Nicheloom

Market intelligence for builders — see what's gaining traction before it's crowded.

splish

Splish: an unofficial fork of Inco's Splash, tuned for Apple M5 GPUs. Up to 1.5x faster (about 1.25x at one request) on Qwen3.8-27B. Apache-2.0

Details

External ID
1388980448
Source
GITHUB
Company
—
Product
splish
Website domain
github.com
Launched
Sept. 26, 2026
Cohort
—
Upvotes
12
Upvotes percentile
0.3866256725595696
Tags
apple-silicon, llm-inference, m5, metal, qwen, speculative-decoding
Fetched at
Sept. 30, 2026, 5:02 p.m.
Updated at
Sept. 30, 2026, 5:02 p.m.

Enrichment

Theme
gpu compute and acceleration tools
Vertical
Horizontal
Function
Model & infra
Audience
Developer
AI stance
AI-native
Project type
Hobby / open-source project
Normalized one-liner
gpu inference engine fork for apple silicon
Manually corrected
False

Could you build this?

No Optimizing LLM inference engines and writing custom GPU kernels tuned for specific Apple Silicon architectures requires low-level systems programming and deep GPU microarchitecture expertise.

What it would actually take: Building an optimized inference fork requires expertise in Apple Metal, Metal Performance Shaders (MPS), or low-level C++/Rust GPU kernel writing. The developer must profile memory bandwidth, threadgroup memory usage, SIMD-group matrix operations, and execution pipelines on Apple Silicon chips. Vibe coding cannot generate novel, highly tuned GPU assembly or low-level kernel optimizations without hallucinating or degrading performance.

Competitors

Other products that read as similar to this one — 187 launches clear the similarity bar, closest 8 shown.

Attention rank: #126 of 188 (itself plus its competitors, highest first — normalized so YC and Product Hunt are compared fairly).

Launched 323 days after the earliest competitor.

Other launches for this product

Same idea, different domain

Nobody's really built a model & infra tool for Fintech yet.