PseudoBrain-Qwen3-Coder-Next-NVFP4 — 1M-Context Frontier 80B-A3B Agentic Coding MoE (Runs 1M Context in ~46 GB Total VRAM)

Created by Ian Kurt Galeon (@defnotean · Hugging Face @defnoteatin · iangaleonofficial@gmail.com)
GitHub & Universal pb27_vllm Engine Plugin: https://github.com/defnotean/pseudobrain
Architecture Paper: PSEUDOBRAIN_ARCHITECTURE_PAPER.pdf · PSEUDOBRAIN_ARCHITECTURE_PAPER.md · Benchmark Receipts: reports/
Open to LLM Inference Systems / Kernel & Engine Engineering roles, research collaborations, and enterprise on-premise 1M-context deployments.

---

1. Why PseudoBrain-Qwen3-Coder-Next Exists

In standard Transformer serving, model weights are paid once, while KV cache memory grows linearly ($O(L)$) and prefill compute grows quadratically ($O(L^2)$) — forcing long-context (128K–1,000,000 token) workloads to OOM even when the GPU easily fits the model weights.

PseudoBrain-Qwen3-Coder-Next-NVFP4 (PB27Qwen3NextForCausalLM) solves this via the PseudoBrain architecture:

1. Hybrid Windowed-Attention + Parallel Gated DeltaNet (GDN): 48 layers: 36 native Gated DeltaNet + 12 full-attention layers ([3, 7, 11, ..., 47]) converted into 16K SWA + Parallel Gated DeltaNet + Quest Exact Memory. Recurrent GDN state is $O(1)$ fixed size regardless of whether context is 4K or 1,000,000 tokens.
2. Three-Tier NVMe/Page-Cache Quest Exact Memory (piece_attention): On the sparse anchor layers, queries attend over [4 StreamingLLM Sinks + Super-Block Landmarks + Top-k Quest Exact-Memory Pages + 16,384 Sliding Window] via FlashAttention-2 LSE-merge (pb27_vllm/memory.py) with 2-Level Hierarchical bf16 Quest Super-Block Bounds (PB27_MEMORY_SUPERBLOCK=64).
3. Piecewise-Proportional, Strictly Non-Overlapping Block RoPE (_reposition_rope_blocks): Enforces a mandatory 64-token footprint per retrieved block plus proportional historical-gap scaling ($T(b_j) = P_{ ext{start}} + j \cdot B + ext{round}(rac{G_{ ext{cum},j}}{G_{ ext{tot}}} S_{ ext{slack}})$) — preserving exact +1 local continuity across contiguous block runs, strictly preventing phase overlap between nearby non-adjacent blocks, and maintaining 99%+ proportional macro-timeline distance fidelity at 1M tokens without static YaRN degradation.
4. Dense Global Aggregation (PB27_MEMORY_GLOBAL_AGG="auto"): Automatically detects high-participation-ratio global queries (_detect_global_agg) and activates (a) Resident Super-Block Summary Landmarks (smean_k, smean_v, 1 summary token per 4,096 tokens in GPU HBM), (b) Stratified Super-Block Timeline Coverage, and (c) Novelty-Rotating Decode Working-Set Sweeps (visit_counts) so multi-step reasoning sweeps across hundreds of distinct blocks over 1M tokens.
5. True $O(1)$ Constant-Speed Decode & Version-Resilient vLLM Plugin: Decouples the scaled prefill retrieval budget from the constant decode budget (PB27_MEMORY_DECODE_STRIDE=64, PB27_MEMORY_DECODE_BLOCKS=64), cutting 1M-token KV cache memory by 93.5% (~48.0 GB → ~3.1 GB), with dual vLLM V1/V2 runner support and a zero-hook ForwardContext fallback.

Hardware & Precision Editions for PseudoBrain-Qwen3-Coder-Next

Repository	Precision	Model Weights	Total VRAM at 1,000,000 Tokens	Target GPUs
defnoteatin/PseudoBrain-Qwen3-Coder-Next-NVFP4	W4A16	~41.5 GB	~45.5 GB	2x RTX 3090 / 4090 (48 GB), RTX 6000 Ada (48 GB), 2x RTX 5090, DGX Spark, AMD MI300X
defnoteatin/PseudoBrain-Qwen3-Coder-Next-FP8	FP8	~80.5 GB	~85.0 GB	H200 (141 GB), DGX Spark (128 GB), AMD Instinct MI300X (192 GB), 2x A100/H100
defnoteatin/PseudoBrain-Qwen3-Coder-Next-NVFP4 (This Repo)	NVFP4	~41.5 GB	~45.5 GB	NVIDIA Blackwell (DGX Spark GB10, 2x RTX 5090, RTX PRO 6000 96 GB, B200)
defnoteatin/PseudoBrain-Qwen3-Coder-Next-BF16	BF16	~160.0 GB	~165.0 GB	AMD MI300X (192 GB), 2x H200 / B200, Multi-GPU CUDA / ROCm

---

2. Two-Command Quickstart (vLLM on NVIDIA CUDA & AMD ROCm)

Option A: Install Plugin Directly into Existing vLLM Environment (2 Commands)

# 1. Install the universal PseudoBrain vLLM V1 plugin (auto-registers via vllm.general_plugins)
pip install --no-deps "git+https://github.com/defnotean/pseudobrain.git#subdirectory=serving/pb27_vllm"

# 2. Launch 1,000,000-token OpenAI-compatible server
VLLM_ALLOW_LONG_MAX_MODEL_LEN=1 vllm serve defnoteatin/PseudoBrain-Qwen3-Coder-Next-NVFP4 \
  --served-model-name pb27 \
  --max-model-len 1048576 \
  --enforce-eager \
  --no-enable-prefix-caching \
  --kv-cache-memory-bytes 6G \
  --max-num-seqs 16 \
  --enable-auto-tool-choice \
  --tool-call-parser qwen3_coder

(Note: Every Hugging Face checkpoint also bundles pb27_vllm_plugin/ directly inside the repository if you prefer pip install --no-deps -e ./pb27_vllm_plugin from a local snapshot download.)

Option B: One-Command Docker Launch (serve.sh — Auto-Detects x86_64 & aarch64)

git clone https://github.com/defnotean/pseudobrain.git
cd pseudobrain
bash serving/pb27_vllm/serve.sh defnoteatin/PseudoBrain-Qwen3-Coder-Next-NVFP4 pb27 8000

Option C: Connect OpenCode / Agentic Coding Harnesses (~/.config/opencode/opencode.jsonc)

{
  "$schema": "https://opencode.ai/config.json",
  "model": "pseudobrain/pb27",
  "small_model": "pseudobrain/pb27",
  "compaction": { "auto": true, "prune": true },
  "provider": {
    "pseudobrain": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "PseudoBrain 1M Local Server",
      "options": {
        "baseURL": "http://127.0.0.1:8000/v1",
        "apiKey": "local-loopback",
        "timeout": 600000
      },
      "models": {
        "pb27": {
          "name": "PseudoBrain-Qwen3-Coder-Next-NVFP4 (1M Context)",
          "tool_call": true,
          "limit": { "context": 1048576, "output": 16384 }
        }
      }
    }
  }
}

---

3. Verified PseudoBrain Benchmark Results (Committed Receipts in reports/)

3.1 Same-Hardware (DGX Spark GB10, TP1) Speed & Memory vs. Stock at 65K and 1,000,000 Tokens

Measured directly against upstream runtime-qualification.json on the exact same DGX Spark (GB10, 128 GB Unified Memory), TP1 hardware:

Metric on DGX Spark (GB10, TP1)	Stock Qwen3.8-Flash-Next-NVFP4 (runtime-qualification.json)	PseudoBrain-Qwen3.8-Flash-Next-NVFP4 (Ours)	Speedup / Savings
65K Prefill Time (TTFT)	38.93 s (1,680 tok/s)	21.02 s (3,108 tok/s)	1.85× faster prefill (-46% latency)
65K Decode Speed (tok/s)	35.8 – 39.1 tok/s (mean 37.8 tok/s, MTP-3)	53.4 – 88.3 tok/s *(mean 65.9 tok/s, peak 183.5 tok/s w/ DFlash2)*	1.74× – 2.26× faster decode
1,000,000 (1M) Prefill Time (TTFT)	1,128.0 – 1,132.2 s (18.9 min, 926 tok/s)	543.06 s (9.0 min, 1,801 tok/s)	2.08× faster 1M prefill (saves ~10 min per 1M prompt)
1,000,000 (1M) Decode Speed (tok/s)	33.3 – 37.1 tok/s (mean 34.9 tok/s)	44.6 – 67.7 tok/s *(mean 57.1 tok/s, 67.7 tok/s on 1M needle)*	1.64× – 1.94× faster decode at 1M tokens
KV Cache Memory at 1M Tokens	Full 1,048,576-token KV cache (max-running-requests 1, requires static YaRN factor=4.0)	Fixed 16,384-token GPU window (~486 MiB active) + NVMe Quest $O(1)$ archive (--max-num-seqs 16, no static YaRN degradation)	~96%–99.2% less GPU KV memory at 1M + 16 concurrent sequences

3.2 8K to 1,000,000 Token Exact Retrieval (needle@0.1/0.5/0.9, multi4, trace4)

Context Length ($L$)	needle@0.1	needle@0.5	needle@0.9	multi4 (4 constants)	trace4 (4-hop chain)	Flash-Next Decode tok/s	Nex-35B-A3B Recall	GPU KV Pool Used
8,192	1.0	1.0	1.0	1.0 (4/4)	1.0	93.01 tok/s	5/5 (100%)	8.1%
32,768	1.0	1.0	1.0	1.0 (4/4)	1.0	66.51 tok/s	5/5 (100%)	8.1%
65,536	1.0	1.0	1.0	1.0 (4/4)	1.0	67.86 tok/s	5/5 (100%)	8.1%
120,000	1.0	1.0	1.0	1.0 (4/4)	1.0	50.50 tok/s	5/5 (100%)	8.1%
200,000	1.0	1.0	1.0	1.0 (4/4)	1.0	50.81 tok/s	5/5 (100%)	8.1%
250,000	1.0	1.0	1.0	1.0 (4/4)	1.0	69.87 tok/s	5/5 (100%)	8.1%
500,000	1.0	1.0	1.0	1.0 (4/4)	1.0	64.91 tok/s (183.5 peak)	5/5 (100%)	8.1%
1,000,000	1.0	1.0	1.0	0.75–1.0	1.0 (Nex)	63.76 tok/s (67.7 needle)	5/5 (100%)	8.1%

3.3 General Intelligence, Reasoning, & Software Engineering Parity

• MMLU (n=300, 0-shot): 85.00% (255/300, within statistical stderr of stock ~85.0–87.9%).
• GSM8K (n=150, strict CoT): 96.00% (144/150, +2.71% above official 93.29% report).
• MMLU-Pro (n=150, 10-choice college reasoning): 81.33% raw / 85.31% non-truncated (vs. 73.23% Base / 80.6% Instruct).
• Algorithmic Coding (custom_code_bench, 33 medium-hard Python problems): 96.97% (32/33) on PseudoBrain-Qwen3.8-Flash-Next at 74.74 tok/s; 90.91% (60/66) on PseudoBrain-27B.
• Real-World Post-Cutoff Software Engineering (PB-SWE-Live, 59 tasks, 100-step agent loop): 26 / 59 (44.1%) vs. Stock 28 / 59 (47.5%) (paired McNemar $p \approx 0.7$, statistically indistinguishable).

---

4. Complete PseudoBrain 1M-Context Model Lineup (10 Families × 4 Precisions = 40 Models)

All 40 repositories share the single unified pb27_vllm plugin (pip install --no-deps "git+https://github.com/defnotean/pseudobrain.git#subdirectory=serving/pb27_vllm"), with slim BF16 GDN delta overlays hosted at defnoteatin/pb27-checkpoints:

Model Family	Params (Total / Active)	1M KV Reduction	4-Bit W4A16 (RTX 30/40/AMD)	8-Bit FP8 (4090/H100/MI300X)	4-Bit NVFP4 (Blackwell/Spark)	16-Bit BF16 (Universal)
PseudoBrain-Qwen3.8-Flash-Next	36B / ~1.7B	96.0%+ (~48.0 GB → ~486 MiB active window)	W4A16 (~21.0 GB)	FP8 (~36.5 GB)	NVFP4 (~21.0 GB)	BF16 (~72.0 GB)
PseudoBrain-Qwen3-Coder-Next	80B / 3.0B	93.5% (~48.0 GB → ~3.1 GB)	W4A16 (~41.5 GB)	FP8 (~80.5 GB)	NVFP4 (~41.5 GB)	BF16 (~160.0 GB)
PseudoBrain-Qwen3-Coder-30B-A3B	30.5B / 3.3B	98.4% (192.0 GB → ~3.1 GB)	W4A16 (~16.5 GB)	FP8 (~31.0 GB)	NVFP4 (~16.5 GB)	BF16 (~61.0 GB)
PseudoBrain-Nex-35B-A3B	35B / ~3.0B	97.6% (~20.0 GB → ~486 MiB active window)	W4A16 (~18.5 GB)	FP8 (~35.0 GB)	NVFP4 (~22.5 GB)	BF16 (~70.0 GB)
PseudoBrain-27B	27B / 27B	99.2% (64.0 GB → ~486 MiB active window)	W4A16 (~15.5 GB)	FP8 (~28.5 GB)	NVFP4 (~19.5 GB)	BF16 (~54.0 GB)
PseudoBrain-Qwen3-32B	32.8B / 32.8B	98.4% (256.0 GB → ~4.1 GB)	W4A16 (~18.5 GB)	FP8 (~34.5 GB)	NVFP4 (~18.5 GB)	BF16 (~66.0 GB)
PseudoBrain-Gemma-4-31B	31B / 31B	92.2% (~45.0 GB → ~3.5 GB)	W4A16 (~17.5 GB)	FP8 (~32.5 GB)	NVFP4 (~17.5 GB)	BF16 (~62.0 GB)
PseudoBrain-Gemma-4-26B-A4B	25.2B / 3.8B	90.2% (~22.5 GB → ~2.2 GB)	W4A16 (~14.2 GB)	FP8 (~26.5 GB)	NVFP4 (~14.2 GB)	BF16 (~51.0 GB)
PseudoBrain-Qwen3.5-9B	9B / 9B	99.2% (32.0 GB → ~256 MiB active window)	W4A16 (~5.5 GB)	FP8 (~9.5 GB)	NVFP4 (~5.5 GB)	BF16 (~18.0 GB)
PseudoBrain-Qwen3.5-4B	4B / 4B	99.2% (24.0 GB → ~192 MiB active window)	W4A16 (~2.8 GB)	FP8 (~4.8 GB)	NVFP4 (~2.8 GB)	BF16 (~8.5 GB)

---

5. License, Citation & Contact

• Code & Plugin License: Apache-2.0 (https://github.com/defnotean/pseudobrain). Model weights inherit their upstream base model license (apache-2.0).
• Author: Ian Kurt Galeon
  • GitHub: @defnotean
  • Hugging Face: @defnoteatin
  • Email (Recruiting, Research & Enterprise Inquiries): iangaleonofficial@gmail.com