Insights
Data-driven analysis: market state, supply, and what the numbers mean for buyers.
LatestCMP 170HX, B70, V100: Cheap Cards Punching UpSame model, same W4A16 quant, same vLLM engine: a $1,050 CMP 170HX runs 255.9 tok/s and a $9,000 RTX 5090 gets 137.9.Sep 28, 2026 · 9 min readRead →- Sep 28, 2026 · 8 min readQwen3.8-27B on 22 GPUs: 44 Sourced RunsOne model, 22 GPUs, 44 sourced runs, 16.8 to 435 tok/s. The card matters less than the quant, the engine and the context you run it at.
- Sep 28, 2026 · 9 min readOne 180B MoE, Five Formats: 41 Sourced RunsQwen3.8-Flash-Next's split entities merged into one page: 41 sourced rows across three checkpoints, 20.3 to 474 tok/s, and what fits a 24GB card.
- Sep 28, 2026 · 6 min readSeptember 2026: What 164 Runs Changed164 new rows, 142 of them sourced, two new models measured, and a used-GPU market that moved up 24%: the month in board data.
- Sep 15, 2026 · 7 min readContext Is Not a Cliff: 790× the Window, +18%Ten context points from 327 to 258,790 tokens on one DGX Spark: decode throughput barely notices context length. What actually moves the number.
- Sep 15, 2026 · 7 min readOne Spark, 552B: Engine Beats Card CountOne DGX Spark runs a 552B MoE at 47 tok/s; two reach 154, three drop to 46. At this scale the engine and format decide throughput, not the box count.
- Sep 15, 2026 · 6 min read3.5 Bits Beat 4 Bits: Formats, Not PrecisionEXL3 at 3.5 bpw outran MXFP4 by 2× on the same 552B MoE. Three clean ladders where only the format changed — and how to pick a quantization.
- Sep 15, 2026 · 7 min readTwo RTX 3060s: 16 tok/s Dense, 94 tok/s MoEThe same dual-3060 rig runs a 30B dense model at 16 tok/s and a 30B mixture-of-experts at 94. That 6x spread decides whether the build is worth it.
- Sep 11, 2026 · 9 min readMoE Expert Offload: Same Weights, 26× ApartSame weights, same box: 3.3 tok/s with whole-layer expert offload vs 150 with row-selective UVA reads, and what decides big-MoE speed on consumer GPUs.
- Sep 10, 2026 · 3 min readGPU prices, Sep 10 2026: entry cards cooledNewegg-observed snapshot for Sep 10, 2026 — RTX 3060 $454.62, 5060 Ti 16GB $779.99, RTX 5090 $4,899.99, Arc Pro B60 $959, each row with its source.
- Sep 10, 2026 · 3 min readThe trending local models nobody has measuredHugging Face's trending list is full of models with no published local benchmark. See which ones the community has actually run - and which are blank.
- Sep 7, 2026 · 6 min readMoE Expert Cache: VRAM Stops Being the CeilingExpert caching moves cold MoE experts to system RAM — FreeToken paper plus llama.cpp RFC #24528. Community: 90+ tps on a 16GB 5070 Ti. As of 2026-09-07.