Local AI, timed on machines I own https://measuredlocally.com/ 2026-10-10T12:00:00Z Alex https://measuredlocally.com/rtx-spark-laptop-prices2026-10-10T12:00:00Z

Microsoft opened orders on Wednesday and five other makers follow on 16 October. The 128GB version is the one that runs the big models, and I would wait for independent tests before buying one.

https://measuredlocally.com/dgx-spark-quad-recipe-speed2026-10-10T12:00:00Z

I moved my four DGX Sparks to a recipe that another owner published on GitHub, and ordinary writing roughly doubled on the same hardware and the same model. The recipe is only a few weeks old, and these are one owner's figures from one cluster.

https://measuredlocally.com/dgx-spark-vs-rtx-spark-laptop2026-10-10T12:00:00Z

Microsoft's RTX Spark laptop now costs less than NVIDIA's own DGX Spark with the same memory. The DGX Spark is the one that can be joined to other boxes, and I do not know yet how many of the recipes its owners share will work on Windows.

https://measuredlocally.com/dgx-spark-64gb2026-10-03T12:00:00Z

NVIDIA announced a DGX Spark with 64GB of memory, half of what the original has, and it costs more than the full one did at launch. The price is the part that stood out to me, and I could not find how NVIDIA measured the gain it claims from linking two or four together.

https://measuredlocally.com/mac-studio-m5-ultra-vs-sparks2026-10-03T12:00:00Z

MacStories tested GLM 5.3 Flash, the model my cluster runs, on a 256GB M5 Ultra Mac Studio, in a review I had missed. One Mac Studio writes about as fast as my four DGX Sparks and my Sparks read a prompt 2.04 times as fast, but the comparison is a rough one.

https://measuredlocally.com/glm-5-3-security-report2026-10-03T12:00:00Z

Anthropic published a security report on GLM 5.3, the family of open models my Sparks run. By Anthropic's count the model turned a known Chrome flaw into a working attack in 50 of 410 attempts, and as far as I could see on 3 October nothing had changed for anyone running it at home.

https://measuredlocally.com/mac-studio-m5-ultra-512gb2026-09-28T12:00:00Z

Apple says its latest Mac Studio can hold 512GB of memory in one box, the same as my whole cluster of four DGX Sparks, and that a cluster of Studios runs AI up to 3x faster than one. I looked up how Apple measured that, and I would wait for independent measurements before ordering.

https://measuredlocally.com/mimo-v2-6-open-weights2026-09-28T12:00:00Z

Xiaomi's MiMo-V2.6-Pro sat at number one on Artificial Analysis's open-weights leaderboard on 27 September, and DGX Spark owners had both versions running within days. One owner reported 68.3 tokens a second on eight Sparks with speculative decoding, up from 17.8, and the early problem was tool use.

https://measuredlocally.com/gguf-in-transformers2026-09-28T12:00:00Z

Hugging Face says its Transformers library can now load the GGUF files that llama.cpp made popular for running models at home. For now it only keeps the file compressed for some Qwen models on Apple silicon, and I think it is most useful to people who work in Python.

https://measuredlocally.com/deepseek-v4-1-flash-two-sparks2026-09-19T12:00:00Z

One of the regular recipe authors on NVIDIA's developer forum published a way to run DeepSeek V4.1 Flash on two DGX Sparks instead of four, with a second trick that applies to any headless Spark. I had not run it myself, and the speeds and the one quality check were the author's own.

https://measuredlocally.com/ternary-bonsai-2-27b2026-09-19T12:00:00Z

Prism ML released a 27 billion parameter model built from Qwen 3.8 with every weight set to -1, 0 or +1, and says that brings it under 6GB. The maker says it keeps 98.2% of the original's scores, and on 19 September that was still the maker's own number.

https://measuredlocally.com/openfaas-four-dgx-sparks2026-09-19T12:00:00Z

OpenFaaS, a UK software company, has written up why it spent about £10,000 on four DGX Sparks: customer data it will not send to a cloud provider, and red teaming its own products. It also found two 4-bit builds of GLM 5.3 Flash running at different speeds, so two builds of the same model are not interchangeable.

https://measuredlocally.com/deepseek-v4-1-flash-on-sparks2026-09-13T12:00:00Z

DeepSeek released a 552 billion parameter model with open weights on 10 September, and within a day an owner had it running on four DGX Sparks. The speeds were similar to GLM 5.3 Flash on my own four, but they came from one owner's first day on a four bit build that had not been checked for quality, as far as I could see.

https://measuredlocally.com/mac-studio-m5-ultra-wait2026-09-13T12:00:00Z

On 11 September Apple's own store was quoting 16 to 18 weeks for any Mac Studio with the full M5 Ultra chip, which meant January. The 512GB option, the one Mac that could hold a four bit build of DeepSeek V4.1 Flash, could not be ordered at all.

https://measuredlocally.com/glm-5-3-flash-nvfp4-builds2026-09-13T12:00:00Z

NVIDIA published its own NVFP4 build of GLM 5.3 Flash on 9 September. I checked 18 builds of that model and of Qwen 3.8 Flash Next for the fault I had found on my own cluster, and found it in three GLM builds, the most downloaded one among them, and in neither of NVIDIA's own.

https://measuredlocally.com/rtx-spark-october-launch2026-09-06T12:00:00Z

NVIDIA confirmed at IFA in Berlin that RTX Spark PCs would ship in October, and released PAIR, a free tool that spreads AI work across the PCs on a home network. Nobody outside NVIDIA had benchmarked an RTX Spark, and I think the free speed-up for RTX cards mattered more on the day.

https://measuredlocally.com/nvidia-buys-hugging-face2026-09-06T12:00:00Z

NVIDIA agreed to buy Hugging Face for $12.93 billion. For those of us running models at home that is our main supplier being bought by our hardware vendor, and I would make sure any model I rely on also lives on my own disk.

https://measuredlocally.com/qwen3-8-flash-next-on-sparks2026-09-06T12:00:00Z

By 6 September owners on NVIDIA's forum had Qwen3.8 Flash Next running on a single DGX Spark at 37 to 44 tokens a second, and one post had GLM 5.3 Flash on one box at 64. The 64 came from a build squeezed to about 2 bits per weight, and I would want to see the quality numbers before copying it.

https://measuredlocally.com/dgx-spark-vs-mac-studio2026-09-01T12:00:00Z

A buyer with two new DGX Sparks asked on Reddit whether to send them back now that Apple had announced the M5 Ultra Mac Studio. I put my own cluster's numbers next to what Mac Studio owners measure, and I think the answer depends on the work: short questions suit the Mac, and agents and long documents suit the Sparks.