Level 1 - Absolute Beginner
Nvidia makes computer chips that power artificial intelligence, or AI. Nvidia is now telling companies that its AI computers will cost more money.
The price will go up by more than 15 percent. This means the computers will cost a lot more than before.
The reason for the higher price is that memory parts inside the computers cost more money now.
Big companies like Microsoft and Google buy these computers. They will have to pay the new, higher price.
- chip
- a small electronic part inside a computer that helps it work
- computer
- a machine that can store and use information
- artificial intelligence
- computer systems that can learn and act a bit like humans
- price
- the amount of money something costs
- percent
- a way to show a part of a hundred, like 15 percent
- memory
- the part of a computer that stores information
- cost
- the amount of money needed to buy or make something
- company
- a business that makes or sells things
Level 2 - Elementary
Nvidia, the world's biggest maker of AI computer chips, has told its customers that prices for servers built with its newest chips will rise by more than 15 percent.
The price increase will affect powerful new systems known as Vera Rubin and Grace Blackwell, which are used to train and run artificial intelligence programs.
The main reason for the higher prices is the rising cost of memory chips. Memory chips are an important part inside every AI server, and they have become much more expensive lately.
The new prices will apply to systems that ship early next year. Nvidia's announcement comes just before the company reports its newest earnings, so investors are watching closely to see if higher costs will hurt profits.
- server
- a powerful computer that provides data or services to other computers
- system
- a set of connected parts, like hardware and software, working together
- train
- to teach a computer program using large amounts of data
- program
- a set of instructions that tells a computer what to do
- memory chip
- a small part that stores information inside a computer
- apply
- to become active or used in a certain situation
- earnings
- the money a company makes over a period of time
- investor
- a person or group that puts money into a company hoping to make a profit
Level 3 - Intermediate
Nvidia has informed customers that it will raise prices on servers built around its most advanced AI chips by more than 15 percent in many configurations, according to people familiar with the matter. The increases will apply to systems based on the company's upcoming Vera Rubin platform as well as its current Grace Blackwell lineup.
The driving force behind the hikes is not the chips themselves but the memory that surrounds them. Prices for the high-bandwidth memory and other components packed into AI servers have climbed sharply this year, squeezing the contract manufacturers who assemble these systems for data center operators.
Those manufacturers, which build servers for customers including Microsoft, Google, and Oracle, have already begun passing the new pricing along. The exact size of the increase will vary depending on the chip generation and how much memory a particular configuration includes.
The timing is notable: Nvidia is scheduled to report its latest quarterly results this week, and investors will be looking closely at whether rising input costs are beginning to erode the company's famously wide profit margins, even as demand for AI computing infrastructure keeps growing.
- configuration
- the specific way a system's parts are arranged or set up
- platform
- a base system of hardware or software that other products are built on
- lineup
- a set or series of related products offered by a company
- high-bandwidth
- able to move very large amounts of data quickly
- component
- one part of a larger machine or system
- contract manufacturer
- a company hired to build products on behalf of another company
- data center
- a large facility that houses computer servers and equipment
- margin
- the difference between what something costs to make and what it sells for
Level 4 - Advanced
Nvidia has quietly notified its contract manufacturing partners that list prices for servers built around its flagship AI accelerators will climb by more than 15 percent in many configurations, a shift that stands to ripple through the balance sheets of every major cloud provider racing to expand AI infrastructure. The increases target both the forthcoming Vera Rubin platform and the current Grace Blackwell lineup.
Unlike previous price actions tied to Nvidia's own silicon, this round is being driven almost entirely by upstream cost pressure: a tightening global supply of high-bandwidth memory has pushed component costs sharply higher across the industry, squeezing the margins of the contract manufacturers who integrate memory, networking, and compute into finished server racks.
Those manufacturers, which supply hyperscale operators including Microsoft, Alphabet's Google, and Oracle, have begun relaying the increases downstream, with the precise markup varying by chip generation and the density of memory packed into each configuration. For an industry that has treated ever-cheaper compute as a given, the reversal underscores how a single memory bottleneck can cascade through an entire supply chain.
The disclosure lands just ahead of Nvidia's latest quarterly earnings report, sharpening investor scrutiny of whether the company's historically wide gross margins can withstand rising input costs even as demand for AI infrastructure shows little sign of slowing. Some analysts argue the increases simply reflect a seller's market; others warn they could accelerate customers' efforts to diversify away from a single dominant supplier.
- accelerator
- specialized hardware built to speed up a specific kind of computing task, like AI
- balance sheet
- a financial statement showing what a company owns and owes
- upstream
- earlier in a supply chain, closer to raw materials or components
- hyperscale
- describing extremely large-scale cloud computing operations
- bottleneck
- a point in a process that slows down or limits the whole system
- scrutiny
- close and careful examination
- gross margin
- the percentage of revenue left after subtracting the direct cost of making a product
- diversify
- to add variety, such as relying on more than one supplier
Talk about this article with KatePractice this story's vocabulary in a real spoken conversation with your AI tutor