wip(standalone): GPU step 16 -- CPU cold drain; hybrid beats CPU once shaders are driver-cached

CPU uploads now bypass the 4-batch / 2 ms cap until the first resident view.
Luna's race (inside the Codex sandbox): CPU 1.354 s vs GPU 3.152 s, warm-up
queue 1.16-1.21 s on every launch. My runs outside the sandbox, same binary
(Step16FairRace, RunGpuFarViewer, interleaved): CPU 1.497 / 1.359 / 1.334 s;
GPU 3.088 (first launch of this build, warm-up 1,287 ms) / 1.262 / 1.260 s
(warm-up 3.2 / 3.1 ms). AMD DxCache was written at my GPU launches (08:46:57,
09:33:37) and never during Luna's runs: the sandbox cannot persist the driver
shader cache, which is why every one of Luna's launches paid the compile.
Audits: hashes, walk 5,642/5,642, 0 diffs vs CPU/WARP, survival 5.10 GB.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01MmSoF64BbRFmk9C3p1VsBJ

Standalone/Research/OwnerShapeWorld/gpu_field_port/CompareStep15Captures.py

@@ -1,10 +1,14 @@
 from pathlib import Path
+import argparse
 
 import numpy as np
 from PIL import Image
 
 
 ROOT = Path(__file__).resolve().parent
+PARSER = argparse.ArgumentParser()
+PARSER.add_argument("--prefix", default="step15")
+PREFIX = PARSER.parse_args().prefix
 VIEWS = (
     "chamber",
     "tunnel",
@@ -20,8 +24,8 @@ VIEWS = (
 
 rows = []
 for view in VIEWS:
-    cpu_path = ROOT / f"step15_cpu_{view}.png"
-    gpu_path = ROOT / f"step15_gpu_{view}.png"
+    cpu_path = ROOT / f"{PREFIX}_cpu_{view}.png"
+    gpu_path = ROOT / f"{PREFIX}_gpu_{view}.png"
     cpu = np.asarray(Image.open(cpu_path).convert("RGB"), dtype=np.int16)
     gpu = np.asarray(Image.open(gpu_path).convert("RGB"), dtype=np.int16)
     if cpu.shape != gpu.shape:
@@ -41,7 +45,7 @@ for view in VIEWS:
     )
 
 lines = [
-    "# Step 15 full-residency capture comparison",
+    f"# {PREFIX.title()} full-residency capture comparison",
     "",
     "All ten CPU/GPU pairs were captured after `[InitialClipmapReady]`. The files were compared numerically as RGB arrays and were not opened in an image viewer.",
     "",
@@ -54,6 +58,6 @@ for row in rows:
         f"| {view} | {size} | {mae:.4f} | {rmse:.4f} | {changed:.4f}% | {over32:.4f}% | {maximum} |"
     )
 
-report = ROOT / "step15_capture_comparison.md"
+report = ROOT / f"{PREFIX}_capture_comparison.md"
 report.write_text("\n".join(lines) + "\n", encoding="utf-8")
 print(report.read_text(encoding="utf-8"), end="")

Standalone/Research/OwnerShapeWorld/gpu_field_port/RunGpuFarAgreement.ps1

@@ -1,5 +1,5 @@
 [CmdletBinding()]
-param([string]$ExecutableName='VoxelForgeGpuFarAgreement.exe',[string]$LogStem='whole_far_rx7900_warp',[switch]$WholeFar,[int]$ClipRadius=3)
+param([string]$ExecutableName='VoxelForgeGpuFarAgreement.exe',[string]$LogStem='whole_far_rx7900_warp',[switch]$WholeFar,[switch]$WholeNear,[int]$ClipRadius=3)
 
 $ErrorActionPreference='Stop'
 if([IO.Path]::GetFileName($LogStem)-ne$LogStem){throw 'LogStem must be a filename stem.'}
@@ -14,6 +14,8 @@ if([IO.Path]::GetFileName($ExecutableName)-ne$ExecutableName-or[IO.Path]::GetExt
 {throw 'ExecutableName must be an executable filename.'}
 if(-not(Test-Path -LiteralPath $ExecutablePath -PathType Leaf)){throw "GPU field runner not found: $ExecutablePath"}
 $Arguments=@("--clip-radius=$ClipRadius");if($WholeFar){$Arguments+='--whole-far'}
+if($WholeNear){$Arguments+='--whole-near'}
+if($WholeFar-and$WholeNear){throw 'Choose either WholeFar or WholeNear, not both.'}
 
 for($Attempt=0;$Attempt -lt 2;$Attempt++)
 {

Standalone/Research/OwnerShapeWorld/gpu_field_port/RunStep15CaptureMatrix.ps1

@@ -1,5 +1,5 @@
 [CmdletBinding()]
-param([string]$ExecutableName='VoxelForgeGpuFieldStep15Timeline.exe')
+param([string]$ExecutableName='VoxelForgeGpuFieldStep15Timeline.exe',[string]$Prefix='step15')
 
 $ErrorActionPreference='Stop'
 $HostProcess=Get-Process -Id $PID
@@ -27,7 +27,7 @@ foreach($View in $Views)
 {
     foreach($Mode in @('cpu','gpu'))
     {
-        $Stem="step15_${Mode}_$($View.Name)"
+        $Stem="${Prefix}_${Mode}_$($View.Name)"
         $CaptureName="$Stem.png"
         $Arguments=@('--clip-radius=3','--workers=8','--quit-after-clipmap','--capture-after-clipmap',
             "--camera=$($View.Camera)","--$Mode-far",

Standalone/Research/OwnerShapeWorld/gpu_field_port/RunStep15CpuReference.ps1

@@ -1,5 +1,5 @@
 [CmdletBinding()]
-param([string]$ExecutableName='VoxelForgeGpuFieldStep15Timeline.exe')
+param([string]$ExecutableName='VoxelForgeGpuFieldStep15Timeline.exe',[string]$LogPrefix='step15_cpu_reference')
 
 $ErrorActionPreference='Stop'
 $HostProcess=Get-Process -Id $PID
@@ -13,7 +13,7 @@ if(-not(Test-Path -LiteralPath $ExecutablePath -PathType Leaf)){throw "Viewer no
 $Reference=$null
 foreach($Workers in @(1,8))
 {
-    $Name="step15_cpu_reference_w$Workers"
+    $Name="${LogPrefix}_w$Workers"
     $StdoutPath=Join-Path $PSScriptRoot ($Name+'.stdout.log')
     $StderrPath=Join-Path $PSScriptRoot ($Name+'.stderr.log')
     $Arguments=@('--stream-benchmark','--shape-world','--force-generate','--stream-hash',

Standalone/Research/OwnerShapeWorld/gpu_field_port/RunStep15GpuSurvival.ps1

@@ -1,5 +1,5 @@
 [CmdletBinding()]
-param([string]$ExecutableName='VoxelForgeGpuFieldStep15Timeline.exe')
+param([string]$ExecutableName='VoxelForgeGpuFieldStep15Timeline.exe',[string]$LogName='step15_gpu_survival')
 
 $ErrorActionPreference='Stop'
 $HostProcess=Get-Process -Id $PID
@@ -11,7 +11,7 @@ if([IO.Path]::GetFileName($ExecutableName)-ne$ExecutableName-or[IO.Path]::GetExt
 {throw 'ExecutableName must be an executable filename.'}
 if(-not(Test-Path -LiteralPath $ExecutablePath -PathType Leaf)){throw "Viewer not found: $ExecutablePath"}
 $Arguments=@('--clip-radius=3','--workers=8','--quit-after-clipmap','--gpu-far-audit','--gpu-far-prewarm')
-$Name='step15_gpu_survival'
+$Name=$LogName
 $StdoutPath=Join-Path $PSScriptRoot ($Name+'.stdout.log')
 $StderrPath=Join-Path $PSScriptRoot ($Name+'.stderr.log')
 $EightGB=8L*1024L*1024L*1024L

Standalone/Source/StandaloneViewer.cpp

@@ -3472,8 +3472,7 @@ float3 Lit=Rock*(Ambient+0.88*Diffuse)+float3(0.50,0.69,0.82)*Head*0.75;
             };
             std::sort(Dirty.begin(),Dirty.end(),[&](const auto& A,const auto& B)
             {const double DA=Distance(A),DB=Distance(B);return DA==DB?A.Level<B.Level:DA<DB;});
-            const bool bDrainColdStart=bGpuFarEnabled&&!bInitialClipmapReported
-                &&!bDisableGpuColdStartDrain;
+            const bool bDrainColdStart=!bInitialClipmapReported&&!bDisableGpuColdStartDrain;
             const auto Start=std::chrono::steady_clock::now();int32 Uploaded=0;
             for(const FStandaloneBatchKey& Key:Dirty)
             {