Accelerator Design Patterns — Design Space Exploration

Design space exploration

For Accelerator Design Patterns, senior architects do not pick one answer — they map the design space, estimate metric movement, and choose based on product constraints.

Option A — conservative

• Conservative: helps lower risk

• Risk: less upside

• Validate with: baseline suite

Option B — balanced

• Balanced: helps good perf/watt

• Risk: may miss peak

• Validate with: multi-workload sweep

Option C — aggressive

• Aggressive: helps peak wins

• Risk: PPA/DV risk

• Validate with: stress suite

Option D — software-first

• Software-first: helps low silicon

• Risk: fragile

• Validate with: controlled apps

DESIGN SPACE — Accelerator Design Patterns
low risk -> balanced -> aggressive
with software-first as alternate axis

Common pitfalls

• Aggressive hardware before workload proof

• Balanced by habit without numbers

Architecture deep dive

Accelerators win on locality and bandwidth contracts, not peak OPS alone.

Concept diagram

ACCELERATOR DATAFLOW

Host CPU ── commands ──► Queue / scheduler
   ▲                         │
   │ completion              ▼
Coherent memory ◄── DMA ── Local SRAM ──► Compute array
                         ▲       │
                         └ tiles ┘

Peak TOPS matters only when data reaches the array at the needed rate.

Metric graph

UTILIZATION BREAKDOWN

compute active   ██████████████████  58%
DMA wait         ██████████          31%
host sync        █████               15%
cache/coherency  ████                12%
idle bubbles     ███████             22%

Low utilization is usually a system integration problem.

Metrics and artifacts

• accelerator utilization

• DMA bandwidth

• kernel launch overhead

• coherency invalidation rate

Mini case study

NPU met TOPs target but end-to-end inference slow — DMA and weight fetch dominated. Architecture added on-chip SRAM tile and double-buffering.

Debug branches

• If util low, check launch overhead and host sync first.

• If BW high, examine weight layout and sparsity support.

Senior review question

Ask: what single metric would prove this concept is working or failing on your workload?

Key takeaways

• Connect every architecture claim to a workload and measurable metric.

• State verification and PPA impact before proposing design changes.

Common pitfalls

• Feature-driven design without MPKI/IPC/bandwidth evidence.

• Ignoring coherency and NoC traffic in cache and accelerator sizing.

Study notes

Re-read this topic with one concrete workload.