GPU Senior Architect / Heterogeneous Compute Architect – £250k+, Benefits – Cambridge, UK

The Opportunity

Architects capable of operating across GPU microarchitecture, AI acceleration, memory systems and SoC-level trade-offs are exceptionally rare.

We’re supporting a global semiconductor company in the search for a GPU Senior Architect / Heterogeneous Compute Architect to help shape future generations of graphics and AI compute technology.

This is a senior, technically influential position for an architect with the depth to rethink how graphics and machine-learning workloads are executed within future mobile platforms.

The successful individual will combine deep GPU architecture expertise with an understanding of AI acceleration, memory systems, workload behaviour and production silicon delivery.

International relocation support is available, but the successful candidate must be willing to relocate permanently to the UK and work from Cambridge.

The Opportunity

As a GPU Senior Architect / Heterogeneous Compute Architect, you will influence the architectural direction of future mobile graphics and AI compute platforms.

You will assess where specialist graphics and AI processing resources should remain independent and where sharing execution, scheduling, memory or data-movement capabilities could deliver better system-level results.

The role requires genuine architectural depth rather than technology exposure. You will be expected to evaluate alternatives, make defensible technical decisions and drive concepts through modelling, implementation, tape-out and silicon bring-up.

You’ll work across GPU hardware, AI acceleration, software, runtime, performance and wider SoC architecture teams, balancing innovation against performance, power, silicon area and implementation constraints.

Key Responsibilities

• Define architectural approaches that bring graphics and machine-learning acceleration together within future mobile compute platforms.
• Analyse emerging graphics and AI workloads to determine their implications for future hardware architecture.
• Evaluate the trade-offs between dedicated processing engines and hardware resources shared across graphics and AI workloads.
• Develop execution architectures incorporating scalar, vector and matrix/tensor processing capabilities.
• Shape how workloads are partitioned, scheduled and dispatched across different compute resources.
• Define mechanisms for resource allocation, quality of service and workload isolation when graphics and AI applications operate concurrently.
• Influence the long-term roadmap for future mobile GPU architectures.
• Drive GPU microarchitecture decisions across shader cores, execution pipelines, scheduling systems and cache structures.
• Define memory-system strategies covering shared caches, bandwidth allocation, latency and contention management.
• Contribute to scalable on-chip communication and interconnect architectures for heterogeneous compute traffic.
• Reduce unnecessary data movement across graphics, AI and wider SoC compute domains.
• Evaluate architectural options against performance, power consumption and silicon-area constraints.
• Collaborate with CPU, AI software, runtime, graphics-software and system-architecture teams.
• Ensure architectural decisions account for SoC-level power, thermal, physical-design and floorplanning constraints.
• Align hardware architecture with the requirements of graphics APIs, AI runtimes and machine-learning frameworks.
• Support performance modelling, workload characterisation and architectural exploration.
• Provide technical leadership from initial concept through implementation, tape-out and silicon bring-up.
• Communicate complex architectural decisions clearly across multidisciplinary engineering teams.

Essential Skills & Experience

To be considered for this opportunity, you should demonstrate substantial architecture experience across GPU, AI acceleration or heterogeneous compute.

You will likely possess:

• 15+ years of relevant experience within GPU architecture, AI accelerators or heterogeneous compute systems.
• Deep understanding of GPU microarchitecture, including SIMD/SIMT execution, shader cores, scheduling and memory systems.
• Experience defining or influencing production GPU architectures across multiple product generations.
• Strong knowledge of tensor or matrix computation and modern AI acceleration techniques.
• Experience evaluating dedicated versus shared compute-resource architectures.
• Strong understanding of workload scheduling, dispatch, resource allocation and concurrent workload behaviour.
• Experience with cache hierarchies, bandwidth management, on-chip interconnect and data-movement optimisation.
• Expertise in performance modelling, workload analysis and power-performance-area trade-offs.
• Experience developing high-volume production silicon.
• Practical understanding of tape-out, physical-design constraints and silicon bring-up.
• The ability to work effectively across hardware, software, runtime and system-architecture teams.
• Strong technical leadership, communication and influencing capability.
• The ability to challenge established architectural assumptions and build support for a new technical direction.

What This Role Isn’t

This role is unlikely to be suitable if your experience is predominantly focused on:

• GPU software or graphics-driver development without meaningful hardware architecture responsibility.
• General SoC architecture without deep GPU microarchitecture experience.
• AI accelerator development without sufficient understanding of GPU execution and graphics workloads.
• RTL implementation without ownership of architectural or microarchitectural decisions.
• Performance analysis without experience influencing the underlying hardware architecture.
• Engineering management with limited recent technical involvement.
• Academic research without evidence of delivery into production silicon.
• CPU architecture without substantial exposure to GPU or heterogeneous compute systems.

This search requires someone capable of connecting GPU architecture, AI acceleration, memory systems and production silicon delivery.

Why Consider This Opportunity?

Opportunities to influence architecture at this level are exceptionally rare.

You’ll join a global semiconductor company investing significantly in future graphics and AI compute capabilities, working alongside experienced architects, researchers and silicon engineers.

The position offers the opportunity to:

• Influence the long-term direction of future mobile compute architectures.
• Solve complex problems spanning graphics, AI and heterogeneous computing.
• Challenge established assumptions around specialist and shared processing resources.
• Work across the complete architectural lifecycle, from workload analysis through to production silicon.
• Operate with significant technical influence across multiple engineering disciplines.
• Help define technologies intended for deployment at substantial global scale.

For an architect motivated by difficult technical problems, long-term product influence and the opportunity to shape a new architectural direction, this is a genuinely significant appointment.

Package & Benefits

• Base salary: £180,000–£250,000+
• Flexibility above the stated range for an exceptional Chief Architect-level candidate
• Final compensation determined by experience, technical depth and architectural track record
• 33 days’ annual leave, including UK public holidays
• Group personal pension
• Private medical insurance
• Life insurance
• Medical-expense claim scheme
• Employee assistance programme
• Cycle-to-work scheme
• Company sports club and social events
• Additional time off for learning and development
• International relocation support available
• The opportunity to work alongside recognised experts within GPU and heterogeneous-compute architecture
• Significant influence over future generations of mobile graphics and AI technology

This is a Cambridge-based position. Permanent relocation to the UK is required and an internationally remote arrangement is not available.

Application Process

My client offers a streamlined 3-stage interview process:

• Stage 1: Technical Screening (45 minutes)
• Stage 2: Technical Panel Interview (2 hours 10 minutes)
• Stage 3: HR / Final Discussion

The process can typically be completed within 2-4 weeks, depending on candidate availability.