb Ethernet?
Yes. It includes 10GbE and 2.5GbE networking.
Does it have Wi-Fi 7?
Yes.
Is it good for Blender?
Yes. The RTX PRO 6000’s professional GPU architecture and 96GB framebuffer make it extremely well suited to large compatible Blender scenes and GPU-rendering workloads.
Is it good for Unreal Engine?
Yes. The combination of a high-core-count CPU, 96GB GPU memory, system RAM, fast NVMe storage, and professional networking makes it a strong Unreal Engine workstation.
Is it good for DaVinci Resolve?
Yes. Its professional GPU and large framebuffer make it suitable for demanding GPU-heavy Resolve environments.
Is it good for Adobe Premiere Pro?
Yes. The workstation provides a high-core-count CPU, professional NVIDIA GPU, DDR5 memory, and fast storage for demanding video-production workflows.
Is it good for CAD?
Yes. It is built for professional graphics and visualization. Exact software certification should be checked against the application and driver version being deployed.
Is it good for scientific computing?
It can be an excellent platform for CUDA-accelerated technical workloads. Some scientific applications may require specialized CPU, memory, precision, or server hardware.
Is this a gaming PC?
No. It can run games, but gaming is not its primary purpose. A GeForce-based Sirius system provides better value for gaming-first customers.
Why does it use a 1200W PSU?
The workstation contains a very high-power professional GPU, high-end Intel processor, liquid cooling, multiple fans, Gen5 storage, and professional expansion hardware.
Should I use a UPS?
Yes. A properly sized UPS is strongly recommended for a high-value professional workstation.
Does Windows come standard?
The supplied configuration does not specify an operating system. Windows or Linux should be listed only when it is actually included with the final product.
What warranty does it include?
The workstation includes a 3-Year Sirius Labor Warranty, applicable manufacturer warranties, and Lifetime Technical Support.
Who should buy Sirius Pro AI Ultra 6096?
It is best for AI engineers, machine-learning professionals, researchers, data scientists, CUDA developers, 3D artists, VFX professionals, architects, engineers, Unreal Engine developers, and organizations that need 96GB of professional local GPU memory.
Final Take
The Sirius Pro AI Ultra 6096 Workstation is built around a simple question:
How much of your workload can remain directly on the GPU?
With 96GB of ECC GDDR7, the NVIDIA RTX PRO 6000 dramatically increases the answer.
That GPU-memory capacity gives AI engineers, developers, researchers, artists, and technical professionals considerably more room for models, datasets, scenes, textures, frame buffers, and professional compute than conventional consumer graphics cards.
The Intel Core Ultra 9 285K provides 24 CPU cores for compilation, preprocessing, encoding, development, application logic, and heavy multitasking.
The 64GB DDR5-6000 memory provides a strong professional starting point for the CPU-side environment.
The Crucial T705 2TB PCIe 5.0 NVMe SSD provides a fast primary working drive, while the ProArt platform leaves substantial room for additional NVMe storage.
The ASUS ProArt Z890-CREATOR WIFI brings the connectivity a professional workstation needs, including Thunderbolt 5, 10GbE, 2.5GbE, Wi-Fi 7, PCIe 5.0, and extensive M.2 expansion.
The Sirius 360mm AIO and nine-fan Black Knight Satellite cooling system are designed around sustained professional workloads rather than short benchmark bursts.
And the 1200W PSU provides the power capacity required for a platform built around one of NVIDIA’s highest-power professional desktop GPUs.
This workstation is not about maximizing gaming FPS.
It is about removing GPU-memory limitations from serious professional work.
96GB ECC Blackwell Graphics. 24-Core Intel Processing. Thunderbolt 5. 10Gb Networking. Local AI. Sirius Built.

Reviews
There are no reviews yet.