Author: Chris DePuy
-
DeepSeek V4 Flash DSpark on 2× DGX Spark: ~60-67 tok/s with Speculative Decoding
For local inferencing, here is a setup that has proven to be quite stable and fast. It’s the 2x DGX Spark running DeepSeek V4 Flash. It keeps the model running locally and, I’d say, is near Frontier, and doesn’t get lost often when using agents. It’s pretty good at planning, too. The reviews quote ~40…
-
SGLang vs vLLM vs llama.cpp vs Atlas — The Inference Engine Shootout for DGX Spark Clusters
SGLang vs vLLM vs llama.cpp vs Atlas — The Inference Engine Shootout for DGX Spark Clusters The NVIDIA DGX Spark (GB10) has transformed from a curiosity into a genuine building block for local AI infrastructure. What began as a 128GB unified-memory desktop appliance has, through community effort and NVIDIA’s own software maturation, become a node…
-
Cloudflare is Helping to Paywall the Internet
Cloudflare is Helping to Paywall the Internet The 1990s Internet made information free. AI agents are making it billable. Since the mid-1990s, when the commercial Internet took off, the dominant bargain was simple: content for attention. You browsed, you clicked, you saw ads — and the publisher got paid by advertisers. Most information was free…
-
DGX Spark Clusters by the Numbers: A Sizing Guide Across 1, 2, 3, and 4 Nodes
NVIDIA’s DGX Spark (GB10) started as a deskside curiosity — a 128GB unified-memory workstation drawing ~38W from the GPU during inference. Over the last several weeks, a wave of open-source recipes and community benchmarks has turned the Spark into a modular building block. Users are connecting 2, 3, and 4 units directly — no switch…