{"id":1615,"date":"2026-08-14T13:24:31","date_gmt":"2026-08-14T12:24:31","guid":{"rendered":"https:\/\/www.senetic.co.uk\/blog\/?p=1615"},"modified":"2026-08-14T13:25:05","modified_gmt":"2026-08-14T12:25:05","slug":"which-computer-for-ai-types-specifications-and-applications","status":"publish","type":"post","link":"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/","title":{"rendered":"Which computer for AI? Types, specifications and applications"},"content":{"rendered":"\n<p class=\"wp-block-paragraph\">Which computer for AI? It depends on the use case and model size: for office work, a laptop or AI PC with an NPU is usually sufficient; running larger models locally requires a workstation or desktop AI supercomputer; and training models from scratch calls for a GPU server or the cloud. This guide is aimed at businesses, IT teams, SMEs and public institutions planning an AI deployment.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Here, you will find four classes of computers designed for AI workloads, their practical limits, key specifications such as memory and bandwidth, and a table showing the resources required for specific use cases. This is an important decision: the right computer enables language models to run on local resources, helps protect data privacy and makes flexible, efficient use of available compute power.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/ec.europa.eu\/eurostat\/statistics-explained\/index.php?title=Use_of_artificial_intelligence_in_enterprises\" target=\"_blank\" rel=\"noreferrer noopener nofollow\">According to a Eurostat report,<\/a> in 2025, 20% of businesses in the European Union used AI technologies. Two years earlier, this figure was 8%. Among large enterprises, adoption has already reached 55%, compared with 17% among small businesses. Adoption is accelerating, albeit unevenly, and more and more teams are now facing the question of hardware.<\/p>\n\n\n\n<div id=\"ez-toc-container\" class=\"ez-toc-v2_0_86 counter-hierarchy ez-toc-counter ez-toc-grey ez-toc-container-direction\">\n<div class=\"ez-toc-title-container\">\n<p class=\"ez-toc-title\" style=\"cursor:inherit\">Table of Contents<\/p>\n<span class=\"ez-toc-title-toggle\"><a href=\"#\" class=\"ez-toc-pull-right ez-toc-btn ez-toc-btn-xs ez-toc-btn-default ez-toc-toggle\" aria-label=\"Toggle Table of Content\"><span class=\"ez-toc-js-icon-con\"><span class=\"\"><span class=\"eztoc-hide\" style=\"display:none;\">Toggle<\/span><span class=\"ez-toc-icon-toggle-span\"><svg style=\"fill: #999;color:#999\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" class=\"list-377408\" width=\"20px\" height=\"20px\" viewBox=\"0 0 24 24\" fill=\"none\"><path d=\"M6 6H4v2h2V6zm14 0H8v2h12V6zM4 11h2v2H4v-2zm16 0H8v2h12v-2zM4 16h2v2H4v-2zm16 0H8v2h12v-2z\" fill=\"currentColor\"><\/path><\/svg><svg style=\"fill: #999;color:#999\" class=\"arrow-unsorted-368013\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" width=\"10px\" height=\"10px\" viewBox=\"0 0 24 24\" version=\"1.2\" baseProfile=\"tiny\"><path d=\"M18.2 9.3l-6.2-6.3-6.2 6.3c-.2.2-.3.4-.3.7s.1.5.3.7c.2.2.4.3.7.3h11c.3 0 .5-.1.7-.3.2-.2.3-.5.3-.7s-.1-.5-.3-.7zM5.8 14.7l6.2 6.3 6.2-6.3c.2-.2.3-.5.3-.7s-.1-.5-.3-.7c-.2-.2-.4-.3-.7-.3h-11c-.3 0-.5.1-.7.3-.2.2-.3.5-.3.7s.1.5.3.7z\"\/><\/svg><\/span><\/span><\/span><\/a><\/span><\/div>\n<nav><ul class='ez-toc-list ez-toc-list-level-1 ' ><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-1\" href=\"#\" data-href=\"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/#What_will_this_computer_be_used_for\" >What will this computer be used for?<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-2\" href=\"#\" data-href=\"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/#Types_of_computers_for_AI\" >Types of computers for AI<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-3\" href=\"#\" data-href=\"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/#Hardware_requirements_how_much_memory_you_need_and_why_it_is_the_most_important_parameter\" >Hardware requirements: how much memory you need and why it is the most important parameter<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-4\" href=\"#\" data-href=\"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/#Memory_determines_what_fits_Bandwidth_determines_how_quickly_the_model_responds\" >Memory determines what fits. Bandwidth determines how quickly the model responds<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-5\" href=\"#\" data-href=\"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/#Brands_and_product_ranges_at_a_glance\" >Brands and product ranges at a glance<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-6\" href=\"#\" data-href=\"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/#Frequently_asked_questions\" >Frequently asked questions<\/a><\/li><\/ul><\/nav><\/div>\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"What_will_this_computer_be_used_for\"><\/span><strong>What will this computer be used for?<\/strong><span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Before looking at specifications, define the use case. It determines the choice, not the technical specifications alone. <strong>The best AI computer is one whose configuration and budget match the specific artificial intelligence applications.<\/strong><\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>AI assistance in everyday work.<\/strong> Meeting transcription, text generation and editing, an assistant built into the operating system, and filters for video conferencing. Models run in the cloud or locally, but they are small and place little demand on the hardware.<\/li>\n\n\n\n<li><strong>Running ready-made models locally.<\/strong> A company downloads an open-source model and uses it in-house for document analysis, handling customer enquiries or image generation. This category also includes data analysis and simpler machine learning projects. Data does not leave the organisation. This is where the real hardware requirements begin.<\/li>\n\n\n\n<li><strong>Fine-tuning models with proprietary data.<\/strong> Fine-tuning and LoRA enable a model to learn a particular company&#8217;s terminology, procedures and context. They require considerably more memory than simply running a ready-made model.<\/li>\n\n\n\n<li><strong>Training a model from scratch.<\/strong> Building a proprietary model from the ground up. This is a task for data centre infrastructure, not a desktop computer.<\/li>\n<\/ul>\n\n\n\n<p class=\"wp-block-paragraph\">The first two scenarios can be handled by hardware many companies already have or can purchase without major investment. The third requires an informed decision on the device class. In practice, the fourth means the cloud or a server room. This breakdown also helps AI developers and data analysts choose, as they most often work in the two middle scenarios.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Types_of_computers_for_AI\"><\/span><strong>Types of computers for AI<\/strong><span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>AI PC with an NPU<\/strong><\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">A computer or laptop with a dedicated NPU designed for artificial intelligence tasks. It processes them efficiently without placing a load on the CPU or graphics card, and without sending data to the cloud. The entry threshold is 40 TOPS (trillion operations per second). This is the requirement for Copilot+ certification, now available in laptops with Intel Core Ultra, AMD Ryzen AI and Snapdragon X processors.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Suitable for:<\/strong> office-work assistance, transcription, translation, operating a system assistant and running small local models.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Its limit:<\/strong> it cannot handle models larger than a few billion parameters. An AI PC is not intended for fine-tuning or training models; that is not its role.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Example: Dell Pro 16<\/strong> with an AMD Ryzen AI chip<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Workstation with an NVIDIA RTX card<\/strong><\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">A traditional workstation with a professional NVIDIA graphics card. For AI computers, NVIDIA RTX cards are typically chosen because their Tensor cores accelerate artificial intelligence and machine learning model calculations. Equally important is the CUDA software ecosystem, which has become the standard for most tools and libraries. The GPU performs the workload here, and the key specification is the amount of VRAM on the card \u2013 from 16 GB in entry-level variants to 96 GB in RTX PRO Blackwell cards. The advantage is very high memory bandwidth, which translates into real-world performance.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Suitable for:<\/strong> image and video generation, working with language models of up to approximately 32 billion parameters, rendering and computation \u2013 wherever response time matters. For language models, the practical minimum is 12\u201316 GB VRAM.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Its limit:<\/strong> models of 70B and larger. They will fit only on the most expensive cards with 96 GB VRAM, while models above 120B will not. Scaling up then means a multi-GPU configuration with appropriate power, cooling and cost.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Example: Lenovo ThinkStation P3 Tower Gen 2<\/strong> \u2013 a workstation with a professional NVIDIA RTX card and 16 GB of VRAM.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Desktop AI supercomputer<\/strong><\/h3>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>The newest class of devices, built around the NVIDIA GB10 Grace Blackwell superchip, brings work with large models to the desktop. <\/strong>It combines the CPU and GPU with a shared pool of 128 GB unified memory, fully available to the model without a rigid split between RAM and VRAM. The device is the size of a book, consumes as much power as a high-performance laptop and runs quietly, while supporting the full NVIDIA ecosystem: CUDA, NGC containers, PyTorch and Ollama. The greatest advantage of this class is complete control over data when running AI models locally.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Suitable for:<\/strong> running large models locally (up to approximately 200 billion parameters), prototyping, fine-tuning models with LoRA and QLoRA, and working with data that cannot leave the company. It is also well suited to testing proprietary AI applications and running short-iteration experiments, where rapidly validating successive ideas matters.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Its limit:<\/strong> when training models from scratch and serving many users simultaneously in production. This hardware is for a team working on a model, not an application server for the entire organisation.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Example: <\/strong><a href=\"https:\/\/www.senetic.co.uk\/product\/90MS0371-M00030\" target=\"_blank\" rel=\"noreferrer noopener\"><strong>Asus Ascent GX10<\/strong><\/a> &#8211; a compact desktop AI supercomputer powered by the GB10 superchip, aimed at developers and AI and data engineers.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>GPU server and cloud<\/strong><\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Server infrastructure with H100, H200 or Blackwell-class accelerators \u2014 owned in-house or rented in the cloud. It offers substantially greater compute power and bandwidth, but also has requirements: a rack, power and cooling measured in kilowatts.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Suitable for:<\/strong> training models from scratch, serving multiple users, deployments that require many operations to run simultaneously, and real-time workloads.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Where the limits lie:<\/strong> costs and data. The cloud means information leaves the company, which remains one of the main barriers to AI adoption.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Before moving to a data centre, however, it is worth considering an intermediate step. Desktop AI supercomputers can be paired to provide 256 GB of shared memory \u2014 enough to work with models containing around 405 billion parameters without leaving the office or investing in infrastructure.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Example: HP ZGX Nano G1n<\/strong> &#8211; two such units connected via a ConnectX-7 interface provide 256 GB of shared memory.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Hardware_requirements_how_much_memory_you_need_and_why_it_is_the_most_important_parameter\"><\/span><strong>Hardware requirements: how much memory you need and why it is the most important parameter<\/strong><span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Memory determines whether a model can run at all. <\/strong>In a workstation, this is the VRAM on the graphics card; in desktop AI supercomputers, it is unified memory shared by the CPU and GPU. From the user&#8217;s perspective, one thing matters: how much memory is actually available to the model. System memory and VRAM are not interchangeable, as they serve different roles and must be selected separately.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Labels such as 7B and 70B indicate the number of model parameters: 7 and 70 billion, respectively. Parameters are values the model developed during training and must load into memory in order to operate. The more parameters there are, the better the response quality, but the greater the hardware requirements.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The requirement depends on the model&#8217;s parameter count and quantisation, i.e. the precision used to store it. The rule is simple: with 4-bit quantisation (Q4), a model needs roughly 0.5\u20130.6 GB for every billion parameters; with 8-bit quantisation (Q8), around 1 GB; and at full FP16 precision, around 2 GB. You also need to allow a 10\u201320% margin.<\/p>\n\n\n\n<figure class=\"wp-block-table\"><table class=\"has-fixed-layout\"><tbody><tr><td><strong>Model<\/strong><\/td><td><strong>Q4 (4-bit)<\/strong><\/td><td><strong>Q8 (8-bit)<\/strong><\/td><td><strong>FP16<\/strong><\/td><\/tr><tr><td>7B \/ 8B<\/td><td>5\u20138 GB<\/td><td>8\u201310 GB<\/td><td>14\u201316 GB<\/td><\/tr><tr><td>13B<\/td><td>8\u201310 GB<\/td><td>14 GB<\/td><td>26 GB<\/td><\/tr><tr><td>30\u201334B<\/td><td>approx. 20 GB<\/td><td>approx. 34 GB<\/td><td>approx. 64 GB<\/td><\/tr><tr><td>70B<\/td><td>40\u201348 GB<\/td><td>approx. 70 GB<\/td><td>approx. 140 GB<\/td><\/tr><tr><td>120B (MoE)<\/td><td>65\u201375 GB<\/td><td>\u2014<\/td><td>\u2014<\/td><\/tr><tr><td>200B (FP4)<\/td><td>fits within 128 GB<\/td><td>\u2014<\/td><td>\u2014<\/td><\/tr><tr><td>405B<\/td><td>256 GB (two units)<\/td><td>\u2014<\/td><td>approx. 810 GB<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\"><em>The values in the Q4 column already include headroom for the KV cache and overhead. The Q8 and FP16 columns show the size of the model weights alone, to which an additional 10\u201320% must be added. The 405B model fits on two connected units only with 4-bit (Q4) quantisation.<\/em><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Image generation has lower requirements: Stable Diffusion 1.5 runs from 4\u20136 GB, SDXL is comfortable at 12\u201316 GB, while newer FLUX-class models require 24\u201348 GB. Training a custom LoRA for SDXL fits within 16\u201324 GB.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">In data science environments, 32 GB of RAM remains the standard, although working with large local models may require more. LoRA or QLoRA fine-tuning is feasible even for a 70B model on a single unit with 128 GB of unified memory. Full training of large models remains the domain of data centre hardware.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Memory_determines_what_fits_Bandwidth_determines_how_quickly_the_model_responds\"><\/span><strong>Memory determines what fits. Bandwidth determines how quickly the model responds<\/strong><span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">The second parameter, rarely discussed but crucial to everyday usability, is memory bandwidth. Memory capacity determines whether a model can load at all; bandwidth determines how quickly it generates subsequent words in its response.<\/p>\n\n\n\n<figure class=\"wp-block-table\"><table class=\"has-fixed-layout\"><tbody><tr><td><strong>Device<\/strong><\/td><td><strong>Memory bandwidth<\/strong><\/td><\/tr><tr><td>Desktop AI supercomputer (GB10)<\/td><td>273 GB\/s<\/td><\/tr><tr><td>Apple Mac Studio M3 Ultra<\/td><td>819 GB\/s<\/td><\/tr><tr><td>NVIDIA RTX 4090<\/td><td>approx. 1008 GB\/s<\/td><\/tr><tr><td>NVIDIA RTX 5090<\/td><td>approx. 1792 GB\/s<\/td><\/tr><tr><td>NVIDIA H100 (HBM3)<\/td><td>approx. 3350 GB\/s<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">At first glance, the conclusion seems obvious: if the H100 has twelve times the bandwidth, why would anyone need a device with 273 GB\/s? However, this comparison is misleading because these systems do not compete with each other.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>H100 is an accelerator for a rack-mounted server<\/strong>, which costs tens of thousands of dollars for the chip alone and requires data centre infrastructure. A desktop AI supercomputer sits on a desk and consumes as much power as a high-performance laptop. This class is popular not because of its speed, but because 128 GB of memory in a small-computer form factor was previously unattainable at this price. Previously, a company had three options: an RTX card with 24\u201332 GB, which is too little for large models; a server with H100, which is too expensive and requires a data centre; or the cloud, meaning data leaves the organisation.<strong> This new device category fills the gap in the middle.<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><em>It is a bit like comparing a delivery van with a sports car. The sports car is faster, but that is not the point. What matters is how much it can hold and how much it costs.<\/em><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">In practice, it works like this: the 70B model generates a response slowly in a single stream, noticeably more slowly than on a server. However, it does so locally, without sending data to the cloud or investing in server-room infrastructure. With multiple concurrent queries, for example in pipelines or agents, total throughput increases. This is more than sufficient for prototyping, fine-tuning and team-based work on a model.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The remaining components are of secondary, though not negligible, importance. A fast NVMe drive reduces model loading times and affects overall performance, so an AI computer specification should include the GPU, processor, RAM and NVMe SSD. In preconfigured systems, the remaining components are selected for the specific platform, so there is no need to assess them separately. The only choice you genuinely need to make is drive capacity, which should be matched to the size of the models and datasets you work with.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Brands_and_product_ranges_at_a_glance\"><\/span><strong>Brands and product ranges at a glance<\/strong><span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>All AI desktop supercomputers currently available are based on the same NVIDIA GB10 superchip and have the same 128 GB of unified memory.<\/strong> Manufacturers do not differentiate them by performance. <strong>Individual configurations differ in storage capacity, chassis and cooling, operating system, and warranty and support coverage.<\/strong> This is important when making a purchase: comparing these systems in terms of computing power makes no sense; what matters is how well they fit a given organisation&#8217;s processes and requirements.<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><a href=\"https:\/\/www.senetic.co.uk\/category\/computers-for-ai-30635\/?f_page=1&amp;f_size=24&amp;f_order=bestsellers&amp;cat=custom_ai_computers_78420152401&amp;f_s_manufacturer=nvidia\" target=\"_blank\" rel=\"noreferrer noopener\"><strong>NVIDIA DGX Spark<\/strong><\/a> &#8211; the reference version and benchmark for the entire device class. NVIDIA bases its documentation and technical materials on this system, so performance descriptions and tests published online most often refer to this particular variant. It comes with a high-capacity, high-speed drive and a distinctive chassis that sets it apart from partner designs.<\/li>\n\n\n\n<li><strong>Lenovo ThinkStation PGX<\/strong> &#8211; a ThinkStation range backed by extensive service and technical support. It is the most widely represented brand in this category, with numerous variants available that differ in storage capacity, operating system and warranty package. A natural choice for organisations that already have established procurement and service processes with Lenovo.<\/li>\n\n\n\n<li><strong>HP ZGX Nano G1n<\/strong> &#8211; also available in a version without a wireless module, designed for environments with stringent security requirements. This is a rare option in this class and a genuine advantage wherever security policies rule out wireless connectivity, for example in public administration, the financial sector or air-gapped networks. The system is supplied with a Linux-based operating system ready to work with models.<\/li>\n\n\n\n<li><strong>Dell Pro Max<\/strong> &#8211; a natural choice for organisations purchasing hardware under corporate agreements with Dell, simplifying formalities for larger deployments. The standard configuration includes a high-capacity drive, with no need to purchase additional storage.<\/li>\n\n\n\n<li><a href=\"https:\/\/www.senetic.co.uk\/category\/computers-for-ai-30635\/?f_page=1&amp;f_size=24&amp;f_order=bestsellers&amp;cat=custom_ai_computers_78420152401&amp;f_s_manufacturer=asus\" target=\"_blank\" rel=\"noreferrer noopener\"><strong>ASUS Ascent GX10<\/strong><\/a> &#8211; a compact design. Available variants with different drive capacities allow organisations to start with a smaller configuration and lower the entry threshold if they are only beginning to test local model workloads. The design allows units to be stacked, making subsequent expansion easier.<\/li>\n\n\n\n<li><strong>Gigabyte AI TOP ATOM<\/strong> and <strong>MSI EdgeXpert<\/strong> &#8211; variants focused on high-capacity, high-speed storage. This matters when working with large datasets and multiple models simultaneously, where model files alone can occupy hundreds of gigabytes. They are also well suited where the system needs to operate outside the server room, close to where data is generated.<\/li>\n\n\n\n<li><strong>Acer Veriton GN100<\/strong> &#8211; a design that dissipates heat effectively. Compared with other systems in this class, it heats up the least, which translates into stable operation during long-running tasks such as model fine-tuning, when the hardware remains under load for many hours without interruption.<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Frequently_asked_questions\"><\/span><strong>Frequently asked questions<\/strong><span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>What is the difference between a standard computer and an AI PC?<\/strong><\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">An AI PC has a dedicated NPU that processes artificial intelligence tasks locally and efficiently. A standard computer can also handle AI, but it does so using the processor or graphics card \u2013 more slowly and with higher energy consumption.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>How much RAM is needed for AI workloads?<\/strong><\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">For office applications with AI assistance, 16 GB is sufficient, while 32 GB is more comfortable. In data science environments, 32 GB of RAM is standard, and working with local models often requires more. However, not only system memory matters: the memory available to the GPU \u2013 VRAM or unified memory \u2013 is also critical.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>How much VRAM is required to run the 7B, 13B or 70B model?<\/strong><\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">With 4-bit quantisation: approximately 5\u20138 GB for the 7B model, 8\u201310 GB for 13B, and 40\u201348 GB for 70B. In practice, for language models, the minimum starts at 12\u201316 GB VRAM, depending on quantisation.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>For AI, which matters more: the processor or the graphics card?<\/strong><\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">The graphics card or a dedicated accelerator. The processor handles data preparation, but the GPU performs the actual computations. Tensor Core-based hardware accelerates machine-learning tasks more effectively than the processor alone.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>What is unified memory and how does it differ from VRAM?<\/strong><\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">VRAM is memory permanently assigned to a graphics card. Unified memory is a shared pool used by both the processor and graphics subsystem. This means that a system with 128 GB of unified memory can run a model that would not fit on any single graphics card. In return, it has lower bandwidth, so responses are generated more slowly.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>What is an NPU and how many TOPS do I need?<\/strong><\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">An NPU is a chip specialised in artificial intelligence calculations. TOPS indicates its performance. For office applications with AI assistance, 40 TOPS is sufficient; this is the threshold for Copilot+ certification. For larger models, TOPS is no longer a meaningful metric; memory capacity and bandwidth become the key factors.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Does AI work on Windows, or is Linux required?<\/strong><\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">AI runs on Windows. However, professional tools for working with large models are developed primarily for Linux, which is why desktop AI supercomputers are usually supplied with an Ubuntu-based operating system.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>How large can models be with 128 GB of unified memory?<\/strong><\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Models of around 200 billion parameters at 4-bit quantisation. Models in the region of 405 billion parameters require two units to be connected.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Can two units be connected in a cluster?<\/strong><\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Yes. Devices based on GB10 feature a built-in NVIDIA ConnectX-7 network interface with 200 GbE ports, allowing two units to be connected to provide 256 GB of shared memory.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>When should you choose an AI PC, a workstation or a server?<\/strong><\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">AI PC \u2013 when AI is intended to support everyday office work. RTX workstation \u2013 when you work with images, video or models of up to 32 billion parameters and speed is a priority. Desktop AI supercomputer \u2013 when you need large local models and fine-tuning. Server or cloud \u2013 when you train models.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><strong>Does data leave the company when using a local model?<\/strong><\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">No. A model running on your own hardware processes data exclusively locally. This is the main reason why companies working with sensitive data choose on-premises solutions rather than the cloud.<\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<p class=\"wp-block-paragraph\"><em>Data on AI use in enterprises: Eurostat, \u201cUse of artificial intelligence in enterprises\u201d, data for 2025. Specifications and parameters current as of August 2026.<\/em><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Which computer for AI? It depends on the use case and model size: for office work, a laptop or AI PC with an NPU is usually sufficient; running larger models locally requires a workstation or desktop AI supercomputer; and training models from scratch calls for a GPU server or the cloud. This guide is aimed<\/p>\n","protected":false},"author":7,"featured_media":1614,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_yoast_wpseo_meta-robots-noindex":"","_yoast_wpseo_meta-robots-nofollow":"","_yoast_wpseo_meta-robots-adv":"","_yoast_wpseo_bctitle":"","_yoast_wpseo_canonical":"","_yoast_wpseo_opengraph-title":"","_yoast_wpseo_opengraph-description":"","_yoast_wpseo_opengraph-image":"","_yoast_wpseo_twitter-title":"","_yoast_wpseo_twitter-description":"","_yoast_wpseo_twitter-image":"","footnotes":""},"categories":[7,163],"tags":[],"class_list":["post-1615","post","type-post","status-publish","format-standard","has-post-thumbnail","category-computer-hardware","category-hardware-technologies"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v28.2 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Which computer for AI? Specs and Brands Compared | Senetic<\/title>\n<meta name=\"description\" content=\"AI workstation specs vary widely. See which ones make a real difference for business use, and which brand series are worth considering.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/\" \/>\n<meta property=\"og:locale\" content=\"pl_PL\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Which computer for AI? Specs and Brands Compared | Senetic\" \/>\n<meta property=\"og:description\" content=\"AI workstation specs vary widely. See which ones make a real difference for business use, and which brand series are worth considering.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/\" \/>\n<meta property=\"og:site_name\" content=\"Senetic Blog IT - Microsoft, Office 365, Azure, Ubiquiti, MikroTik\" \/>\n<meta property=\"article:published_time\" content=\"2026-08-14T12:24:31+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2026-08-14T12:25:05+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/www.senetic.co.uk\/blog\/app\/uploads\/2026\/08\/fb217fa6-9bc6-4462-9985-d01ac7281e93.jpg\" \/>\n\t<meta property=\"og:image:width\" content=\"1319\" \/>\n\t<meta property=\"og:image:height\" content=\"727\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/jpeg\" \/>\n<meta name=\"author\" content=\"Adam Kubarski\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Adam Kubarski\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"18 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/1615,which-computer-for-ai-types-specifications-and-applications\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/1615,which-computer-for-ai-types-specifications-and-applications\\\/\"},\"author\":{\"name\":\"Adam Kubarski\",\"@id\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/#\\\/schema\\\/person\\\/951e696088c815246a24768e3be85b5e\"},\"headline\":\"Which computer for AI? Types, specifications and applications\",\"datePublished\":\"2026-08-14T12:24:31+00:00\",\"dateModified\":\"2026-08-14T12:25:05+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/1615,which-computer-for-ai-types-specifications-and-applications\\\/\"},\"wordCount\":2859,\"commentCount\":0,\"image\":{\"@id\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/1615,which-computer-for-ai-types-specifications-and-applications\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/app\\\/uploads\\\/2026\\\/08\\\/fb217fa6-9bc6-4462-9985-d01ac7281e93.jpg\",\"articleSection\":[\"Computer Hardware\",\"Hardware &amp; Technologies\"],\"inLanguage\":\"pl-PL\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/1615,which-computer-for-ai-types-specifications-and-applications\\\/#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/1615,which-computer-for-ai-types-specifications-and-applications\\\/\",\"url\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/1615,which-computer-for-ai-types-specifications-and-applications\\\/\",\"name\":\"Which computer for AI? Specs and Brands Compared | Senetic\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/1615,which-computer-for-ai-types-specifications-and-applications\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/1615,which-computer-for-ai-types-specifications-and-applications\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/app\\\/uploads\\\/2026\\\/08\\\/fb217fa6-9bc6-4462-9985-d01ac7281e93.jpg\",\"datePublished\":\"2026-08-14T12:24:31+00:00\",\"dateModified\":\"2026-08-14T12:25:05+00:00\",\"author\":{\"@id\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/#\\\/schema\\\/person\\\/951e696088c815246a24768e3be85b5e\"},\"description\":\"AI workstation specs vary widely. See which ones make a real difference for business use, and which brand series are worth considering.\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/1615,which-computer-for-ai-types-specifications-and-applications\\\/#breadcrumb\"},\"inLanguage\":\"pl-PL\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/1615,which-computer-for-ai-types-specifications-and-applications\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"pl-PL\",\"@id\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/1615,which-computer-for-ai-types-specifications-and-applications\\\/#primaryimage\",\"url\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/app\\\/uploads\\\/2026\\\/08\\\/fb217fa6-9bc6-4462-9985-d01ac7281e93.jpg\",\"contentUrl\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/app\\\/uploads\\\/2026\\\/08\\\/fb217fa6-9bc6-4462-9985-d01ac7281e93.jpg\",\"width\":1319,\"height\":727,\"caption\":\"AI computer\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/1615,which-computer-for-ai-types-specifications-and-applications\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Which computer for AI? Types, specifications and applications\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/#website\",\"url\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/\",\"name\":\"Senetic Blog IT - Microsoft, Office 365, Azure, Ubiquiti, MikroTik\",\"description\":\"\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"pl-PL\"},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/#\\\/schema\\\/person\\\/951e696088c815246a24768e3be85b5e\",\"name\":\"Adam Kubarski\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"pl-PL\",\"@id\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/app\\\/uploads\\\/2026\\\/03\\\/cropped-Adam-Kubarski-96x96.jpg\",\"url\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/app\\\/uploads\\\/2026\\\/03\\\/cropped-Adam-Kubarski-96x96.jpg\",\"contentUrl\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/app\\\/uploads\\\/2026\\\/03\\\/cropped-Adam-Kubarski-96x96.jpg\",\"caption\":\"Adam Kubarski\"},\"description\":\"Head of E-commerce With over 15 years of experience in e-commerce across various industries, focusing on long-term growth, effective management, and business process optimisation. Previously, he also worked on his own e-commerce projects, blending entrepreneurial experience with a managerial approach. Outside of work, he regularly plays football and reads self-development books and biographies of leaders.\",\"sameAs\":[\"https:\\\/\\\/www.linkedin.com\\\/in\\\/adamkubarski\\\/\"],\"url\":\"https:\\\/\\\/www.senetic.co.uk\\\/blog\\\/author\\\/adam-kubarski\\\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Which computer for AI? Specs and Brands Compared | Senetic","description":"AI workstation specs vary widely. See which ones make a real difference for business use, and which brand series are worth considering.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/","og_locale":"pl_PL","og_type":"article","og_title":"Which computer for AI? Specs and Brands Compared | Senetic","og_description":"AI workstation specs vary widely. See which ones make a real difference for business use, and which brand series are worth considering.","og_url":"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/","og_site_name":"Senetic Blog IT - Microsoft, Office 365, Azure, Ubiquiti, MikroTik","article_published_time":"2026-08-14T12:24:31+00:00","article_modified_time":"2026-08-14T12:25:05+00:00","og_image":[{"width":1319,"height":727,"url":"https:\/\/www.senetic.co.uk\/blog\/app\/uploads\/2026\/08\/fb217fa6-9bc6-4462-9985-d01ac7281e93.jpg","type":"image\/jpeg"}],"author":"Adam Kubarski","twitter_card":"summary_large_image","twitter_misc":{"Written by":"Adam Kubarski","Est. reading time":"18 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/#article","isPartOf":{"@id":"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/"},"author":{"name":"Adam Kubarski","@id":"https:\/\/www.senetic.co.uk\/blog\/#\/schema\/person\/951e696088c815246a24768e3be85b5e"},"headline":"Which computer for AI? Types, specifications and applications","datePublished":"2026-08-14T12:24:31+00:00","dateModified":"2026-08-14T12:25:05+00:00","mainEntityOfPage":{"@id":"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/"},"wordCount":2859,"commentCount":0,"image":{"@id":"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/#primaryimage"},"thumbnailUrl":"https:\/\/www.senetic.co.uk\/blog\/app\/uploads\/2026\/08\/fb217fa6-9bc6-4462-9985-d01ac7281e93.jpg","articleSection":["Computer Hardware","Hardware &amp; Technologies"],"inLanguage":"pl-PL","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/#respond"]}]},{"@type":"WebPage","@id":"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/","url":"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/","name":"Which computer for AI? Specs and Brands Compared | Senetic","isPartOf":{"@id":"https:\/\/www.senetic.co.uk\/blog\/#website"},"primaryImageOfPage":{"@id":"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/#primaryimage"},"image":{"@id":"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/#primaryimage"},"thumbnailUrl":"https:\/\/www.senetic.co.uk\/blog\/app\/uploads\/2026\/08\/fb217fa6-9bc6-4462-9985-d01ac7281e93.jpg","datePublished":"2026-08-14T12:24:31+00:00","dateModified":"2026-08-14T12:25:05+00:00","author":{"@id":"https:\/\/www.senetic.co.uk\/blog\/#\/schema\/person\/951e696088c815246a24768e3be85b5e"},"description":"AI workstation specs vary widely. See which ones make a real difference for business use, and which brand series are worth considering.","breadcrumb":{"@id":"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/#breadcrumb"},"inLanguage":"pl-PL","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/"]}]},{"@type":"ImageObject","inLanguage":"pl-PL","@id":"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/#primaryimage","url":"https:\/\/www.senetic.co.uk\/blog\/app\/uploads\/2026\/08\/fb217fa6-9bc6-4462-9985-d01ac7281e93.jpg","contentUrl":"https:\/\/www.senetic.co.uk\/blog\/app\/uploads\/2026\/08\/fb217fa6-9bc6-4462-9985-d01ac7281e93.jpg","width":1319,"height":727,"caption":"AI computer"},{"@type":"BreadcrumbList","@id":"https:\/\/www.senetic.co.uk\/blog\/1615,which-computer-for-ai-types-specifications-and-applications\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/www.senetic.co.uk\/blog\/"},{"@type":"ListItem","position":2,"name":"Which computer for AI? Types, specifications and applications"}]},{"@type":"WebSite","@id":"https:\/\/www.senetic.co.uk\/blog\/#website","url":"https:\/\/www.senetic.co.uk\/blog\/","name":"Senetic Blog IT - Microsoft, Office 365, Azure, Ubiquiti, MikroTik","description":"","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.senetic.co.uk\/blog\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"pl-PL"},{"@type":"Person","@id":"https:\/\/www.senetic.co.uk\/blog\/#\/schema\/person\/951e696088c815246a24768e3be85b5e","name":"Adam Kubarski","image":{"@type":"ImageObject","inLanguage":"pl-PL","@id":"https:\/\/www.senetic.co.uk\/blog\/app\/uploads\/2026\/03\/cropped-Adam-Kubarski-96x96.jpg","url":"https:\/\/www.senetic.co.uk\/blog\/app\/uploads\/2026\/03\/cropped-Adam-Kubarski-96x96.jpg","contentUrl":"https:\/\/www.senetic.co.uk\/blog\/app\/uploads\/2026\/03\/cropped-Adam-Kubarski-96x96.jpg","caption":"Adam Kubarski"},"description":"Head of E-commerce With over 15 years of experience in e-commerce across various industries, focusing on long-term growth, effective management, and business process optimisation. Previously, he also worked on his own e-commerce projects, blending entrepreneurial experience with a managerial approach. Outside of work, he regularly plays football and reads self-development books and biographies of leaders.","sameAs":["https:\/\/www.linkedin.com\/in\/adamkubarski\/"],"url":"https:\/\/www.senetic.co.uk\/blog\/author\/adam-kubarski\/"}]}},"_links":{"self":[{"href":"https:\/\/www.senetic.co.uk\/blog\/wp-json\/wp\/v2\/posts\/1615","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.senetic.co.uk\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.senetic.co.uk\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.senetic.co.uk\/blog\/wp-json\/wp\/v2\/users\/7"}],"replies":[{"embeddable":true,"href":"https:\/\/www.senetic.co.uk\/blog\/wp-json\/wp\/v2\/comments?post=1615"}],"version-history":[{"count":1,"href":"https:\/\/www.senetic.co.uk\/blog\/wp-json\/wp\/v2\/posts\/1615\/revisions"}],"predecessor-version":[{"id":1618,"href":"https:\/\/www.senetic.co.uk\/blog\/wp-json\/wp\/v2\/posts\/1615\/revisions\/1618"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.senetic.co.uk\/blog\/wp-json\/wp\/v2\/media\/1614"}],"wp:attachment":[{"href":"https:\/\/www.senetic.co.uk\/blog\/wp-json\/wp\/v2\/media?parent=1615"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.senetic.co.uk\/blog\/wp-json\/wp\/v2\/categories?post=1615"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.senetic.co.uk\/blog\/wp-json\/wp\/v2\/tags?post=1615"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}