SVE 128 bit Implementation of gemm_q4_k_8x8_q8_k kernel

[anubhavfujitsu]

Overview

This PR adds 128-bit SVE (Scalable Vector Extensions) support for the ggml_gemm_q4_K_8x8_q8_K kernel using i8mm and vector instructions.

Additional information

By running a Q4_K_M quantized model of Llama-3.2-3B-Instruct, I have checked the generation output.
I also verified that the perplexity matches between the NEON and SVE implementations.

SVE 128 (this PR)	NEON (OSS)
12.5089 +/- 0.69243	12.5089 +/- 0.69243

The command used to measure the perplexity is:

./llama-perplexity -m model.gguf -f wikitext-2-raw/wiki.test.raw --chunks 10

Performance Check

This PR Improves the Prompt Eval time (TTFT) of LLM Inference between ~4 to 5%, as compared to NEON (Original Version) for Llama-3.2-3B-Instruct-Q4_K_M.

The performance was measured on Nvidia Grace machine.
Performance is improved as follows. The value is tokens/second.

Task	Threads	NEON (Original)	SVE 128 (This PR)	Speedup
PP128	4	77.45	81.64	1.0541
PP128	8	141.37	149.71	1.0590
PP128	16	267.43	279.88	1.0466
PP128	32	436.96	455.90	1.0433

Requirements

• I have read and agree with the contributing guidelines
• AI usage disclosure: No

[ggml-gh-bot]

Hi @anubhavfujitsu, thanks for your contribution!

Per our contribution guidelines, the automated PR checker found the following issue(s) that need your attention:

• Multiple open PRs from a new contributor: We limit new contributors (those without a previously merged PR) to 1 open PR at a time. You currently have 2 open PRs.

---

Please note that maintainers reserve the right to make final decisions on PRs. If you believe there is a mistake, please comment below.