Why 128-bit address pointers make no sense: x86 still uses only 48 bits of 64
srimisra · x · 2026-11-15
- 32-bit hit a hard 4 GB wall, forcing the jump to 64-bit, which can address 18 quintillion bytes — nobody is near that wall.
- x86 doesn't even use all 64 bits: most machines use 48-bit addresses (256 TiB), newer chips go to 57-bit (128 PiB).
- 128-bit pointers would double pointer size, costing memory and cache, slowing pointer-heavy code for nothing.
- 128-bit already exists in CPUs — as vector registers (SSE to AVX-512). IBM i uses 16-byte pointers with a smaller address space; RISC-V's RV128I is only on paper.
Conclusion: we'll move to 128-bit when we actually run out of 64.
More from Infra
- Court Refuses to Dismiss DRAM Supply Conspiracy Suit Against Samsung, SK Hynix, Micron — SumitGup · 2026-11-15
- 2009 video shows Jensen Huang sizing CUDA at billions when NVDA traded at $0.2 — GregoryDiamos · 2026-11-15
- Ex-NVIDIA engineer recalls 2012: Jensen Huang backed the deep learning cluster behind Deep Speech — GregoryDiamos · 2026-11-15
- EU's Open Model EUROPA Eyes ~1T Params, Per-GPU Throughput Up 14x — mawcopolow · 2026-11-15
- Buy vs. Rent vs. API for Inference: Utilization Decides Your $/GPU-hour — Panther1537 · 2026-11-15
- AI data center demand to drive ~4,800 GW of new renewables by 2031 — ingliguori · 2026-11-15