comp Recruitment Room comp Remote · Barbados - Barbados
Homeoffice
Anzeige

Beschreibung

Quelle: OnJob.io · Recruitment Room
GPU Programming Software Engineer ROLE TYPE: ContractorWORK ARRANGEMENT: RemoteCOMPENSATION: $96,000–$230,400 annually ($50–$120 per hour) ROLE OVERVIEW Contribute to an AI training project by designing and developing GPU-related programming tasks that help train large language models (LLMs). The work focuses on GPU performance, architecture, and software optimization. You will apply your GPU programming expertise to develop, implement, profile, and optimize GPU-based solutions, with an emphasis on efficient execution and high-performance computing. SCOPE OF WORK Design, implement, and optimize GPU-based software solutions using CUDA, WebGPU, or GLSL.Profile and fine-tune GPU kernels and shaders to maximize performance and computational efficiency.Develop host-side logic and software integrations using C++.Create GPU programming tasks that assess performance, architecture, and implementation quality.Apply GPU programming techniques to solve performance-oriented engineering problems. REQUIRED QUALIFICATIONS Advanced experience with at least one GPU programming language or framework capable of running on NVIDIA hardware.Proficiency in C++ and experience developing host-side software integrations.Strong understanding of GPU programming concepts, performance optimization, and computational efficiency.Ability to design, implement, and evaluate GPU-related programming solutions. PREFERRED BACKGROUND Experience in one or more of the following areas: Graphics programming.Machine learning acceleration.Scientific computing.High-performance computing (HPC).Other technical domains involving GPU technologies. TECHNICAL SKILLS CUDAWebGPUGLSLC++GPU kernel and shader optimizationGPU profiling and performance tuningNVIDIA GPU architecture ENGAGEMENT DETAILS Remote contractor engagement.Flexible task execution, subject to project requirements.Experts must meet minimum weekly task submission requirements.Work is output-based, with payment tied to tasks that meet the project specifications.Task completion time may vary depending on complexity, experience, and individual workflow.Minimum submission requirements apply. COMPENSATION STRUCTURE Indicative hourly rate: $50–$120.Illustrative annualized equivalent: $96,000–$230,400, based on 40 hours per week and 48 working weeks per year.Payment is made per qualifying task that meets the required specifications.Actual earnings depend on completed and approved submissions, task complexity, and project availability. Annualized compensation is an illustrative equivalent based on 40 hours/week × 48 weeks/year. It is not guaranteed earnings because compensation is output-based and subject to minimum submission requirements. START TIMELINE Selection is typically completed within 48 hours.Successful candidates should be ready to begin promptly.The first tasks are expected to start within 24–48 hours after onboarding.
Vollständige Beschreibung anzeigen

GPU Programming Software Engineer

ROLE TYPE: Contractor
WORK ARRANGEMENT: Remote
COMPENSATION: $96,000–$230,400 annually ($50–$120 per hour)

ROLE OVERVIEW

Contribute to an AI training project by designing and developing GPU-related programming tasks that help train large language models (LLMs). The work focuses on GPU performance, architecture, and software optimization.

You will apply your GPU programming expertise to develop, implement, profile, and optimize GPU-based solutions, with an emphasis on efficient execution and high-performance computing.

SCOPE OF WORK

  • Design, implement, and optimize GPU-based software solutions using CUDA, WebGPU, or GLSL.

  • Profile and fine-tune GPU kernels and shaders to maximize performance and computational efficiency.

  • Develop host-side logic and software integrations using C++.

  • Create GPU programming tasks that assess performance, architecture, and implementation quality.

  • Apply GPU programming techniques to solve performance-oriented engineering problems.

REQUIRED QUALIFICATIONS

  • Advanced experience with at least one GPU programming language or framework capable of running on NVIDIA hardware.

  • Proficiency in C++ and experience developing host-side software integrations.

  • Strong understanding of GPU programming concepts, performance optimization, and computational efficiency.

  • Ability to design, implement, and evaluate GPU-related programming solutions.

PREFERRED BACKGROUND

Experience in one or more of the following areas:

  • Graphics programming.

  • Machine learning acceleration.

  • Scientific computing.

  • High-performance computing (HPC).

  • Other technical domains involving GPU technologies.

TECHNICAL SKILLS

  • CUDA

  • WebGPU

  • GLSL

  • C++

  • GPU kernel and shader optimization

  • GPU profiling and performance tuning

  • NVIDIA GPU architecture

ENGAGEMENT DETAILS

  • Remote contractor engagement.

  • Flexible task execution, subject to project requirements.

  • Experts must meet minimum weekly task submission requirements.

  • Work is output-based, with payment tied to tasks that meet the project specifications.

  • Task completion time may vary depending on complexity, experience, and individual workflow.

  • Minimum submission requirements apply.

COMPENSATION STRUCTURE

  • Indicative hourly rate: $50–$120.

  • Illustrative annualized equivalent: $96,000–$230,400, based on 40 hours per week and 48 working weeks per year.

  • Payment is made per qualifying task that meets the required specifications.

  • Actual earnings depend on completed and approved submissions, task complexity, and project availability.

Annualized compensation is an illustrative equivalent based on 40 hours/week × 48 weeks/year. It is not guaranteed earnings because compensation is output-based and subject to minimum submission requirements.

START TIMELINE

  • Selection is typically completed within 48 hours.

  • Successful candidates should be ready to begin promptly.

  • The first tasks are expected to start within 24–48 hours after onboarding.

Arbeitgeber: Recruitment Room

Anzeige
Bewerben

Weiter zur externen Bewerbung

Du verlässt das OPROMA-Portal. Den Bewerbungsweg findest du auf der Zielseite.

Stellenquelle
OnJob.io
Zielwebsite
onjob.io

Die Zielseite öffnet sich in einem neuen Tab. Dort gelten die Datenschutzbestimmungen des jeweiligen Anbieters.

Ref-Nr.: 14731620
Aktualisiert: 22.11.2026
Quelle: OnJob.io
weitere Infos
Ob die Stelle bei OnJob.io noch verfügbar ist und weitere Informationen findest du direkt auf der Website der Quelle. Bitte beachte: OPROMA ist nicht der Arbeitgeber dieser Stelle und kann keine Auskünfte zur Bewerbung geben.

Angebote merken und später wiederfinden

Mit deinem kostenlosen OPROMA-Profil.

Kostenloses Profil erstellen