个人 × AI 智能体 × 工具链 = 超级个体 用产品和投资赚点小钱。 分享独立开发、增长技巧，以及 AI 实战经验与教程。
China
Joined August 2016
- Tweets2.5K
- Following415
- Followers42.4K
- Likes3.2K
别再买 DGX Spark 、RTX Spark 这种纯垃圾了。
跑 LLM 比不上多卡 2080ti ，甚至不如多卡V100。
跑图片、视频生成，一张4090都能把他们秒成渣。
4090 拥有 1万6千多个 CUDA 核心，只是显存小了点，但是配合comfyUI的量化权重以及CPU offload ，能顺畅跑各种AIGC模型。
这俩 Spark 唯二的优势就是功耗低，显存大，但算力极低，价格超贵。买这俩的，都是不懂行的。
Cursor 之前送的 1万刀额度，前段时间debug 和优化开的 fable 5.1 1M 上下文最高强度，迅速烧了上千刀，我老实了。
现在乖乖用 300K上下的Opus 5.5 ， 又便宜又好用，而且少一些上下文预算，Agent 自己也会节约一些，不然Agent 会猛蹬。
另外期待 Grok 4.8 发力。
测试了一下 3 Steps Turbo Lora Taomate
单卡 4090 配合我的优化方案，MiniMax H3生成384p草稿，LTX 2.5 进一步去噪和高清，15秒视频需要93秒，如果不带音频，去掉音频vae编解码，15秒只需要73秒，5秒视频35秒。
胜利了，凭经验 debug，胜过AI。
昨天用 Claude Fable 5.1 debug 到 凌晨 1 点，今天早上八点继续，一个劲到12点靠我先前的经验才解决。
我遇到了一个奇怪的bug，某个工作流，5秒的可以对口型，但是 6秒和8秒的就不行。反正诸如此类吧，AI给了种种理由，我调整了各种参数进行重生成验证，也花式剪切音频。最终凭经验发现是没对齐，🌿。
魔改了一下英伟达团队的最新成果Sol-H3-Spark，让原本需要100G显存才能跑起来的模型，在 单卡 4090成功运行，并且生成 5 秒的 720p 视频只需要推理 50 秒左右，保留大部分效果。
关键思路是MiniMax H3 生成的384p潜空间草稿，交给LTX 2.5 进行去噪和高清化，两种模型使用特殊的桥接器粘接。充分利用了MiniMax 的运动和其他优点，又同时利用了LTX生成速度快和高清的优点。并且支持了原生的音频驱动生成，使其可以实现数字人或者MV效果。
🚀 Sol-H3 on DGX Spark: 768p in Under a Minute 🤩
Monday: 8×B300, 5s 768p in 1.65s — faster than playback.
Today: the same stack on one desktop Spark — about 56s hot E2E.
Five seconds of 1344×768 video at 24 FPS with stereo audio, on a single NVIDIA DGX Spark (GB10).
Two-stage, not the datacenter profile:
384p H3 draft → latent ×2 → H3-to-LTX VAE adapter → 768p LTX refine → VAE decode
No decode/re-encode between stages. Stage 2 is conditioned on the draft latent, so Gemma stays off the box. Quantized weights stay resident; sparse attention cuts the rest.
Stage 1 takes any MiniMax-H3 few-step LoRA. Timing is hot E2E (encode → both stages → video/audio VAE); cold start and MP4 mux are separate. Apache 2.0.
Server was realtime. Edge is one box, under a minute.
🔗 nvlabs.github.io/Sana/Sol-En…
Amazing team effort—full credits in the blog.
@haopengl33 @lawrence_cjs @yitongli165665 @shanasaimoe ,Jingyu Xin, @HaochengXiUCB @songhan_mit
我发现这些加速技术都是多卡并行张量计算的，而且大概率消费级显卡还用不了这种特性。
所以合理推测 Fal 后训练的 H3 Max 也是这种方式，也就是说这种优化是针对数据中心机器的，那消费级显卡来跑，可理解为几乎没什么优化。
🚀 Sol-H3: @MiniMax_AI H3 Video Generation Faster Than Playback 🤩
Five seconds of world. 1.653 seconds to infer.
We’re releasing Sol-H3, our fastest end-to-end MiniMax-H3 inference stack yet. On one 8× NVIDIA B300 Blackwell system, it generates five seconds of 1344×768 video with stereo audio in 1.653 seconds. Across 1×, 4×, and 8× B300, Sol-H3 reaches up to a 15.54× speedup versus Base H3.
Compared with 50-step Base H3 Dense on the same 8× B300 system, the four-step Sol-H3 profile delivers:
• 5s: 18.250s → 1.653s (11.04×)
• 10s: 50.660s → 3.732s (13.57×)
• 15s: 99.513s → 6.612s (15.05×)
Sol-H3 also scales across GPU counts:
• 4× B300: 2.918s / 6.993s / 12.542s for 5s / 10s / 15s (12.11–15.54×)
• 1× B300: 13.745s / 37.813s / 52.260s for 5s / 10s / 15s (9.45–14.29×)
All figures are medians of three measured runs after one warmup at 1344×768 and 24 FPS with stereo audio. Base H3 uses 50 scheduler points (49 DiT forwards); Sol-H3 uses four DiT forwards, so this is a full-profile comparison—not an attention-only runtime change. Sol-H3 uses Dense attention on 1× B300 and SOL with INT8 QKV / FP8 output transport on 4× / 8×. Timing includes text encoding, DiT denoising, and video/audio VAE decoding; model loading, compilation warmup, and final MP4 encoding are excluded.
Sol-H3 brings Sol-Engine × Sol-Attn into one full-stack runtime:
• dynamic sparse attention with no retraining
• fused norm, RoPE, MLP, and sparse-attention setup
• fused INT8 QKV / FP8 output communication across 8 GPUs
• parallel, batched VAE decoding
• precomputed AdaLN caching
Inside the stack:
• sparse-attention setup: 1.206 → 0.285 ms (−76.4%)
• VAE decode: 7.55 → 0.602 s
• ~24 GB memory freed per GPU
Any MiniMax-H3 few-step LoRA can plug into the same engine, and the code is deployment-friendly under Apache 2.0.
For us, the bigger milestone is crossing from “fast generation” into “faster than playback.” That opens the path toward continuous 24 FPS generation and truly interactive video systems.
We’re excited to partner with @reactorworld to release Sol-H3 and make it available as an API day-0.
Try it now on Reactor: reactor.inc/sandbox?model=fa…
🔗 nvlabs.github.io/Sana/Sol-En…
Amazing team effort—full credits in the blog.
@shanasaimoe @lawrence_cjs @yitongli165665 @haopengl33 @HaochengXiUCB @songhan_mit
