3bb2fcc856
llamafile: powerpc: add FP16 MMA path for Q4/Q8 matmul ( #19709 )
...
Avoid xvi8ger4pp signed→unsigned bias correction by dequantizing Q4/Q8
inputs to FP16 and using FP16×FP16→FP32 MMA. This removes
post-processing overhead and improves performance.
Performance Impact:
1.5 ~ 2x improvement in PP_Speed for Q4 and Q8 Models,
measured with llama-bench and llama-batched-bench.
Q8 Model: granite-4.0-h-micro-Q8_0.gguf (from huggingface)
Q4 Model: Meta-Llama3-8b Q4 model (generated with llama-quantize from
f32 model)
llama-bench Q8 Model Results:
model size params backend threads test Base t/s Patch t/s
granitehybrid 3B Q8_0 3.16 GiB 3.19 B CPU 10 pp8 64.48 ± 4.72 73.99 ± 0.27
granitehybrid 3B Q8_0 3.16 GiB 3.19 B CPU 10 pp16 80.11 ± 0.32 112.53 ± 0.40
granitehybrid 3B Q8_0 3.16 GiB 3.19 B CPU 10 pp32 89.10 ± 0.27 152.95 ± 0.68
granitehybrid 3B Q8_0 3.16 GiB 3.19 B CPU 10 pp64 93.65 ± 0.25 187.83 ± 0.83
granitehybrid 3B Q8_0 3.16 GiB 3.19 B CPU 10 pp128 99.93 ± 0.02 201.32 ± 0.11
granitehybrid 3B Q8_0 3.16 GiB 3.19 B CPU 10 pp256 102.32 ± 0.40 208.32 ± 0.41
granitehybrid 3B Q8_0 3.16 GiB 3.19 B CPU 10 pp512 103.42 ± 0.40 209.98 ± 0.14
granitehybrid 3B Q8_0 3.16 GiB 3.19 B CPU 10 tg128 20.35 ± 0.01 19.57 ± 0.01
llama-bench Q4 Model Results:
model size params backend threads test Base t/s Patch t/s
llama 8B Q4_0 4.33 GiB 8.03 B CPU 10 pp8 34.77 ± 0.10 41.23 ± 0.08
llama 8B Q4_0 4.33 GiB 8.03 B CPU 10 pp16 40.81 ± 0.04 64.55 ± 0.15
llama 8B Q4_0 4.33 GiB 8.03 B CPU 10 pp32 44.65 ± 0.05 90.84 ± 0.22
llama 8B Q4_0 4.33 GiB 8.03 B CPU 10 pp64 47.49 ± 0.03 114.39 ± 0.11
llama 8B Q4_0 4.33 GiB 8.03 B CPU 10 pp128 49.29 ± 0.24 120.13 ± 0.19
llama 8B Q4_0 4.33 GiB 8.03 B CPU 10 pp256 49.77 ± 0.23 121.51 ± 0.11
llama 8B Q4_0 4.33 GiB 8.03 B CPU 10 pp512 49.89 ± 0.23 117.52 ± 0.10
llama 8B Q4_0 4.33 GiB 8.03 B CPU 10 tg128 13.40 ± 0.01 13.37 ± 0.00
Llama perplexity Results:
Model Base Final PPL Estimate Patch Final PPL Estimate
granite-4.0-h-micro-Q8_0 1.3862 +/- 0.04424 1.3868 +/- 0.04432
Meta-Llama3-8b Q4 1.3801 +/- 0.04116 1.3803 +/- 0.04116
Signed-off-by: Shalini.Salomi.Bodapati <Shalini.Salomi.Bodapati@ibm.com>
2026-02-19 14:28:53 +08:00
7afdfc9b84
ggml-cpu: Enable FP16 MMA kernels on PPC ( #19060 )
2026-01-27 11:52:34 +08:00
d34d5ca1e9
llamafile: add rvv support for sgemm kernels ( #18199 )
...
Co-authored-by: Rehan Qasim <rehan.qasim@10xengineers.ai>
2025-12-22 20:20:23 +02:00
3a0d10533a
Q4/Q8 Tiled Gemm Optimization. ( #16999 )
2025-12-05 19:41:51 +08:00
a6a58d6478
llamafile: PowerPC Sgemm Optimization ( #15558 )
...
This patch improves GEMM for FP32 Data Type on PowerPC
Implements GEMM on large blocks with configurable block size mc, nc, kc
(default: 256, 256, 256).
Packing Function optimized to access blocks as per memory layout.
GEMM Optimized to work on larger blocks.
Isolated Packing from GEMM Operations for better MMA utilization.
Verified functionality and correctness uing llama-cli and stand alone
test case (performs matmul and compares final mattrix C result with base).
Minor code refactoring changes:
Replace macro with inline function
Code Indent made consistent with 4 spaces
Performance Testing:
Observed 50% ~ 70% improvement in Prompt Processing Speed mesured using
llama-bench with Meta-Llama3-8B FP32 Model. Similar gains observed with
Mistral-7b-Instruct-v0.3 Model.
model Size Params Backend Threads Test Patch Base
llama 8B all F32 29.92 GiB 8.03 B CPU 20 pp512 98.58 60.3
llama 8B all F32 29.92 GiB 8.03 B CPU 20 pp1024 95.88 57.36
llama 8B all F32 29.92 GiB 8.03 B CPU 20 pp2048 85.46 53.26
llama 8B all F32 29.92 GiB 8.03 B CPU 20 pp4096 68.66 45.78
llama 8B all F32 29.92 GiB 8.03 B CPU 20 pp6144 57.35 40.44
25 ~ 30% improvement in llama-batched-bench with Metla-Llama3-8B in
Prompt Processing Speed for large prompts (256, 512, 1024, 2048, 4096)tokens with various batch
sizes ( 1, 2, 4, 8, 16)
Signed-off-by: Shalini Salomi Bodapati <Shalini.Salomi.Bodapati@ibm.com>
2025-08-26 23:35:25 +08:00
55c509daf5
ggml : refactor llamafile_sgemm PPC code ( #14673 )
...
Remove un-necessary templates from class definition and packing functions
Reduce deeply nested conditionals, if-else switching in mnapck function
Replace repetitive code with inline functions in Packing functions
2 ~ 7% improvement in Q8 Model
15 ~ 50% improvement in Q4 Model
Signed-off-by: Shalini Salomi Bodapati <Shalini.Salomi.Bodapati@ibm.com>
2025-07-14 16:16:42 +03:00
60ef23d6c1
ggml-cpu: enable IBM NNPA Vector Intrinsics ( #14317 )
...
* ggml-cpu: add nnpa compile flag
Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
(cherry picked from commit 4a9f60c2018d4a7987f90ff0d651622f58bbcbb801b929491b157f856c34157f856c3418d79e1a30fbb733451fbd288e8fa5https://github.com/ggml-org/llama.cpp/pull/14317#discussion_r2164176555
Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
* ggml: remove dependency on ggml-cpu from ggml-base
Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
* ggml-cpu: rename all fp16<->fp32 macros to prefix with ggml_cpu
ref: https://github.com/ggml-org/llama.cpp/pull/14317#discussion_r2164449406
Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
* ggml-cpu: remove mistaken fallback macro
fallback logic was already implemented but i was too sleepy to realise
Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
* ggml: move ggml_table_f32_f16 to ggml-cpu
ref: https://github.com/ggml-org/llama.cpp/pull/14317#discussion_r2164775006
Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
* ggml-cpu: move ggml_table_f32_f16 back to ggml-base due to ci failures
Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
* Revert "ggml-cpu: move ggml_table_f32_f16 back to ggml-base due to ci failures"
This reverts commit 32a35335649e40d984adhttps://github.com/ggml-org/llama.cpp/pull/14317#discussion_r2164775006
Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
(cherry picked from commit 9e40d984adf71b21d2f72dce119178
2025-06-25 23:49:04 +02:00
faed5a5f5d
llamafile : support s390x SIMD instruction set ( #14273 )
2025-06-19 11:48:54 +02:00
6adc3c3ebc
llama : add thread safety test ( #14035 )
...
* llama : add thread safety test
* llamafile : remove global state
* llama : better LLAMA_SPLIT_MODE_NONE logic
when main_gpu < 0 GPU devices are not used
---------
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
2025-06-16 08:11:43 -07:00
3f3769ba76
ggml : Enable MMA for BF16 in llamafile_sgemm ( #13148 )
...
This patch upstreams llamafile's cpu matrix multiplication kernels for ppc64le using MMA builtins for BF16 data type.
This change results in 9x - 40x gains
in total speed S t/s (ie all tokens/total time), across various batch sizes tested using llama-batched-bench benchmark.
The patch is tested with Meta-Lllama-3-8B,
and Mistral-7B models (BF16 models generated by using llama-quantize from corresponding FP32 models) on an IBM POWER10 machine.
Signed-off-by: Shalini Salomi Bodapati <Shalini.Salomi.Bodapati@ibm.com>
2025-05-02 19:53:12 +03:00
13731766db
llamafile : ppc64le GEMV forwarding for FP32. ( #12594 )
...
This patch enables usage of MMA when one of the
dimensions of the matrix(ie either M or N) is 1. This
is useful in case of token generation where N < 2.
The concept of 'GEMV Forwarding' is used where when one
of the matrix has a single row/column, the elements are
broadcasted, instead of using packing routine to prepack
the matrix elements.
This change results in 5% - 15% improvement in total
speed(ie all tokens/total time), across various batch
sizes. This is in comparision with the corresponding
dot product implementation.
The patch is tested with FP32 models of Meta-Lllama-3-8B,
Mistral-7B, Llama-2-7B-chat-hf on a IBM POWER10 machine.
Signed-off-by: Amrita H S <amritahs@linux.vnet.ibm.com>
2025-03-28 09:43:22 +02:00
c7b43ab608
llamafile : ppc64le MMA implementation for Q4_0. ( #12489 )
...
This change upstreams llamafile's cpu matrix
multiplication kernels for ppc64le ISA using MMA
builtins. This patch handles matrix multiplication
between quantised datatypes, block_q4_0 and
block_q8_0.
This change results in 5% - 50% improvement
in total speed(ie all tokens/total time), across
various batch sizes.
The patch is tested with Meta-Lllama-3-8B,
Mistral-7B, Llama-2-7B-chat-hf models on a
IBM POWER10 machine.
Signed-off-by: Amrita H S <amritahs@linux.vnet.ibm.com>
2025-03-27 08:51:47 +02:00
8a8c4ceb60
llamafile: use member variable instead of constant for iq4nlt ( #11780 )
2025-02-13 18:05:04 +01:00
8cef75c743
llamafile : ppc64le MMA INT8 implementation ( #10912 )
...
This change upstreams llamafile's cpu matrix
multiplication kernels for ppc64le using MMA
builtins for quantised int8 datatype.
This change results in 10% - 70% improvement
in total speed(ie all tokens/total time), across
various batch sizes.
The patch is tested with Meta-Lllama-3-8B,
Mistral-7B, Llama-2-7B-chat-hf models on a
IBM POWER10 machine.
Signed-off-by: Amrita H S <amritahs@linux.vnet.ibm.com>
2025-01-08 12:54:19 +02:00
0827b2c1da
ggml : fixes for AVXVNNI instruction set with MSVC and Clang ( #11027 )
...
* Fixes for clang AVX VNNI
* enable AVX VNNI and alder lake build for MSVC
* Apply suggestions from code review
---------
Co-authored-by: slaren <slarengh@gmail.com>
2024-12-31 15:23:33 +01:00
2cd43f4900
ggml : more perfo with llamafile tinyblas on x86_64 ( #10714 )
...
* more perfo with llamafile tinyblas on x86_64.
- add bf16 suport
- change dispache strategie (thanks:
https://github.com/ikawrakow/ik_llama.cpp/pull/71 )
- reduce memory bandwidth
simple tinyblas dispache and more cache freindly
* tinyblas dynamic dispaching
* sgemm: add M blocs.
* - git 2.47 use short id of len 9.
- show-progress is not part of GNU Wget2
* remove not stable test
2024-12-24 18:54:49 +01:00
9177484f58
ggml : fix arm build ( #10890 )
...
* ggml: GGML_NATIVE uses -mcpu=native on ARM
Signed-off-by: Adrien Gallouët <angt@huggingface.co>
* ggml: Show detected features with GGML_NATIVE
Signed-off-by: Adrien Gallouët <angt@huggingface.co>
* remove msvc support, add GGML_CPU_ARM_ARCH option
* disable llamafile in android example
* march -> mcpu, skip adding feature macros
ggml-ci
---------
Signed-off-by: Adrien Gallouët <angt@huggingface.co>
Co-authored-by: Adrien Gallouët <angt@huggingface.co>
2024-12-18 23:21:42 +01:00
7cc2d2c889
ggml : move AMX to the CPU backend ( #10570 )
...
* ggml : move AMX to the CPU backend
---------
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
2024-11-29 21:54:58 +01:00
db4cfd5dbc
llamafile : fix include path ( #0 )
...
ggml-ci
2024-11-16 20:36:26 +02:00
ae8de6d50a
ggml : build backends as libraries ( #10256 )
...
* ggml : build backends as libraries
---------
Signed-off-by: Xiaodong Ye <xiaodong.ye@mthreads.com>
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
Co-authored-by: R0CKSTAR <xiaodong.ye@mthreads.com>
2024-11-14 18:04:35 +01:00
shalinib-ibm
Taimur Ahmad
Aaron Teo
Diego Devesa
amritahs-ibm
Jeffrey Morgan
Srihari-mcw
Djip007
Georgi Gerganov