#PCIe bandwidth
74 related articles
·6 min
How the Creator of Redis Squeezed a 284-Billion-Parameter DeepSeek Into an Ordinary Laptop
Redis author antirez built the pure-C inference engine DS4 "Dwarf Star," compressing DeepSeek V4 Flash from 500GB to 80.8GB via asymmetric quantization for local deployment on 128GB unified memory at 26.7 tokens/sec.
Read more →
·7 min
Alibaba Open-Sources 2.4T Qwen Model as Chinese AI Sweeps the Global Competitive Landscape
Alibaba open-sources a 2.4 trillion parameter Qwen model and launches the Qwen Token Plan. Chinese models surge, Kimi K3 tops global rankings, and China's AI is reshaping the global competitive landscape.
Read more →
·3 min
V100 GPU Showdown: 4×16G PCIe vs. 2×32G SXM — Which Gives You More Bang for Your Buck?
Benchmarking 4×V100 16G PCIe vs. 2×V100 32G SXM adapter for local LLM inference. Prefill speed, decode speed, power limits, and bandwidth bottlenecks analyzed.
Read more →
·2 min
MiniMax M3 Local Deployment: 23B Active Parameters ≠ 23GB VRAM — A Complete Guide to MoE Architecture
MiniMax M3 is a 428B MoE model. Its 23B active parameters reflect per-token compute, not VRAM needs. Learn the MoE trade-offs, quantization options, and deployment paths to avoid the most common misconception.
Read more →
AirLLM: How a 4GB GPU Can Run a 70B Mo…
·2 min
AirLLM: How a 4GB GPU Can Run a 70B Model — Core Principles and Use Cases
AirLLM is an open-source project that uses layer-by-layer inference to run 70B LLMs on a single 4GB GPU. Learn how it works, its tradeoffs, and ideal use cases.
Read more →
Apple M7 Ultra Leaked: Can 1.5TB Unifi…
·2 min
Apple M7 Ultra Leaked: Can 1.5TB Unified Memory Challenge NVIDIA Blackwell?
Apple's M7 Ultra chip rumored to feature 1.5TB unified memory targeting NVIDIA Blackwell-level AI performance. A deep dive into UMA advantages, what "matching Blackwell" really means, and Apple's AI strategy.
Read more →
·2 min
How Large a Local LLM Can You Run on a Laptop with 8GB VRAM? A Configuration Guide
Complete guide to running local LLMs on a laptop with 8GB VRAM: real usable memory, quantization estimation, Q4 7B/8B model recommendations, Ollama setup, GPU offloading, and agent development tips.
Read more →
Apple M7 Ultra Chip Leaked: Can 1.5TB …
·2 min
Apple M7 Ultra Chip Leaked: Can 1.5TB Unified Memory Reshape Local AI Inference?
Reddit leaks suggest Apple's M7 Ultra chip could feature up to 1.5TB unified memory. We analyze the architecture, pricing debate, bandwidth limits, and ecosystem trade-offs for local LLM inference.
Read more →
Building an eGPU for Local LLM on a $1…
·2 min
Building an eGPU for Local LLM on a $1,000 Budget: A Practical Buying and Setup Guide
Building an eGPU for local LLMs on a $1,000 budget? This guide covers GPU selection, dock costs, RTX 3090 vs 3060 value, and top tools like Ollama and llama.cpp.
Read more →
JAX Host Offloading: A Practical Guide…
·2 min
JAX Host Offloading: A Practical Guide to Breaking Through LLM Training Memory Bottlenecks
Memory capacity is the core bottleneck in LLM training. This guide explores JAX-based host offloading — covering optimizer state offloading, activation strategies, PCIe bandwidth trade-offs, and how it complements activation recomputation.
Read more →
·2 min
Local Image-to-3D on Apple Silicon: 20 Seconds, 2GB RAM — Modelr Open-Source App Tested
A developer ports Tencent's Hunyuan3D to Apple MLX. M4 Max benchmarks: shape generation in 20s, 5.6GB peak RAM. Quantized versions run on iPhone. Modelr is the first local image-to-3D app for Apple Silicon.
Read more →
How Fixing Three Bugs Made Qwen3.5-122…
·3 min
How Fixing Three Bugs Made Qwen3.5-122B Run Reliably on Mac Studio
A developer fixed three critical bugs to make Qwen3.5-122B run reliably as a daily driver on Mac Studio. Explore memory management, inference stability, and quantization precision.
Read more →
·4 min
GPT-5.6 In-Depth Analysis: Model Architecture, Performance Benchmarks, and a Complete Guide to the Agentic Era
A comprehensive analysis of GPT-5.6: the Sol/Terra/Luna product-line positioning, MoE architecture, 1.5M token context, coding agent benchmarks, and the shift from Chatbot to Agent.
Read more →
·3 min
SlickToken: A Deep Dive into GPU Fleet and Agentic Workflow Planning
SlickToken is a GPU fleet and agentic workflow planning tool for AI teams, supporting offline simulation, load testing, and capacity planning—no internet required to protect enterprise data.
Read more →
·4 min
AMD Ryzen AI Halo Hands-On: Can x86 Architecture Challenge NVIDIA's DGX Spark?
In-depth review of the AMD Ryzen AI Halo mini AI box: powered by the Ryzen AI Max Plus 395 (Strix Halo) chip with 128GB unified memory, priced at $4,000. Compared against NVIDIA's DGX Spark across token generation, prefill speed, and x86 advantages.
Read more →
·3 min
Apple Executive Explains: Why the Mac Mini Has Become AI Developers' Top Choice for Local Inference
An Apple chip executive explains the surge in Mac Mini demand among AI developers: unified memory architecture breaks the VRAM bottleneck, superior energy efficiency enables long-term local deployment—the Mac Mini is becoming a top pick for local LLM inference.
Read more →
·4 min
A Guide to Avoiding Pitfalls in PCA Visualization of Self-Supervised Vision Models: Why the Colors Are Meaningless
Running self-supervised vision models (SSL) on a MacBook CPU isn't hard. This article reveals the core misconception of PCA visualization through ViT-S experiments: colors can't convey semantics across images, and changing resolution reverses hues entirely.
Read more →
·6 min
DeepSeek Builds Its Own AI Chips: The Strategic Logic and Challenges Behind Compute Independence
DeepSeek is entering AI chip development, targeting compute autonomy. This article analyzes its motivations, software-hardware synergy, chip R&D challenges, and impact on China's AI vertical integration.
Read more →
·3 min
TB/s Memory Tier Accelerates SGLang: How X-Mem Reduces TTFT by 6.7x
Netpreme integrates X-Mem™ MPU into SGLang HiCache, achieving up to 6.7× TTFT reduction and 33–50% TPS gains at 98% prefix cache hit rates. Here's the technical breakdown.
Read more →
·4 min
Demystifying Unified Memory Architecture: How Mini PCs Run 70B Large Models
Why can a mini PC with unified memory run a 70B model while an RTX 4090 can't? A deep dive into the VRAM wall and unified memory architecture for smarter local AI hardware choices.
Read more →