CUDA Developer Needed – GPU-Accelerated Bitcoin WIF Key Recovery Tool (Verification Required) -- 2

Job ID: 40192112

Budget: €30 – €250 EUR

Job Title:
CUDA Developer Needed – GPU-Accelerated Bitcoin WIF Key Recovery Tool (Verification Required)


Project Description:

I am looking for an experienced CUDA / GPU developer to build and optimize a high-performance Bitcoin WIF private key recovery program.

This project requires real GPU computation, correct Bitcoin cryptography handling, and verifiable results. This is not a demo or theoretical project. The program must be fully functional and tested.

Only apply if you have proven experience with CUDA, cryptography, or Bitcoin key handling.


Technical Requirements:

- Written in C++ with CUDA
- Runs on NVIDIA GPUs
- Command-line interface (CLI)
- Supports Bitcoin WIF (Base58Check)
- Supports compressed and uncompressed private keys
- Correct checksum validation
- Generates private keys, converts to WIF, derives public key and Bitcoin address
- Compares generated address with target address
- Stops immediately when a valid match is found
- Outputs:
- Matching private key (WIF)
- Corresponding Bitcoin address
- Performance statistics (keys per second, elapsed time)


Performance Requirements:

- Must use GPU (CPU-only solutions will be rejected)
- GPU usage must be verifiable (nvidia-smi or similar)
- Reasonable performance optimization expected
- Clean, readable, well-structured source code


Mandatory Verification Requirement (NO EXCEPTIONS):

Before any payment is released, the program must be proven to work using a known test private key.

The developer must successfully recover the following dummy private key:

KEYL5.txt + [7 unknown Base58 characters] XXXXXXX + 4EFT7Udr2sXCsi1tUiFRHvBU5srrapgrsUXZx98B

Target Bitcoin address 13fCqZ1Xzkv4vnUvRjw7YkYB6DLgMrFSkC

WIF Key Structure:
L5atq + [7 unknown Base58 characters] + [known 40-character suffix]

- L5atq: A 5-character Base58 prefix, sourced from a file containing ~164,000 known good prefix values
- 7-character section: Unknown, to be brute-forced using CUDA
- Final 40 characters: Fixed, the same for all candidates

The full 52-character WIF is validated, decoded, converted to a Bitcoin address, and matched against a known target address.

How the Brute-Force Works:
- The tool reads batches of 64 prefix lines at a time from the input file (each 5 characters long)
- For each prefix, the kernel:
- Brute-forces all 58^7 possible 7-character Base58 combinations
- Appends the known 40-character suffix
- Constructs a candidate WIF key
- Each candidate key is:
- Decoded via Base58Check (with checksum)
- Converted to a private key
- Used to generate a public key (this is your part)
- Used to derive a Bitcoin address
- Matched against the user-provided Bitcoin address
- If a match is found:
- The program immediately halts all threads
- The valid private key (in WIF) is printed and saved


Verification must be done by:
- Screen recording of the program running and finding the key
OR
- Full console logs and output files clearly showing correct recovery

Hardcoded outputs, simulated logic, mocked results, or fake success messages will be considered non-delivery.


Milestone Structure:

Milestone 1:
- Code compiles and runs
- Accepts CLI parameters
- CUDA kernel launches correctly

Milestone 2 (Critical):
- Successfully recovers the provided test private key
- Results verified by me

Milestone 3 (Final):
- Performance optimization
- Code cleanup and documentation

Payment is released only after milestone verification.


Ownership and Confidentiality:

- All source code, binaries, and related materials become my exclusive property after final payment
- Code must not be reused, resold, shared, or published
- No restrictive third-party licenses


Who Should Apply:

- CUDA developers
- GPU optimization specialists
- Developers with Bitcoin or cryptography experience

Do NOT apply if:
- You plan to simulate results
- You cannot provide real GPU proof
- You lack CUDA experience


How to Apply:

Include in your bid:
- Relevant CUDA / GPU experience
- Examples of similar work (if any)
- Your GPU hardware model

Generic or copy-paste bids will be ignored.