--- id: "/tags/infrastructure-chips-efficiency/" title: "Infrastructure Chips Efficiency" creator: null creator_url: null creator_x: null url: "https://skillissue.dev/tags/infrastructure-chips-efficiency/" level: null level_max: null category: null subcategory: null region: null license: null license_scope: null access: null models: null source_date: null date_added: null resource_kind: null verification_basis: null access_evidence: null domain_top: null domain_sub: null last_verified: null --- # Infrastructure Chips Efficiency [Beginner](/levels/beginner/index.md) [Intermediate](/levels/intermediate/index.md) [Advanced](/levels/advanced/index.md) [Expert](/levels/expert/index.md) 122 resources · page 1 of 4 - ### [Apple machine learning research](/model-training/training-infrastructure/apple-machine-learning-research/index.md) Creator not supplied; see original source. Apple Machine Learning Research publications: papers on speech, language, vision, privacy and health. Not applicable (entity) · unknown · unknown resource · en - ### [Apple MLX](/model-training/training-infrastructure/apple-mlx/index.md) Creator not supplied; see original source. Apple MLX GitHub repo: array framework for machine learning on Apple silicon. Not applicable (entity) · unknown · unknown resource · en - ### [OpenInference](/model-training/training-infrastructure/arize-openinference/index.md) Creator not supplied; see original source. Arize OpenInference GitHub repo: OpenTelemetry instrumentation and semantic conventions for LLM tracing. Not applicable (entity) · unknown · unknown resource · en - ### [Truss](/model-training/training-infrastructure/baseten-truss/index.md) Creator not supplied; see original source. Baseten Truss GitHub repo: packaging and serving ML models in production. Not applicable (entity) · unknown · unknown resource · en - ### [checkpoint-engine](/model-training/training-infrastructure/checkpoint-engine/index.md) Creator not supplied; see original source. Moonshot checkpoint-engine GitHub repo: middleware for updating model weights in running LLM inference engines. Not applicable (entity) · unknown · unknown resource · en - ### [Colossal-AI](/model-training/training-infrastructure/colossal-ai/index.md) Creator not supplied; see original source. HPC-AI Tech Colossal-AI GitHub repo: parallel training system for large AI models. Not applicable (entity) · unknown · unknown resource · en|zh - ### [DeepGEMM](/model-training/training-infrastructure/deepgemm/index.md) Creator not supplied; see original source. DeepSeek DeepGEMM GitHub repo: GPU matrix multiplication kernel library including FP8 GEMMs. Not applicable (entity) · unknown · unknown resource · en - ### [DeepGEMM-Ascend](/model-training/training-infrastructure/deepgemm-ascend/index.md) Creator not supplied; see original source. DeepSeek DeepGEMM-Ascend GitHub repo: matrix multiplication kernels for Huawei Ascend hardware. Not applicable (entity) · unknown · unknown resource · en - ### [DeepSeek Open Infra Index](/model-training/training-infrastructure/deepseek-open-infra-index/index.md) Creator not supplied; see original source. DeepSeek Open Infra Index GitHub repo: index of released infrastructure tools such as FlashMLA, DeepEP, DeepGEMM and 3FS. Not applicable (entity) · unknown · unknown resource · en - ### [DeepSpeed](/model-training/training-infrastructure/deepspeed/index.md) Creator not supplied; see original source. DeepSpeed GitHub repo: deep learning optimisation library for large-scale training and inference, origin of ZeRO. Not applicable (entity) · unknown · unknown resource · en - ### [py-why/dowhy](/model-training/training-infrastructure/dowhy/index.md) Creator not supplied; see original source. PyWhy DoWhy GitHub repo: Python library for causal inference with explicit assumptions and refutation tests. Not applicable (entity) · unknown · unknown resource · en - ### [Etched](/model-training/training-infrastructure/etched/index.md) Creator not supplied; see original source. Etched: company building transformer-specific AI inference chips, racks and software. Not applicable (entity) · unknown · unknown resource · en - ### [FastDeploy](/model-training/training-infrastructure/fastdeploy/index.md) Creator not supplied; see original source. PaddlePaddle FastDeploy GitHub repo: inference and deployment toolkit for LLMs and VLMs. Not applicable (entity) · unknown · unknown resource · en|zh - ### [Fireworks cookbook](/model-training/training-infrastructure/fireworks-cookbook/index.md) Creator not supplied; see original source. Fireworks cookbook GitHub repo: recipes for fine-tuning, function calling and building with Fireworks AI. Not applicable (entity) · unknown · unknown resource · en - ### [FlashMLA](/model-training/training-infrastructure/flashmla/index.md) Creator not supplied; see original source. DeepSeek FlashMLA GitHub repo: multi-head latent attention decoding kernels for Hopper GPUs. Not applicable (entity) · unknown · unknown resource · en - ### [FlashQLA](/model-training/training-infrastructure/flashqla/index.md) Creator not supplied; see original source. QwenLM FlashQLA GitHub repo: linear attention kernel library built on TileLang. Not applicable (entity) · unknown · unknown resource · en - ### [Huawei Ascend](/model-training/training-infrastructure/huawei-ascend/index.md) Creator not supplied; see original source. Huawei Ascend community site: Ascend AI processors, CANN software stack, frameworks and developer tools. Not applicable (entity) · unknown · unknown resource · zh - ### [KTransformers](/model-training/training-infrastructure/ktransformers/index.md) Creator not supplied; see original source. KTransformers GitHub repo: framework for heterogeneous CPU-GPU inference of large LLMs. Not applicable (entity) · unknown · unknown resource · en - ### [Kunlunxin (昆仑芯)](/model-training/training-infrastructure/kunlunxin/index.md) Creator not supplied; see original source. Kunlunxin site: Baidu-affiliated maker of AI accelerator chips (in Chinese). Not applicable (entity) · unknown · unknown resource · zh - ### [Liger Kernel](/model-training/training-infrastructure/liger-kernel/index.md) Creator not supplied; see original source. LinkedIn Liger Kernel GitHub repo: Triton kernels for LLM training. Not applicable (entity) · unknown · unknown resource · en - ### [Liquid AI research](/model-training/training-infrastructure/liquid-ai-research/index.md) Creator not supplied; see original source. Liquid AI research page: papers and technical reports on Liquid Foundation Models. Not applicable (entity) · unknown · unknown resource · en - ### [llm.c](/model-training/training-infrastructure/llm-c/index.md) Creator not supplied; see original source. llm.c GitHub repo (Andrej Karpathy): LLM training in plain C and CUDA. Not applicable (entity) · unknown · unknown resource · en - ### [LMDeploy](/model-training/training-infrastructure/lmdeploy/index.md) Creator not supplied; see original source. InternLM LMDeploy GitHub repo: toolkit for compressing, deploying and serving LLMs. Not applicable (entity) · unknown · unknown resource · en|zh - ### [llama.cpp](/model-training/training-infrastructure/local-on-device-ai/index.md) Creator not supplied; see original source. llama.cpp GitHub repo: LLM inference in C/C++ for local and on-device hardware. Not applicable (entity) · unknown · unknown resource · en - ### [Megatron-LM](/model-training/training-infrastructure/megatron-lm/index.md) Creator not supplied; see original source. NVIDIA Megatron-LM GitHub repo: research code for training transformer models at scale with tensor and pipeline parallelism. Not applicable (entity) · unknown · unknown resource · en - ### [MiniCPM](/model-training/training-infrastructure/minicpm-technical-repository/index.md) Creator not supplied; see original source. OpenBMB MiniCPM GitHub repo: on-device small language models with deployment resources. Not applicable (entity) · unknown · unknown resource · zh - ### [MiniCPM-V](/model-training/training-infrastructure/minicpm-v-technical-repository/index.md) Creator not supplied; see original source. OpenBMB MiniCPM-V GitHub repo: on-device multimodal models MiniCPM-V and MiniCPM-o for vision and speech. Not applicable (entity) · unknown · unknown resource · zh - ### [MLX / mlx-lm](/model-training/training-infrastructure/mlx-mlx-lm/index.md) Creator not supplied; see original source. Apple mlx-lm GitHub repo: Python package for running and fine-tuning LLMs with MLX. Not applicable (entity) · unknown · unknown resource · en - ### [MobileCLIP research implementation](/model-training/training-infrastructure/mobileclip-research-implementation/index.md) Creator not supplied; see original source. Apple ml-mobileclip GitHub repo: official implementation of MobileCLIP efficient image-text models. Not applicable (entity) · unknown · unknown resource · en - ### [Mooncake](/model-training/training-infrastructure/mooncake/index.md) Creator not supplied; see original source. Mooncake GitHub repo: KV-cache-centric disaggregated serving platform used for Kimi. Not applicable (entity) · unknown · unknown resource · en - ### [Motif Technologies](/model-training/training-infrastructure/motif-technologies/index.md) Creator not supplied; see original source. Motif Technologies GitHub organisation: training and serving code for its Korean models. Not applicable (entity) · unknown · unknown resource · en - ### [Multiverse Computing model registry](/model-training/training-infrastructure/multiverse-computing-model-registry/index.md) Creator not supplied; see original source. Multiverse Computing Hugging Face organisation: compressed model cards such as HyperNova. Not applicable (entity) · unknown · unknown resource · en [Next page](/tags/infrastructure-chips-efficiency/page/2/index.md) [Plain-text Markdown copy](/tags/infrastructure-chips-efficiency/index.md)