Compile & Deploy VanitySearch with CUDA (RTX 5090) on Windows 11 VPS

Job ID: 40255337

Budget: $250 – $750 USD

I am looking for an experienced C++ / CUDA developer to compile and deploy the open-source project **VanitySearch** on my Windows 11 VPS (RDP environment) with full GPU acceleration.

The VPS is equipped with an **RTX 5090 GPU**, and I require proper CUDA configuration and optimization for this architecture.

---

Scope of Work

1. Clone the VanitySearch GitHub repository.
2. Compile in **Release mode** using Visual Studio.
3. Configure and enable CUDA acceleration for RTX 5090.
4. Resolve any:

* CUDA version mismatches
* Driver compatibility issues
* Compute capability configuration issues
* Project (.vcxproj) adjustments if required
5. Test the executable to confirm GPU acceleration is functioning properly.
6. Deploy and verify operation on my Windows 11 VPS.
7. Provide brief documentation or screen recording of the setup process.

---

Environment

* Windows 11 (RDP VPS)
* RTX 5090 GPU
* Latest NVIDIA drivers (if updates required, include in scope)
* Visual Studio (latest stable)
* CUDA Toolkit (likely CUDA 12+)

---

Required Skills

* Strong C++ experience
* Visual Studio project configuration (.sln, .vcxproj)
* NVIDIA CUDA Toolkit (modern versions)
* Experience with new-generation RTX GPUs
* Understanding of compute capability configuration
* Windows system configuration and troubleshooting

---

Important

RTX 5090 is a modern GPU. You must be comfortable:

* Adjusting compute capability flags
* Configuring CUDA 12+
* Troubleshooting architecture compatibility issues
* Optimizing build configuration for GPU acceleration

If you have previously compiled GPU-based cryptographic or key-search tools, that is a strong advantage.

---

Deliverables

* Working compiled executable
* Verified GPU acceleration
* Clean deployment on VPS
* Brief setup documentation

---

Budget

Fixed Price: $200 – $350

Open to hourly ($50–$80/hour, 4–6 hour cap)

---

Quick Screening Question

Please briefly answer in application:

1. What CUDA compute capability would you target for an RTX 5090?
2. How would you verify that GPU acceleration is actually being used after compilation?

Proposals without this answer will be ignored.