Optimization ledger¶
Current measured plan¶
The largest Base40 bottleneck is DiT denoising. Work is prioritized in this order:
- remove qualified work with NFE/scheduler/distillation design;
- reduce weight and activation bytes with selective quantization or reuse;
- accelerate attention and dominant GEMMs with shape-qualified kernels;
- overlap transfers and phase work when the memory budget allows;
- optimize VAE, conditioning and service lifecycle once Turbo makes them material.
Accepted or exposed¶
| Method | Measured speed | Quality decision |
|---|---|---|
| SageAttention2, Base40 | 1.283× | four blind rounds tied; production path |
| SageAttention2, Base50 | 1.313× | four blind rounds tied; production path |
| strict Triton RoPE | 1.049× | SSIM 0.994831; sample-aligned audio identical |
| Base40 Pattern3 warmup20 | 1.182× | SSIM 0.990041; explicit experimental option |
| Base40 → Base24 | 1.594× | fast profile; not lossless, motion-sensitive risk |
| TP2 on two RTX 3080 cards | 1.567× | exact parallelism; communication limits scaling |
| two independent RTX 4090 workers | 1.893× observed batch concurrency | throughput only |
Experimental¶
| Method | Signal | Why it is not default |
|---|---|---|
| AdaLN-BF16 hybrid INT8 | 1.376× on one 15 s Base40 run | SSIM 0.857694; needs cross-scene qualification |
| Base16 | 2.278× | quality margin is insufficient |
| SGLang INT8+Sage on one 3080 | 515.357 s hot Base40 signal | cross-executor quality and more scenes remain |
| W4A4/SVDQuant | 1.877× full-linear microbenchmark | full residual/scale path has not reached video E2E |
| selected-layer cache | whole-stack cache proves reuse exists | per-layer history can be 55–110 GiB; reference KV drifts |
Rejected or stopped¶
| Route | Measurement | Decision |
|---|---|---|
| global INT8 + Sage2 | 1.745× | reject: motion −21.99%, audio −4.244 dB |
| FP8-SGL on RTX 4090 | 7.315 → 9.320 s/step | reject: 27.4% slower and no memory reduction |
| ConvRot INT8 | 1.38% slower on 4090; 2.12% slower on 3080 | reject on these architectures |
| keep 24 BF16 blocks resident | no gain or slower, ~42.5 GB peak | reject |
| request-local fixed preprocessing cache | later steps only 0.283% faster | stop: no E2E value |
| remove audio tokens | 0.09% timing noise, changed video | reject |
| cache warmup4 | ~1.462× | reject: SSIM 0.8431 |
| whole-stack Taylor/scaling cache | best residual error still ~20.6% | stop before E2E |
Fewer steps require a new contract¶
Dropping NFE alone is not the method. A valid 16/24/32-step route must jointly tune the scheduler, flow shift, solver and—when needed—a trained adapter, then requalify spatial instruction following, motion, audio and final decode. Base24 is the current practical anchor; Base16 shows the speed ceiling but not enough quality headroom. Official 4- and 8-step adapters are evidence that few-step inference is a model-and-sampler contract, not just a smaller integer.
The full 30-route matrix preserves exact workloads and negative results.