Ainiketan ainiketan.in
Papers Dictionary This Week Learning Paths ∑ Playground
Support Us
Papers Dictionary This Week Learning Paths ∑ Playground Support Us ☕
← Dictionary / SRAM (Static RAM)

SRAM (Static RAM)

Appears in 1 paper

Tiny, ultra-fast on-GPU cache (e.g., 192KB per core).

As used in Paper 21 — Mamba: Linear-Time Sequence Modeling with Selective State Spaces →

Tiny, ultra-fast on-GPU cache (e.g., 192KB per core). All computation happens here. If your algorithm doesn't fit in SRAM, you pay HBM-access penalties. Flash Attention and Mamba both optimise for this.

Appears in papers

Paper 21 — Mamba: Linear-Time Sequence Modeling with Selective State Spaces →
Browse Dictionary
← All terms A–Z
Share
WhatsApp
Ainiketan

Where India learns AI — deeply, freely, together.

जहाँ हर जिज्ञासु AI सीखे — खुलकर, गहराई से, साथ में।

Free forever No ads No login Open source

Learn

All 24 Papers Math Tutorials Dictionary Learning Paths This Week in AI

Community

Student Journal Soon Paper Club Soon Research Questions Soon Mentor Network Soon Teacher Packs Soon

Site

About Scholarship Fund Impact Corrections Support Us ☕ Terms & Copyright
☕
Buy us a chai

This site is free forever. If it helped you, support it for others.

GitHub Sponsors →
Weekly digest

5 things in AI every week. Plain English. Free.

© 2026 Ainiketan · Built for India, for free, forever · Suggest a correction

Content license: CC BY 4.0 · Hosted on Vercel · Privacy-friendly analytics (no cookies)

All summaries are original writing by Ainiketan — we link to sources and do not reproduce copyrighted text. Copyright concerns: askainiketan@gmail.com · Terms & Copyright