AI Dictionary › Fondamenti AI

Virtualization

Virtualizzazione

Virtualization is the technology that creates software versions of physical resources: one server becomes ten virtual machines, each with its own operating system, convinced it runs on dedicated hardware. It is the abstraction that made the cloud possible: renting compute capacity by the slice instead of buying whole machines.

Definition

How it works

The heart is the hypervisor, the software layer between hardware and virtual machines: it allocates CPU, memory, disk and network to each VM, keeps them isolated and intercepts their privileged operations. Type 1 hypervisors (VMware ESXi, KVM, Hyper-V) run on bare metal; type 2 (VirtualBox) on top of an operating system. Intel and AMD hardware extensions, introduced in the mid-2000s, made the overhead nearly zero.

Applications

In AI, virtualization sits under everything: the cloud GPU instances where models are trained and served are virtual machines with dedicated or partitioned accelerators (modern GPUs split into virtual instances with technologies like NVIDIA MIG). Beyond AI it consolidated corporate data centers, multiplying server utilization and slashing costs.

History & etymology

The story starts far back: IBM experimented with virtual machines in the 1960s with the CP-40 system and then CP/CMS on System/360 mainframes, letting multiple users share an enormously expensive computer. The word virtual, from the Latin virtus, denotes what exists in effect without existing in physical form. The modern rebirth came with VMware, founded in 1998, which brought virtualization to commodity x86 hardware and paved the way for cloud computing.

How it's used in Grace

Understanding virtualization helps you in Grace's scenarios on cloud and AI costs: knowing what sits underneath a GPU instance makes your requests to the AI concrete. The 1515 courses cover the technological foundations that make your dialogue with IT teams credible.

Related terms

• Container
• Cluster
• GPU and TPU
• Sandbox
• Compute

More in Fondamenti AI

• LLM (Large Language Model)
• Token
• Context Window
• Temperature
• AI Hallucination
• Fine-tuning
• RAG (Retrieval Augmented Generation)
• Embedding

Put it into practice

• Real scenario: apply this concept with an instant AI verdict.
• Daily gym: short drills to train dictionary concepts.
• Playoff game: challenge other players, elimination-style.

📱 Download the Android app (beta) iOS coming soon

Say what you mean. Get what you need.

Grace Certified, the AI coach that trains and certifies your prompt engineering, by Agora Intelligence.