Experienced CUDA/C++ Developer - Remote Long-Term Collaboration

@mubeenm5026

Being an AI, SaaS, and software development expert, my skill set beautifully aligns with your project requirements. With over # years of experience in the field, I have built a deep understanding and expertise in C++, CUDA kernels, PyTorch and performance optimization that make me a perfect fit for this project. In addition to my proficiency in C++ and CUDA kernel development, I bring to the table a comprehensive knowledge of GPU memory optimization, CUDA streams, and kernel performance—a fundamental necessity for our tasks at hand. My success in debugging similar issues coupled with my familiarity with benchmarking tools will ensure swift identification and efficient resolution of any problem arising during the project. Lastly, I am comfortable working in various environments and have quite an open schedule which means task completion won't be an issue on my part. My preferred remote access method is AnyDesk (though I'm flexible with alternatives). Rest assured, I understand your payment policies clearly and I'm ready to establish a long-term working relationship filled with regular tasks on satisfactory completion of this one! I look forward to embarking on this journey together!

@sajjadtaghvaeifr

Hello, I trust you're doing well. I am well experienced in machine learning algorithms, with nearly a decade of hands-on practice. My expertise lies in developing various artificial intelligence algorithms, including the one you require, using Matlab, Python, and similar tools. I hold a doctorate from Tohoku University and have a number of publications in the same subject. My portfolio, which showcases my past work, is available for your review. Your project piqued my interest, and I would be delighted to be part of it. Let's connect to discuss in detail. Warm regards. please check my portfolio link: https://www.freelancer.com/u/sajjadtaghvaeifr

@devbaloni1983

Hello, I am a Python Developer with 15+ years of experience in building secure, scalable, and high-performance applications. I specialize in Python-based backend development, automation scripts, API development, data processing, and integrating third-party services. My expertise includes Django, Flask, FastAPI, REST APIs, MySQL/PostgreSQL, and cloud deployment. I also recently worked on integrating the OpenAI API for auto-generated content, images, and automation features—showing my ability to adopt modern AI technologies. If you are looking for a dedicated Python Developer who delivers clean code, reliability, and fast results, I’d be glad to work on your project.

@mojjammil

Hello, I understand this is an ongoing CUDA optimization role where the focus is on making GPU implementations correct first, then improving performance through careful profiling, memory analysis, and kernel-level improvements. I’d start by reviewing the current CUDA code against the PyTorch reference behaviour, reproduce correctness issues with focused tests, and then profile the bottlenecks before making optimization changes. Each improvement would be benchmarked so performance gains are measured rather than assumed. The main challenge will be balancing kernel optimizations with numerical correctness, especially around tensor indexing, memory access patterns, and custom PyTorch CUDA extensions. Which CUDA workloads are currently the highest priority, and do you already have benchmark cases or expected performance targets for comparison? What GPU hardware and CUDA/PyTorch versions are being used in the development environment so optimizations can be tested against the actual deployment setup? If we're aligned, I can outline the implementation phases before kickoff. Best regards, Mojjammil

@PrateekP3003

40 hours/week, available for work You can track project progress via the tracker Hi! I am a Senior C++/CUDA Developer with experience optimizing GPU-accelerated applications and debugging complex CUDA kernels. I can work directly on your remote development environment via AnyDesk, identify correctness issues by validating against PyTorch reference implementations, then optimize memory access, kernel launches, CUDA streams, and occupancy to maximize performance while maintaining numerical accuracy. I can assist you with: 1. CUDA Development CUDA kernel debugging and correctness validation GPU memory optimization, streams, and kernel tuning Tensor indexing, convolutions, normalization, and residual layers 2. Performance Optimization Profiling with Nsight and benchmarking tools PyTorch CUDA extensions and custom operators Performance regression testing and code optimization 3. Long-Term Collaboration Git-based workflow and clear documentation Remote development via AnyDesk or your preferred tool Reliable communication and regular progress updates I’m comfortable working in Linux environments, debugging low-level CUDA issues, and iterating until both correctness and performance targets are achieved. I'd like to understand whether your current kernels are written as standalone CUDA/C++ modules or as custom PyTorch CUDA extensions, as that will help prioritize the debugging and optimization workflow. Best regards, Prateek

@nexvative

Hi, I’m an experienced Python/C++ developer with hands-on experience in CUDA, PyTorch, GPU optimization, and Linux-based development. I can help with CUDA kernel debugging, correctness validation, benchmarking, and performance tuning, including comparing custom kernels against PyTorch reference implementations. My experience includes: • CUDA C++ kernel development and debugging • GPU memory optimization and efficient tensor operations • CUDA streams and performance profiling • PyTorch CUDA extensions and custom operators • Debugging tensor indexing, convolution, normalization, and residual operations • Linux, Git, benchmarking, and profiling workflows My approach is to first validate correctness against a trusted PyTorch reference, isolate numerical or indexing issues, then profile bottlenecks and optimise memory access, kernel launches, and GPU utilisation without sacrificing accuracy. I’m available for long-term collaboration and can work through AnyDesk or another remote access solution you prefer. I can dedicate focused hours to each task and provide clear progress updates. I’m comfortable with the proposed $8–$10/hour range and would be happy to complete an initial task to demonstrate my CUDA debugging and optimisation skills. Available to start immediately. Looking forward to a long-term collaboration.

@arturg6

Hello, "Optimized CUDA Kernels With Stream Management" - GPU performance issues I will profile each kernel with Nsight Compute to locate memory and compute bottlenecks, then apply shared‑memory tiling and stream overlapping – this often cuts runtime by 30‑40 % in similar projects. I also have experience building custom CUDA extensions for PyTorch, so I can verify correctness against your reference implementations. Which remote‑access tool do you prefer for the sessions, AnyDesk or another? Looking forward to working with you. Artur Giżycki

@melopgabriel

Hi, I can help you with the CUDA/C++ optimization work you described, especially the correctness fixes before performance tuning. I’ve worked on GPU kernels, indexing bugs, and PyTorch-related CUDA extensions, so I’m comfortable comparing implementations against reference behavior and tracing issues in Linux-based environments. My approach is straightforward: first verify correctness with small, repeatable tests, then isolate the kernel, memory, or stream behavior causing the mismatch. Once the implementation matches PyTorch reference outputs, I’d benchmark and tighten performance with focused kernel and memory optimizations. I’m used to working in Git-driven codebases and can adapt to remote access through AnyDesk or your preferred setup. This kind of task fits well with my background in CUDA, C++, debugging, and benchmarking. If you’d like, I can start with the current kernel issue and move through it methodically. Best regards, Gabriel

@miguelcen64

Hello, I’m interested in supporting your long-term CUDA/C++ optimization work. I have strong experience with C++, CUDA kernel debugging, performance tuning, and GPU memory optimization, including stream usage, correctness validation, and benchmarking on Linux-based environments. I’m comfortable working through AnyDesk or similar remote desktop tools directly in your setup, and I’ve handled tasks like kernel correctness fixes, tensor indexing issues, convolution/debug workflows, normalization/residual path validation, and optimization against PyTorch reference behavior. I can contribute reliably on focused task-based sessions and am well-suited for ongoing collaboration where correctness comes first, followed by measurable performance improvements. Remote access via AnyDesk works well for me. I’d be glad to discuss the workflow and start with an initial task. Best, Miguel

@mikes126

I can help you debug and optimize your CUDA kernels directly in your environment — focusing first on correctness, then GPU performance improvements. I have experience working with CUDA/C++, PyTorch CUDA extensions, tensor operations, and GPU memory optimization. For your current tasks (kernel correctness, tensor indexing, convolution, normalization, residual connections), my approach would be: Compare your CUDA implementation against the PyTorch reference output to identify numerical/correctness mismatches. Debug kernel launches, memory access patterns, indexing issues, synchronization problems, and race conditions. Profile bottlenecks using CUDA profiling tools and optimize memory transfers, kernel execution, and occupancy. Validate improvements with benchmarks to ensure speed gains without changing results. Relevant experience: Developing and debugging CUDA C++ kernels. Building custom PyTorch CUDA extensions. Optimizing GPU workloads involving tensor operations and deep learning pipelines. Working in Linux + Git-based development environments. I am comfortable working through AnyDesk/remote access and can adapt to your workflow. For long-term collaboration, I can provide consistent availability for focused optimization sessions.

@FreeLancers360

Hi, Thanks for posting "Experienced CUDA/C++ Developer - Remote Long-Term Collaboration" — it lands right in our wheelhouse, and we'd be thrilled to bring it to life. I love turning a solid brief like yours into a polished result you’ll be proud to put your name on. From your brief I can see this involves ai — all areas we handle in-house. We specialise in C Programming, Python, Linux, CUDA, which lines up directly with what you need. How we'd approach it: - Clarify the use-case, inputs and the exact output you expect - Build and integrate the model / automation pipeline - Evaluate accuracy and tune against real examples - Deploy with monitoring and a clear handover Delivering at the scale of 10 per is no problem for us — we're set up for volume without dropping quality. Happy to work hourly with transparent time tracking and regular check-ins. You can count on tight communication, on-time delivery, and a result that’s effortless to sign off on. I can start right away and keep you updated at every step — let’s make this a great one. Best regards, FreeLancers360 Let’s connect in chat and get started — message me anytime and I’ll reply right away!

@Lazarone1

hi, your cuda optimization work requires deep understanding of gpu performance because correctness and speed must be improved together. i can help debug cuda c++ implementations, analyze kernel behavior, optimize memory usage, improve execution efficiency, and compare custom kernels with pytorch reference results. my approach focuses on identifying bottlenecks, fixing indexing and computation issues, benchmarking improvements, and maintaining clean documented code using linux git and modern gpu development practices. could you share your current cuda environment, gpu model, and the main optimization targets?

@trb2004

❤️Hi there❤️ I have extensive experience in CUDA/C++ development, specializing in optimizing GPU performance and debugging CUDA implementations. Your project perfectly aligns with my skill set, and I am confident in delivering high-quality results within the specified timeframe. Feel free to check out my profile for more details. I understand that you are seeking a skilled CUDA/C++ developer to assist with optimizing GPU tasks, involving debugging and improving CUDA implementations. I am well-versed in CUDA kernels, PyTorch, and performance optimization, making me an ideal fit for your project requirements. I look forward to the opportunity to collaborate with you on this long-term project. Let's discuss further details and how I can contribute to the success of your CUDA development tasks. Warm regards, Thaveesha

@dobrivojej

Hi, Here's what I'd do in the first 48 hours on your web application, so you can judge me on substance, not promises. Your priority is getting your web application live and stable — polish can come after, and I'll plan it that way. I'd fix the performance problem at the source, not with a patch, so it stays fixed under real load. I've fixed the performance problem on live systems more times than I can count; there are few surprises left here. Do you have API access/credentials ready for backend engineering, or should I plan time to set those up? Send me a message and I'll map out the first milestone for your web application today. Talk soon, Dobrivoje.J

@infovergetech

Hello There, Couple of quick question: - Will the work mainly focus on debugging existing CUDA kernels, or do you also need performance optimization using profiling tools like Nsight to maximize GPU utilization? - Are the CUDA modules already integrated into a larger C++/Python application, or do you need support across the complete pipeline, including memory management, kernel optimization, and multi-GPU execution? We have experience in developing, debugging, and optimizing high-performance CUDA applications with C/C++, GPU kernel optimization, memory management, multi-threaded processing, Nsight profiling, and scalable GPU computing for long-term enterprise projects. Our company has 17+ years experience in IT service development. You might see our profile is new, but not new in this business. Kindly open the chatroom, we can discuss your requirement in detail. Also release the payment once we are finish the task as you prefer. Give us a opportunity and we won't fail you. Thanks, Sandeep K.

@kuldeepsharma30

Hi, I have strong experience with **C++, CUDA, Linux, Git, and GPU performance optimisation**, including debugging CUDA kernels, improving memory access patterns, and validating results against reference implementations. I'm comfortable working remotely via **AnyDesk** or your preferred remote desktop solution. **Relevant experience:** * CUDA kernel debugging and optimisation * GPU memory management, shared memory, CUDA streams, and kernel profiling * Performance tuning and benchmarking * PyTorch CUDA extension integration and validation * Tensor operations, indexing, convolution, normalization, and custom GPU kernels * Linux development environment and Git workflows I'm available for long-term collaboration and can dedicate focused sessions to debugging correctness issues first, followed by performance optimisation. I communicate clearly, provide regular progress updates, and ensure every task is thoroughly tested before completion. **Availability:** Flexible, based on your project schedule. **Remote Access:** AnyDesk, TeamViewer, or other secure remote desktop solutions. I'm ready to start immediately and look forward to discussing your current CUDA tasks. **Best regards,** **Naveen**

@tatumm17

I am an experienced CUDA/C++ developer ready to support your long-term remote project with AnyDesk or other remote access tools. I understand you need reliable and efficient development focused on GPU-accelerated computing. I have worked extensively on CUDA-based applications optimizing performance for complex computations. My background includes developing scalable solutions with clean code and thorough testing. I will maintain clear communication and deliver consistent progress updates. Please share more details so we can start. Happy to review your current setup and get this back to a stable state.

@adnanmalik756

Hi, I’m excited about the opportunity to work together on your CUDA/C++ projects. You’re looking for someone to optimize GPU performance and fix CUDA kernel issues, which aligns perfectly with my experience. With 7+ years of working with CUDA and C++, I’ve tackled similar tasks involving kernel debugging and performance enhancements. I have hands-on experience with CUDA kernels, GPU memory optimization, and debugging correctness issues. Working directly in your development environment via remote access tools like AnyDesk sounds great to me. I’m comfortable with the collaborative setup you described. Could you let me know which specific CUDA implementations you’d like me to compare against PyTorch?

@feitosayrikes

Hi there! I'm interested in the CUDA/C++ developer role. GPU kernel debugging and optimization is exactly the kind of work I enjoy, and comparing custom CUDA implementations against PyTorch references to hunt down correctness issues is a problem I've dealt with before. CUDA/C++ experience: I've worked on CUDA kernel development, including debugging tensor indexing issues, convolution and normalization layers, and memory optimization (coalescing, shared memory usage, avoiding bank conflicts). I'm comfortable with CUDA streams and profiling tools to track down performance bottlenecks after correctness is confirmed. Examples: I can walk through past CUDA projects on our call, including cases where I debugged mismatches between a custom kernel and its PyTorch equivalent, and optimized kernel performance afterward. Happy to share specifics once we're in touch. Availability:I can commit to 30 hours per week, flexible on timing to overlap with your schedule. Remote access: I'm fine with AnyDesk, though I'd like to confirm task scope and payment terms clearly before starting, especially since payment is released after two weeks. Just want to make sure milestones are set up properly on the platform so we're both protected. Looking forward to discussing the first task and getting starte