FAQ topic04 / 12

Processors and Boards FAQs

The processor and board set the performance, power draw and lifespan of an embedded system. These FAQs explain CPUs, clock speeds, caches, x86 and Arm, system-on-modules, COM Express and motherboard form factors.

Search this topic

45 answers Search all FAQs

Topics 12

COMPUTER HARDWARE · BOARDS

What Is a Motherboard?

A motherboard is the main circuit board in a computer. It carries the processor socket, memory slots, chipset, firmware and power circuitry, and it links storage, expansion cards and every external port so the parts work together as one system.

6 MIN

PROCESSORS · FUNDAMENTALS

What Is a CPU?

A CPU, or central processing unit, is the chip that runs a computer’s software. It fetches instructions from memory, decodes them and carries them out, billions of times a second. A modern CPU packs several cores, cache memory and often graphics and an NPU into one package.

7 MIN

ELECTRONICS · CHIPS

What Is a Semiconductor?

A semiconductor is a material, most often silicon, whose ability to conduct electricity sits between a conductor and an insulator and can be controlled. Adding impurities, called doping, lets engineers build diodes and transistors from it, and billions of transistors on one piece of silicon make the chips inside every computer.

7 MIN

PROCESSORS · CLOCK SPEED

What Is GHz?

GHz, short for gigahertz, means one billion cycles per second. On a processor it gives the clock speed: a 3 GHz CPU runs three billion clock cycles a second. Higher GHz helps, but cores, instructions per clock, cache and cooling decide real performance just as much.

6 MIN

DISPLAYS · PANEL INTERFACES

What Is LVDS?

LVDS (low-voltage differential signalling) is a way of sending high-speed digital data over pairs of wires using a small voltage difference of about 350 mV. It is defined in TIA/EIA-644. Its best-known use is the internal link between an embedded board and an LCD panel, although newer designs are moving to eDP.

6 MIN

FIRMWARE · RELIABILITY

What Is a Watchdog Timer?

A watchdog timer (WDT) is a hardware countdown that resets a computer or microcontroller if its software stops responding. Healthy software restarts the countdown at regular intervals. If the software hangs and the timer reaches zero, the watchdog triggers a reset, so an unattended system recovers by itself instead of waiting for someone to restart it.

6 MIN

GPUS AND AI · GPU SHARING

What Is NVIDIA MIG?

NVIDIA MIG (Multi-Instance GPU) splits one data centre GPU into as many as seven smaller GPUs, each with its own compute cores, cache and memory. Each instance runs its own workload in isolation with predictable performance, so one physical GPU can serve several users, models or inference services at once.

6 MIN

GPUS AND AI · FORM FACTORS

SXM vs PCIe GPUs: What Is the Difference?

SXM and PCIe are two ways of fitting an NVIDIA data centre GPU. An SXM GPU is a module mounted flat on a server baseboard, with higher power and all-to-all NVLink between GPUs. A PCIe GPU is a standard add-in card for ordinary server slots that uses less power and links to other cards over PCIe or an NVLink bridge.

6 MIN

SERVERS · GPU SYSTEM DESIGN

What Is NVIDIA MGX?

NVIDIA MGX is a modular reference architecture for building accelerated servers. NVIDIA defines standard building blocks for the chassis, GPUs, CPUs, networking, power and cooling, and server makers combine them into their own systems, from 1U and 2U GPU servers to liquid-cooled racks, with x86 or Arm processors.

6 MIN

GPUS AND AI · NVIDIA HOPPER

H100 vs H200: What Is the Difference?

The H100 and H200 are both NVIDIA Hopper data centre GPUs with the same compute cores. The difference is memory: the H200 has 141 GB of HBM3e at 4.8 TB/s, against 80 GB of HBM3 at 3.35 TB/s on the H100 SXM. That lets it hold larger AI models and run memory-bound inference faster.

6 MIN

GPUS AND AI · INTERCONNECTS

What Is NVLink?

NVLink is NVIDIA’s high-speed link between GPUs. It lets two or more GPUs exchange data and share memory directly, at many times the bandwidth of PCI Express, so they can work on one large AI model or simulation. It runs over a bridge between PCIe cards or through NVSwitch chips on multi-GPU server boards.

6 MIN

STORAGE · EMBEDDED FLASH

What Is eMMC Storage?

eMMC storage (embedded MultiMediaCard) is NAND flash memory and its controller packaged in one small chip that is soldered to a device’s circuit board. Defined by JEDEC, it gives phones, tablets, single board computers and embedded modules a low-cost, low-power boot and system drive, typically from 8 to 256 GB, at up to around 400 MB/s.

6 MIN

PROCESSORS · EMBEDDED DESIGN

What Is an Embedded System?

An embedded system is a combination of a processor, memory, input/output and software built into a larger device to do one dedicated job, such as controlling a washing machine, a car’s brakes or a factory machine. Unlike a general-purpose PC, it runs fixed software, often in real time, and is designed around the device it serves.

6 MIN

SYSTEMS · LIFECYCLE

How Long Do Computers Last?

Most consumer laptops last around three to five years and desktops around five to eight. Industrial and embedded computers typically stay in service for seven to ten years or more, because they use fanless cooling, industrial components and processors from long-life embedded roadmaps. Heat, workload, storage wear and software support decide the rest.

6 MIN

PROCESSORS · SILICON

FPGA vs ASIC: What Is the Difference?

An FPGA (field-programmable gate array) is a chip whose logic can be reconfigured after manufacture, so one part can serve many jobs and be updated in the field. An ASIC (application-specific integrated circuit) is designed and fabricated for one job. ASICs are faster, smaller and more efficient at volume, while FPGAs avoid the high upfront cost and long lead time.

6 MIN

GPUS AND AI · NVIDIA

What Is NVIDIA Blackwell?

NVIDIA Blackwell is the GPU architecture NVIDIA announced in March 2024 to follow Hopper and Ada Lovelace. It powers B200 and GB200 data centre accelerators, RTX PRO Blackwell workstation and embedded GPUs, GeForce RTX 50 cards and the Jetson Thor edge AI modules, and adds fifth-generation Tensor Cores with 4-bit (FP4) AI.

6 MIN

OPERATING SYSTEMS · EMBEDDED LINUX

What Is Embedded Linux?

Embedded Linux is the Linux kernel and a trimmed set of software built to run one dedicated device, such as an industrial controller, gateway, camera or edge AI computer, rather than a general-purpose PC. It is assembled for specific hardware, often an Arm processor, and usually boots straight into the device’s application.

6 MIN

EXPANSION · PCI EXPRESS

What Is PCIe?

PCIe (PCI Express) is the high-speed serial interface that connects graphics cards, NVMe SSDs, network cards and other devices to a computer’s processor. Data travels over lanes, and slots come in x1, x4, x8 and x16 widths. Each PCIe generation roughly doubles the speed per lane.

6 MIN

FIRMWARE · BOOT

What Is a BIOS?

A BIOS (basic input/output system) is the firmware stored on a chip on the motherboard that runs first when a computer powers on. It tests the hardware, sets up the processor, memory and devices, then hands over to the operating system. On current computers the BIOS is UEFI firmware, but the name has stuck.

6 MIN

INTERFACES · EXPANSION

What Is PCIe 6.0?

PCIe 6.0 is the sixth generation of PCI Express. It doubles PCIe 5.0’s rate to 64 GT/s per lane by switching to PAM4 signalling with fixed-size FLIT packets and forward error correction, giving about 121 GB/s in each direction on an x16 link. The specification was released in 2022 and the first hardware is reaching data-centre servers.

6 MIN

INTERFACES · EXPANSION

What Is PCIe 5.0?

PCIe 5.0 is the fifth generation of PCI Express, the link between a processor and its GPUs, NVMe SSDs and network cards. It runs at 32 gigatransfers per second per lane, about 3.9 GB/s in each direction, so an x16 slot carries roughly 63 GB/s each way. It is backwards compatible with earlier generations.

6 MIN

FIRMWARE · SECURITY

What Is Intel Platform Trust Technology (PTT)?

Intel Platform Trust Technology (PTT) is Intel’s firmware TPM. It provides TPM 2.0 functions from firmware running in the processor’s security engine instead of a separate chip, so Windows 11, BitLocker and measured boot work without a discrete TPM. AMD’s equivalent is fTPM.

6 MIN

FIRMWARE · REMOTE MANAGEMENT

What Is a Baseboard Management Controller (BMC)?

A baseboard management controller (BMC) is a small, separate processor on a server or industrial motherboard that runs independently of the main CPU and operating system. Through its own network connection it lets administrators power-cycle, monitor, update and reinstall a computer remotely, even when the system has crashed or is switched off.

6 MIN

PRODUCT DEVELOPMENT · MANUFACTURING

OEM vs ODM: What Is the Difference?

An OEM (original equipment manufacturer) sells a product under its own brand and owns its design, whether or not it builds the product itself. An ODM (original design manufacturer) designs and builds a product that other companies buy and sell under their own names. The difference comes down to who owns the design.

6 MIN

PRODUCT DEVELOPMENT · VALIDATION

What Are EVT, DVT and PVT?

EVT, DVT and PVT are the three build-and-test stages between a working prototype and mass production. Engineering validation testing (EVT) proves the design works, design validation testing (DVT) proves production-intent builds meet the specification and pass certification, and production validation testing (PVT) proves the factory can build it repeatably.

6 MIN

EXPANSION · CARD FORMATS

What Is Mini PCIe?

Mini PCIe (mPCIe) is a small plug-in card format for embedded computers and laptops, defined by PCI-SIG in 2003. The 52-pin socket carries one PCIe lane and USB 2.0, plus a SIM interface for cellular modems. It is used for Wi-Fi, 4G modems, GNSS, serial, CAN and video capture cards in full-size and half-size versions.

6 MIN

MEMORY · MODULES

DIMM vs SO-DIMM: What Is the Difference?

A DIMM is the full-size memory module used in desktops, servers and larger industrial boards. A SO-DIMM is the short version, about half the length, used in laptops, mini PCs and compact embedded boards. At the same DDR generation and speed they perform the same; they differ in size, pin count and slot, so they are not interchangeable.

5 MIN

ELECTRONICS · BOARDS

What Is a Printed Circuit Board?

A printed circuit board (PCB) is a rigid or flexible board that holds electronic components and connects them through copper tracks bonded to an insulating base, usually FR-4 glass-fibre epoxy. It replaces loose wiring with a compact, repeatable layout and is the base of every motherboard, single board computer and expansion card.

7 MIN

PROCESSORS · CORES AND THREADS

What Is Hyperthreading?

Hyperthreading is Intel’s name for simultaneous multithreading (SMT): each physical core runs two threads at once, so the operating system sees two logical processors per core. It fills idle time in the core and typically adds around 15 to 30% throughput on parallel work. It does not double performance.

6 MIN

COMPUTER HARDWARE · FORM FACTORS

What Is a Motherboard Form Factor?

A motherboard form factor is the standard that fixes a board’s size, mounting holes, rear I/O area and slot positions, so it fits matching cases and power supplies. The main motherboard sizes are ATX, Micro-ATX and Mini-ITX, with smaller embedded sizes such as 3.5-inch, Pico-ITX and PC/104.

9 MIN

GRAPHICS · GPU CARDS

What Is a Graphics Card?

A graphics card, also called a video card, is an expansion card that carries a graphics processing unit (GPU) with its own memory, power circuits, cooling and display outputs. It plugs into a PCIe slot and draws images for the screens, and it also speeds up video, AI inference and other parallel computing work.

6 MIN

PROCESSORS · MEMORY HIERARCHY

What Is a CPU Cache?

A CPU cache is a small block of very fast SRAM built into the processor. It keeps copies of the data and instructions the cores use most, so they wait less for main memory. Caches come in levels: L1 is smallest and fastest, L2 sits beside each core, and L3 is larger and shared.

6 MIN

PROCESSORS · ARCHITECTURE

32-Bit vs 64-Bit: What Is the Difference?

A 32-bit processor works with 32-bit registers and memory addresses, so it can address about 4 GB of memory. A 64-bit processor uses 64-bit registers and addresses, so it can reach far more memory and handle larger numbers in one step. A 64-bit operating system runs most 32-bit software, but not the reverse.

6 MIN

PROCESSORS · ARM

What Is the Qualcomm Hexagon NPU?

Qualcomm Hexagon is the processor block Qualcomm uses for AI inference in its Snapdragon and Dragonwing chips. It began as a digital signal processor (DSP) for audio, imaging and sensors, and has grown into a dedicated NPU. It works alongside the CPU and Adreno GPU, which together Qualcomm calls the Qualcomm AI Engine.

6 MIN

PROCESSORS · NPU

What Is Intel AI Boost?

Intel AI Boost is Intel’s name for the NPU, or neural processing unit, built into Intel Core Ultra processors. It runs AI inference such as image recognition, speech and background effects at low power, freeing the CPU and GPU. Software reaches it through Intel OpenVINO, Windows ML, DirectML and ONNX Runtime.

6 MIN

PROCESSORS · NPU

What Is AMD Ryzen AI?

AMD Ryzen AI is AMD’s name for the AI hardware and software in its Ryzen processors. At its centre is an NPU based on AMD’s XDNA architecture, built onto the processor beside the CPU cores and Radeon graphics. In our range, Ryzen 8000-series processors carry a 16 TOPS NPU and Ryzen AI 300-series processors a 50 TOPS NPU.

6 MIN

PROCESSORS · INTEL

Intel Core Ultra vs Core i7: What Is the Difference?

Core Ultra is the brand Intel introduced with Meteor Lake, replacing Core i5, i7 and i9 on its newest chips. Core Ultra 7 sits where Core i7 did, but adds a built-in NPU for AI and Intel Arc or Xe graphics. Core i7 ended with the 14th Gen, and newer non-Ultra chips are named Core 5 and Core 7.

6 MIN

PROCESSORS · INTEL

Intel N100 vs N150: What Is the Difference?

The Intel N150 is a refresh of the N100. Both have four efficient cores, four threads, 6 MB of cache and a 6 W base power. The N100 is an Alder Lake-N processor with a 3.4 GHz maximum turbo; the N150, from the Twin Lake refresh, reaches 3.6 GHz and runs its integrated graphics at a higher clock.

6 MIN

GPU · INTEL

What Is Intel Arc? Intel’s GPUs for Edge and Embedded Systems

Intel Arc is Intel’s graphics brand. It covers discrete Intel Arc GPUs, the A-series and newer B-series cards and modules, and the Arc graphics built into Intel Core Ultra processors. Arc GPUs handle displays, video encoding and decoding, and AI inference, and they are supported by Intel’s OpenVINO toolkit for running AI models.

6 MIN

PROCESSORS · GPU

GPU vs CPU: What Is the Difference?

A CPU has a small number of powerful cores that handle many different tasks one after another, such as running the operating system and applications. A GPU has thousands of simpler cores that work in parallel, which makes it far faster for graphics, video and AI inference, where the same calculation is repeated across large amounts of data.

7 MIN

EXPANSION · PCIE

PCIe 4.0 vs PCIe 5.0: What Is the Difference?

PCIe 5.0 doubles the data rate of PCIe 4.0, from 16 to 32 gigatransfers per second per lane. That gives about 4 GB/s per lane in each direction instead of 2 GB/s, or roughly 64 GB/s for an x16 slot. The two are compatible, so a card works in either slot at the slower generation’s speed.

6 MIN

MOTHERBOARDS · PLATFORM

What Is a Motherboard Chipset?

A motherboard chipset is the support chip that connects the processor to most of the board’s other features, such as USB, SATA, extra PCIe lanes, networking and audio. It decides how many ports and expansion slots a board can offer, which processors it accepts and which platform features, such as remote management, are available.

6 MIN

PROCESSORS · MANAGEMENT

What Is Intel vPro?

Intel vPro is a platform for business and edge processors that adds hardware-based remote management, through Intel Active Management Technology (AMT), plus security and stability features for fleets of computers. With AMT, an administrator can power on, restart, reach the BIOS or reinstall a computer over the network, even when its operating system has stopped responding.

6 MIN

EMBEDDED · STANDARDS

What Are the COM Express Module Types?

COM Express is a PICMG standard for computer-on-modules, and its pin-out types define which interfaces a module provides. Type 6 is the most common, with display outputs, USB and PCIe for general embedded systems. Type 7 is headless, with up to four 10GbE ports for edge server use. Type 10 is a smaller, lower-power pin-out for Mini modules.

6 MIN

EDGE AI · HARDWARE

iGPU vs dGPU: What Is the Difference?

An iGPU (integrated GPU) is built into the processor and shares the system memory, so it adds little power, heat or cost. A dGPU (discrete GPU) is a separate graphics processor with its own memory, offering far more performance but needing more power, cooling and space. For edge AI, iGPUs suit lighter inference and dGPUs suit heavy workloads.

7 MIN

Products

Shop this topic

Products for processors and boards

All edge computers

CategoryIndustrial MotherboardsLong-life ATX, Mini-ITX and smaller boards with Intel and AMD processors.View products

CategorySingle Board Computers3.5 inch, Pico-ITX and Arm SBCs for compact embedded designs.View products

CategoryCOM ExpressCOM Express modules that let you upgrade the processor later.View products

CategoryCarrier BoardsCarrier boards for COM Express and NVIDIA Jetson modules.View products