--- title: Running Llama on Windows 98 description: Discover how EXO Labs ran advanced AI models on a vintage Windows 98 machine, showcasing innovation in resourceful computing. image: https://netglass.io/hubfs/AI-Generated%20Media/Images/The%20image%20features%20a%20vintage%20Windows%2098%20Pentium%20II%20computer%20setup%2c%20complete%20with%20a%20CRT%20monitor%20displaying%20lines%20of%20code.jpeg --- [![Netglass Logo](https://netglass.io/hubfs/NetglassLogoTxtWhiteTrans4.svg)](https://netglass.io) - [Philosophy](https://netglass.io) ## [The Netglass "way" Our approach is based on the following tenets.](https://netglass.io) ### [Cloud Strategy Cloud strategy that orchestrates your staff processes and customer preferences](https://netglass.io#strategy) ### [Cloud Security Cloud security, tailored to the way you do business](https://netglass.io#security) ### [Cloud Cost Right-sizing your cloud can help reduce management complexity, and most importantly, cost.](https://netglass.io#cost) - [Cloud](https://netglass.io/cloud) ### [Cloud Services Modernize your operations with these efficiency improving services](https://netglass.io/cloud) ### [Business Continuity / Disaster Recovery Ensure your business stays resilient with our comprehensive disaster recovery and continuity solutions.](https://netglass.io/cloud#bcdr) ### [Phone system modernization Transform your communication with our cutting-edge phone system modernization services.](https://netglass.io/cloud#phone) ### [Network optimization and monitoring Improve and secure your network with our expert planning and real-time monitoring services.](https://netglass.io/cloud#network) ### [Lift and Shift Efficiently migrate your applications to the cloud with our Lift and Shift service.](https://netglass.io/cloud#lift-and-shift) ### [Google Workspace & Microsoft 365 migrations Effortlessly migrate to Google Workspace or Microsoft 365 with our expert, hassle-free services.](https://netglass.io/cloud#office) - [Software](https://netglass.io/software) ## [Experienced software design Leverage our seasoned expertise in software design to create powerful, user-friendly applications.](https://netglass.io/software) ### [Web application scaling Scale your web applications smoothly and efficiently with our expert solutions.](https://netglass.io/software#scaling) ### [Hosting cost reduction Cut hosting costs without compromising performance with our strategic optimization services.](https://netglass.io/software#hosting) ### [Cloud migration of legacy applications Modernize your legacy applications, reach larger audiences, and improve security.](https://netglass.io/software#legacy) ### [Database design Build robust, scalable databases with our expert design services.](https://netglass.io/software#database) ### [Load testing Deep experience in extreme load generation gives you hard data on your system capacity](https://netglass.io/software#load) - [Security](https://netglass.io/security) ## [Security by design Bake robust security into your development processes with our 'security by design' approach.](https://netglass.io/security) ### [Secure SDLC Develop secure software from start to finish with our Secure Software Development Life Cycle (SDLC) services.](https://netglass.io/security#ssdlc) ### [Security orchestration Eliminate "threat overload" by developing workable detection and prioritization strategies](https://netglass.io/security#orchestration) ### [Penetration testing Identify and fix vulnerabilities with our comprehensive penetration testing services.](https://netglass.io/security#pentest) ### [Attack surface reduction Minimize your risk with our expert attack surface reduction strategies and solutions.](https://netglass.io/security#attack) - [Contact](https://netglass.io/contact) - [Blog](https://netglass.io/blog) AI # Running Llama on Windows 98 Discover how EXO Labs ran advanced AI models on a vintage Windows 98 machine, showcasing innovation in resourceful computing. [Phil Emmerson](https://netglass.io/blog/author/phil) Jan 2, 2025 --- Hello and welcome to the Netglass.io blog! We're excited to embark on this journey of sharing insights and innovations in the IT realm with you. Recently, we came across [an intriguing experiment by EXO Labs](https://blog.exolabs.net/day-4), where they successfully ran a modern AI language model, Llama, on a vintage Windows 98 Pentium II machine.  This endeavor involved several challenges, including sourcing compatible hardware, transferring necessary files via FTP, and adapting modern code to compile on outdated systems. The team utilized Borland C++ 5.02, a 26-year-old IDE, to modify Andrej Karpathy's llama2.c, enabling it to function on the legacy hardware. The results were impressive: stories260K model: 39.31 tokens/second stories15M model: 1.03 tokens/second Llama 3.2 1B model: 0.0093 tokens/second This achievement underscores the potential for running advanced AI models on hardware once considered obsolete, aligning with EXO's mission to make frontier AI accessible beyond traditional data centers. Their work also highlights the importance of efficient model architectures, such as BitNet, which utilizes ternary weights to significantly reduce storage requirements and enhance energy efficiency.  This approach could enable large models to operate on minimal hardware, further democratizing AI technology. At Netglass.io, we're inspired by such innovative approaches that challenge conventional boundaries in technology. Stay tuned for more insights and discussions on the ever-evolving landscape of technology. Sources [https://blog.exolabs.net/day-4](https://blog.exolabs.net/day-4) [AI](https://netglass.io/blog/tag/ai) [LLM](https://netglass.io/blog/tag/llm) [Retro Computing](https://netglass.io/blog/tag/retro-computing) ## Similar posts ### Get notified about new cloud insights Stay ahead of the curve and stay informed on the latest advancements in cloud technologies, potential security risks to be aware of, and valuable optimizations that can enhance your cloud experience. Sign up now to receive exclusive updates and insights straight to your inbox. ### Subscribe [![Netglass Logo](https://netglass.io/hubfs/NetglassLogoTxtWhiteTrans4.svg)](https://netglass.io) Netglass.io LLC 332 S Michigan Ave Suite 121 # 5760 Chicago, IL 60604 United States ### Resources - [Blog](https://netglass.io/blog) - [Privacy policy](https://netglass.io/privacy) - [Contact us](https://netglass.io/contact) - [Schedule meeting](https://netglass.io/meeting) © 2024 Netglass.io LLC All rights reserved [Powered by Atlas - a B2B SaaS HubSpot theme](https://www.kalungi.com/atlas-hubspot-theme-for-b2b-saas-software) ```json { "@context" : "https://schema.org", "@type" : "BlogPosting", "author" : { "@type" : "Person", "name" : "Phil Emmerson", "url" : "https://netglass.io/blog/author/phil" }, "dateModified" : "2025-01-02T20:54:05.960Z", "datePublished" : "2025-01-02T15:03:34.000Z", "headline" : "Running Llama on Windows 98", "image" : [ "https://netglass.io/hubfs/AI-Generated%20Media/Images/The%20image%20features%20a%20vintage%20Windows%2098%20Pentium%20II%20computer%20setup%2c%20complete%20with%20a%20CRT%20monitor%20displaying%20lines%20of%20code.jpeg" ], "mainEntityOfPage" : { "@id" : "https://netglass.io/blog/llama-on-win98", "@type" : "WebPage" }, "publisher" : { "@type" : "Organization", "logo" : { "@type" : "ImageObject", "url" : "https://netglass.io/hubfs/netglass-logo-lt-blue.png" }, "name" : "Netglass.io LLC" } } ```