Metraipolis › archive › dc_d2075c9ca0cc
The Missing Fourth Term for the Emulation Tensor Memory Equilibrium (TME) Model: The Residue Deconstruction Cost — arXiv:2610.10924
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.10924
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-09T05:55:59.549Z
- SHA-256 of the archived text
05fc7bc5a60f66f75487b138dc22edd08adaf391a77445cb91a7e2f234c2febf- Stable URL of this page
https://metraipolis.com/documento/dc_d2075c9ca0cc
How to cite
The Missing Fourth Term for the Emulation Tensor Memory Equilibrium (TME) Model: The Residue Deconstruction Cost — arXiv:2610.10924 — arXiv (en), https://arxiv.org/abs/2610.10924 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d2075c9ca0cc (Metraipolis archive dc_d2075c9ca0cc, sha256 05fc7bc5a60f66f75487b138dc22edd08adaf391a77445cb91a7e2f234c2febf)
BibTeX
@misc{dcd2075c9ca0cc,
title = {The Missing Fourth Term for the Emulation Tensor Memory Equilibrium (TME) Model: The Residue Deconstruction Cost — arXiv:2610.10924},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.10924},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d2075c9ca0cc; sha256 05fc7bc5a60f66f75487b138dc22edd08adaf391a77445cb91a7e2f234c2febf},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.10924. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
The Tensor-Memory Equilibrium (TME) model of "FP8 is All You Need (Part 1)" calculates the execution time of Ozaki Scheme II emulation of fp64 as the maximum of a tensor-core term and a High-Bandwidth Memory (HBM) traffic term, plus a per-output reconstruction term. However, it omits the per-input deconstruction cost: every streamed fp64 operand must be scaled, rounded, and reduced modulo each of the $r$ moduli on SIMT pipes before any matrix multiply can issue. In this note we add this fourth term, calibrate its constant from the cuBLAS emulation path, and derive a closed-form operational-intensity threshold $\mathrm{OI}^{*} = c_q r P_{\mathrm{fp64}}/(8P_{\mathrm{int}})$ below which emulation cannot match native fp64 regardless of tensor-core throughput. On the NVIDIA B300 GPU the threshold is $\mathrm{OI}^{*}\approx 0.56$ FLOP/B. As a result, GEMV, SpMV, and low-batch GEMV, which are the memory-bound kernels the original paper claims to accelerate, are limited to 0.3-0.9x of native performance, and the 7-point stencil to 1.8x rather than the claimed 3.1x. Dense GEMM is unaffected as expected. We also show that precomputing and storing the residues moves the same cost into the bandwidth term, and we state the instruction count that an implementation would have to achieve to invalidate the bound.
Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Hardware Architecture (cs.AR) ; Artificial Intelligence (cs.AI); Performance (cs.PF).
Pagina del lavoro: https://arxiv.org/abs/2610.10924 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-09T05:47:01.396Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).