October 3, 2026· ~12 min readKolibri: Aleph Alpha's Apache-2.0 MoE That Puts German First On the Day of German Reunification — a date chosen for symbolism — Aleph Alpha released **Kolibri-1**, an English–German mixture-of-experts transform
August 19, 2026· ~11 min readDeepSeek V4 Flash 0731: Initial Experiences on 2x DGX Spark ## Overview DeepSeek AI released **DeepSeek-V4-Flash-0731**, an updated checkpoint of their mixture-of-experts (MoE) model that power their flaghshi
August 1, 2026· ~14 min readSovereign AI on Your Desk: Why Two DGX Sparks Beat the Cloud In 2024, running a 685-billion-parameter model locally meant a rack of A100s, a six-figure power bill, and a PhD in distributed systems. In 2026, it
May 31, 2026· ~9 min readTwo-Node DGX Spark Cluster: Running DeepSeek V4 Flash at 20 TPS Two NVIDIA DGX Sparks. One 685B-parameter MoE model. A 3× throughput improvement without a single crash. Here's how. ## The Setup: Two GB10 Workstat
May 28, 2026· ~6 min readTaming OOM on DGX Spark: Debugging Unified Memory Pressure in a 2-Node vLLM Cluster Your cluster loads the model. Weights land in GPU memory. Then — crash. `NV_ERR_NO_MEMORY`. The OOM killer takes down `VLLM::Worker_TP`. You restart.
April 13, 2026· ~7 min readvLLM vs SGLang: Choosing an LLM Inference Framework in 2026 Serving large language models at production scale boils down to one problem: getting the most tokens out of your GPU per second, per dollar. Two open