Director of ProductLondon, UK

I shape infrastructure products from the hardware up.

I’m Darren Soothill, a Director of Product working across infrastructure, hardware and storage. This is where product direction meets the bench: the systems I test, the trade-offs I find and the details that change the roadmap.

Series / 01 Complete
LOCAL

Current focus

Strix
Halo

01

Setup that can be repeated

02

Throughput in context

03

Thermals over time

Current product lens

Local AI on AMD Strix Halo

Capability · cost · experience · sustained performance

Product question / 01Local AI / 2026

When does local AI become a useful product choice?

01

Local LLMs on Strix Halo

The product question is bigger than a headline benchmark. Local inference changes privacy, cost, latency and deployment — but only if the hardware and software form a system people can actually use.

I took that question to the bench: what I installed, what broke, how I measured it and which engineering trade-offs deserve to shape the product decision.

Read the 7-part series
  1. 01
    GMKtec EVO-X3 baseline

    Hardware, firmware state and NVMe performance measured as one Linux system.

    Published
  2. 02
    ROCm without folklore

    The working Linux stack, its support boundary and a recovery runbook.

    Published
  3. 03
    Vulkan versus ROCm

    Matched prompt, generation and concurrency results across two models.

    Published
  4. 04
    Find the useful quant

    Fit, context and speed from 0.8B to a 397B-class model.

    Published
  5. 05
    Run the long-session test

    122B at a near-full 256K context for more than 100 minutes.

    Published
  6. 06
    Keep the useful stack

    The final product, runtime and operations decisions.

    Published
  7. 07
    DeepSeek V4 Flash 0731 on evox3

    A failed first load, a measured 10× experiment and the repeat test that changed deployment.

    Published

The archive

Recent field notes

The technical detail behind product decisions: storage, Linux and the bits between a specification sheet and a working system.

View every note
2026.08.04 local-ai

DeepSeek V4 Flash 0731 on evox3: the repeat that changed deployment

DeepSeek V4 Flash 0731 on evox3 and ROCm 7.14: a failed first load, a measured 10× sparse-prefill result, and the repeat test...

2026.08.03 local-ai

A local LLM stack worth keeping

The final Strix Halo stack: which hardware, Linux memory model, runtimes, models and operating controls I would keep after the benchmarks.

2026.08.03 local-ai

The long-session test: 122B at a 256K context

A 101-minute, 165-request soak of Qwen3.5 122B at a nearly full 256K context, including throughput, memory, thermals and request-retention fixes.

2026.08.03 local-ai

Finding the useful quant: what fits is not what wins

A measured 0.8B-to-397B Strix Halo model sweep showing why file size, active parameters, context cost and answer quality must be separate decisions....

How I workDirection ↔ detail

“Good product direction starts with the customer problem — and survives contact with the hardware.”
01

Start with the outcome

Who needs the product, what must change for them and which constraint is real.

02

Get close to the system

Hardware, storage paths, versions and topology reveal the actual trade space.

03

Bring evidence back

Measurements and rough edges turn assumptions into roadmap decisions.

Comparing notes on local AI or storage?

darren@soothill.com