Nicheloom visits: 58,280 all-time · 7,637 in the last 7 days

The opportunity tracker for new startups.

Today: 1,008 launches · Niche of the day: specialized calculation tools · See what launched today →

TRiP

a complete transformer engine in C built from scratch just by me

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 Apr 30, 2026 · 38 upvotes (top 19%) · 0 views

Niche: machine learning inference engines · 108 launches, 8 in the last 7 days ▼ -27% · Explore this niche →

Category and tags

  • Horizontal
  • Model & infra
  • Developer
  • AI-native
  • Hobby / open-source project

Visit website ↗ Source listing ↗

Could you build this?

No Implementing a complete transformer engine from scratch in pure C requires deep specialized knowledge of tensor math, memory layout, matrix multiplication optimization, and neural network mechanics without standard high-level ML libraries.

What it would actually take: Building a custom transformer engine in C requires implementing linear algebra operations (BLAS/GEMM), tensor slicing, backpropagation/forward passes, attention mechanisms (KV-cache, RoPE, Softmax), and tokenizers entirely in native code. It also requires hand-optimized memory management and SIMD/AVX or CUDA intrinsics for usable performance. Deep systems programming and low-level numerical computing expertise are required.

Discussion

6 comments analyzed.

Concerns raised: No benchmark comparisons to other projects, CPU-only implementation (not GPU), Performance data not provided, Flash attention not implemented

Feature requests: Matmul kernel hand-optimization suggestions welcome

Competitors

Other products that read as similar to this one : 3261 launches clear the similarity bar, closest 8 shown.

Attention rank: #602 of 3262 (itself plus its competitors, highest first, normalized so YC and Product Hunt are compared fairly).

Launched 183 days after the earliest competitor.

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.