Architecture & Performance Modeling Engineer, AI Inference Silicon (Senior / Principal)
Listed by Jobgether for an employer it does not name
- LocationUS
- Work typeRemote
- Salary220k to 350k USD a year
- EmploymentFull-time
- Posted16 November 2026
- SourceListed by Jobgether (Lever)
Last checked44 min ago: still listed on Jobgether. First seen 44 min ago.
Moves fastWhen Jobgether takes a role down early, it is usually about 8 days after posting. Apply soon.
At a glance
- Experience5+ years
- VisaNo sponsorship
- Skills and toolsPython, LLMs
Read from the ad by RoleFinder, with rules and an AI reading; the description is the source.
About the role
This position is listed on behalf of a partner company, which manages all applications and next steps. Our partner is looking for an Architecture & Performance Modeling Engineer, AI Inference Silicon (Senior / Principal) based in the United States.
In this role, you will help define the architecture and performance of next-generation AI inference accelerators designed to power increasingly complex large language models.
You will work closely with chip architects to evaluate architectural options, identify performance bottlenecks, and translate quantitative analysis into actionable design recommendations.
Your work will begin before RTL implementation, using analytical models and advanced simulation tools to guide critical hardware decisions.
You will take ownership of performance modeling for a next-generation chip while contributing to the development of a shared modeling framework.
The role spans compute architecture, memory systems, data movement, interconnects, and multi-chip communication.
You will also analyze real-world AI inference workloads, including prefill and decode operations, to optimize latency, throughput, and hardware utilization.
This is an opportunity to influence the design of cutting-edge AI silicon in a highly collaborative, technically ambitious engineering environment.
Accountabilities:
- Collaborate with chip architects to translate architectural questions into structured modeling studies, defining assumptions, alternatives, performance metrics, and recommendations.
- Conduct architectural trade-off analyses across compute engines, memory capacity and bandwidth, dataflow, DMA, command processing, on-chip interconnects, and chip-to-chip communication.
- Evaluate different hardware configurations and identify opportunities to improve performance while respecting power, area, system, and implementation constraints.
- Analyze representative AI inference workloads across prefill and decode phases, different batch sizes and context lengths, and dense and mixture-of-experts (MoE) models.
- Identify performance bottlenecks affecting latency, throughput, and hardware utilization, translating findings into practical architectural improvements.
- Develop and enhance analytical, transaction-level, and cycle-level performance models, selecting the appropriate level of detail to balance accuracy and development speed.
- Build, maintain, and extend reusable models within a native Rust simulation framework.
- Validate and correlate model predictions against RTL simulations, emulation results, and silicon measurements as they become available.
- Maintain regression coverage and clear documentation of modeling assumptions, limitations, discrepancies, and results.
- Communicate findings effectively to architecture and engineering teams, providing ongoing recommendations as workloads, specifications, and design decisions evolve.
- Take primary responsibility for modeling a next-generation chip while contributing to the continuous improvement of the team's modeling infrastructure.
Requirements:
- At least 5 years of professional experience in computer architecture, microarchitecture, or performance modeling for accelerators, CPUs, GPUs, or similar hardware platforms.
- Proven experience using performance models to influence architectural decisions, identify bottlenecks, evaluate design alternatives, and support recommendations with quantitative analysis.
- Strong understanding of computer architecture fundamentals, including pipelines, queues, parallel execution, memory hierarchies, data movement, and interconnect architectures.
- Hands-on experience developing analytical, transaction-level, or cycle-level performance models, including first-order estimates to validate and interpret simulation results.
- Ability to work effectively with incomplete specifications, document assumptions, and balance modeling accuracy with rapid iteration.
- Strong programming skills in Rust, or expert-level C/C++ proficiency combined with a willingness to work extensively in Rust.
- Proficiency in Python for workload analysis, experimentation, data processing, and performance visualization.
- Ability to read and interpret SystemVerilog RTL and verification results, reconcile implementation details with architectural specifications, and calibrate models as hardware designs mature.
- Excellent analytical, problem-solving, and communication skills, with the ability to explain performance results, assess confidence levels, and recommend clear next steps.
- Ability to collaborate closely with multidisciplinary engineering teams and take ownership of complex technical projects.
- Experience with LLM inference is highly desirable, including attention mechanisms, KV caches, quantization, mixture-of-experts workloads, and multi-chip parallelism.
- Familiarity with the relationship between workload behavior, software execution strategies, and hardware performance is a strong advantage.
- Candidates with experience exceeding the stated requirements may be considered for Principal-level responsibilities, with scope and expectations adjusted according to seniority and expertise.
- Applicants must currently be authorized to work in the United States. New visa sponsorship is not available; however, H-1B transfers may be facilitated for eligible candidates.
Benefits:
- Competitive base salary:
- Senior: $220,000–$270,000 per year.
- Principal: $270,000–$350,000 per year.
- Competitive equity opportunities as part of the overall compensation package.
- Flexibility to consider compensation above the stated ranges for candidates with exceptional expertise, experience, or specialized qualifications.
- Fully company-paid medical, dental, and vision insurance for employees and their dependents.
- Company-paid life and disability insurance, with additional voluntary coverage options.
- Access to supplemental hospital, critical illness, and accident insurance.
- Unlimited paid time off, with an emphasis on maintaining a healthy work-life balance and taking time to recharge.
- 13 paid company holidays annually.
- Remote-first working environment, offering flexibility to work from anywhere in the United States.
- Company-provided computer and home office setup.
- 401(k) retirement plan with company matching, available from the first day of employment.
- Opportunity to work on next-generation AI inference silicon and directly influence major architectural decisions before RTL implementation.
- Close collaboration with experienced engineers and architects working on advanced AI hardware systems.
- Opportunity to contribute to the development of hardware infrastructure designed to serve increasingly large and sophisticated AI models.
How Jobgether works:
We use an AI-powered matching process to ensure your application is reviewed quickly, objectively, and fairly against the role's core requirements. Our system identifies the top-fitting candidates, and this shortlist is then shared directly with the hiring company. The final decision and next steps (interviews, assessments) are managed by their internal team.
We appreciate your interest and wish you the best!
Why Apply Through Jobgether?
Data Privacy Notice: By submitting your application, you acknowledge that Jobgether will process your personal data to evaluate your candidacy and share relevant information with the hiring employer. This processing is based on legitimate interest and pre-contractual measures under applicable data protection laws (including GDPR). You may exercise your rights (access, rectification, erasure, objection) at any time.
#LI-CL1
Get open roles ranked against your CV
Add your CV once. RoleFinder ranks the open roles where you want to work, plus remote ones, against it, shows your best 12 with a one-line reason each, and emails you new matches once a week. The file is read once by AI and not kept; employers never see it.
Sign in with Google or email to start. How we handle your CV