Qwen3.6-35B-A3B on a 16 GB M1 Pro with SSD-streamed MoE
Get picks like this daily. Today's launch numbers, the top launches, AI and tech news, and a daily opportunity pick, straight to your inbox.
Want to reach these readers? Sponsor an issue.
HN · launched Jul 17, 2026 · 24 upvotes (top 25%) · 0 views
Niche: llm inference and deployment tools · 188 launches, 6 in the last 7 days · Explore this niche →
Category and tags
- Horizontal
- Model & infra
- Developer
- AI-native
- Hobby / open-source project
Visit website ↗ Source listing ↗
More like this: all 188 launches in llm inference and deployment tools →
Could you build this?
No Streaming active mixture-of-experts (MoE) weights from an NVMe SSD into Apple Silicon unified memory at token generation speed requires low-level kernel bypass, Metal/MPS programming, and custom memory-mapping paging kernels.
What it would actually take: Requires low-level C++/Metal implementations using POSIX mmap or direct asynchronous I/O (aio) to stream selectively gated expert tensors from disk into VRAM just-in-time during forward passes. The critical difficulty is hiding high SSD read latency behind attention computation and designing custom weight quantizations optimized for Apple Silicon memory bus bandwidth.
Discussion
4 comments analyzed.
Concerns raised: Performance slower than expected, M1 chip compatibility requirements unclear
Feature requests: Support for cached pre-filled system prompt KV-cache with prefix hash in requests, Remote INFER protocol support
Competitors
Other products that read as similar to this one : 696 launches clear the similarity bar, closest 8 shown.
Attention rank: #161 of 697 (itself plus its competitors, highest first, normalized so YC and Product Hunt are compared fairly).
Launched 258 days after the earliest competitor.
- qwen3.6-35b-a3b-144T-S · github · 2026-09-20 · 15 upvotes · similarity 0.65
- Samosa Chat · hn · 2026-07-15 · 6 upvotes · similarity 0.63
- Run an 80B Qwen in 4.3 GB of RAM on a Mac, and a 35B on an iPhone · hn · 2026-08-03 · 312 upvotes · similarity 0.56
- glyd-condor · github · 2026-10-14 · 29 upvotes · similarity 0.54
- Running 104GB Qwen3.8-Flash-Next on 48GB Mac with at ~12 tok/s · hn · 2026-09-01 · 240 upvotes · similarity 0.54
- Fine-tune an 8B model on a 4 GB laptop GPU · hn · 2026-08-04 · 139 upvotes · similarity 0.53
- lean-qwenflash · github · 2026-10-24 · 48 upvotes · similarity 0.53
- S3 compatible store with 1M IOPS(4K-R,p99~5ms), BYOC in 5min with rust · hn · 2025-12-07 · 23 upvotes · similarity 0.50
The niche, labels, vibe code rating and competitor comparison on this page are analysis by Nicheloom. If you use them, please credit nicheloom.com. How you may use our data.