What Moves AMD Stock Price: How to Read Any Day's Move

A red day in a chip stock arrives with twenty explanations attached. One feed blames a rival's results, another a government announcement, a third something a chief executive said on a podcast. By the evening the story has settled on whichever version got shared most, and a holder is left guessing which part was real.

Most of those explanations are the same handful of forces wearing different headlines. A chip company's price answers to five things: how fast its biggest business is growing, what it promises for next quarter, what it earns on each sale, what its neighbors report, and what governments allow it to ship. Most of what else gets blamed is decoration on one of those, and the rest is mood about AI spending as a whole.

Those forces can be ranked, and the ranking barely shifts from month to month. Learning it once is worth more than reading a hundred daily explainers, because the next red day will have a different cause and the same short list of candidates.

For AMD, four of the five have a number behind them and a date on which it is published. The data-center line, the guide and the margin arrive with each quarter's results, and its neighbors report on their own schedules. The fifth, what governments decide about chip exports, arrives without a schedule, which is part of why it is the hardest to hold a position through.

What moves AMD stock price?

AMD stock price moves on five things, in rough order of weight: data-center growth, the guide for the next quarter, gross margin, what the rest of the chip sector is doing, and what governments decide about chip exports. A sixth force, sentiment about AI demand, moves it when none of the five has changed.

The order matters more than the list. Data-center growth and the guide explain most results-day moves, and a glance at the rest of the chip sector explains most of the days in between. That keeps a sector-wide selloff from being blamed on something AMD did.

Two things keep this ranking stable, and the first is that AMD's profit has concentrated into the data-center segment, so the number that segment reports dwarfs the rest. The second is that the market prices chips on future demand. Only a few sources say anything credible about it: AMD's own guide, its largest customers, and the governments that decide who may buy what.

That is also why the ranking survives a change of subject. A tariff headline, a rival's earnings and an essay about AI safety look unrelated. Each reaches the price through the same channel: what it implies about how many accelerators get sold next year.

The data-center number, and the guide that comes with it

Data Center is the segment that sells chips into servers and artificial-intelligence systems, and it is now the number the market reads first. In the quarter ended June 27, 2026 it brought in $6.7 billion of AMD's $11.5 billion of revenue, or 58%. It more than doubled from a year earlier (source: AMD Form 10-Q, Q2 2026).

One figure from that same quarter shows why this one segment carries the price. Data Center produced more operating income than the whole company did, because the costs collected outside the segments took about $1.1 billion back out (source: AMD Form 10-Q, Q2 2026). What AMD stock is sets out the full segment split and that arithmetic; for reading a single day, the point is that the data-center line is what the share price now tracks.

The forward number matters even more than the result, and this is the part that trips up new holders. Alongside its second-quarter figures AMD said it "expects revenue to be approximately $13 billion, plus or minus $300 million" for the third quarter. That midpoint sits 41% above the same quarter a year earlier (source: AMD second-quarter 2026 results). That sentence is the guide, or guidance in the language of filings and analyst notes. It is the company's own estimate, published once a quarter, and the market prices the next three months against it from the moment it appears.

So a results day has two scores. The first is what AMD earned in the quarter that just ended, which is history by the time it is published. The second is the guide, what AMD says about the quarter ahead. How to read an AMD earnings report walks through the order the market reads them in.

Gross margin, and why a record quarter can still disappoint

Gross margin is the share of revenue left after the cost of making the chips. For the second quarter of 2026 AMD reported a GAAP gross margin of 54%, up 14 percentage points from a year earlier, and a non-GAAP margin of 56%. It guided the third quarter to a non-GAAP margin of about 56% (source: AMD second-quarter 2026 results).

GAAP means the generally accepted accounting principles AMD's filings follow. The non-GAAP margin leaves out items such as the amortization of intangible assets bought with past acquisitions, which is why it runs higher.

Margin is the reason revenue growth alone tells you little. A company can sell far more chips and earn less on each one, and the market treats that differently from the same revenue at a steady margin. Data-center accelerators carry different economics from console chips, so as the mix shifts the whole company's margin shifts with it, without any single product getting cheaper or more expensive.

A 14-point move in a year is large for this line, and most of it is the comparison rather than the business. The year-ago quarter carried about $800 million of export-control inventory charges, which held its margin near 40%. Strip those out and the starting point is nearer 50%, leaving about three and a half points for everything else, mix included (source: AMD second-quarter 2026 results). That is the first thing to check on any margin jump: whether the prior-year quarter was carrying a one-off charge, because a depressed base flatters the change.

The guided margin deserves as much attention as the guided revenue. A company forecasting flat margin on higher revenue is telling you the mix it expects. If the margin guide slips while the revenue guide holds, the change is in what it expects to sell.

There is a simple way to use this. Read the two guided numbers together, and treat a fall in either as the news. Revenue answers how much AMD sells, and margin answers how good the selling is.

The August 2026 case: revenue up, stock down

The clearest lesson in this pair of numbers came on August 5, 2026. AMD had published second-quarter results the previous afternoon showing revenue up 50% and data-center revenue more than doubled. The stock closed at $482.05 that day, down about 7% (source: Stock Analysis, AMD price history). Yahoo Finance attributed the fall to results and a third-quarter guide that fell short of the highest analyst forecasts, which had reached up to $14 billion (source: Yahoo Finance, August 5, 2026).

Nothing in that sequence was contradictory: the quarter was a record and the guide sat below the highest hopes. A price already carrying those hopes had to give some back. Any reader who sees "record revenue" in a headline and a falling price on the chart is looking at the gap between a result and an expectation.

Why Nvidia and Intel news moves AMD stock

AMD trades as part of a group, and news about one member is read as news about the demand all of them share. A read-across is that borrowed conclusion: Nvidia reports strong accelerator demand, and every accelerator maker rises with it.

AMD's own annual report names the companies it competes with in each market: Intel and Nvidia in server products, Altera in programmable chips, and "a variety of smaller fabless silicon companies" offering accelerators. It adds that "some of our customers are internally developing their own data center microprocessor products and accelerator products" (source: AMD Form 10-K, fiscal 2025). The same filing names Taiwan Semiconductor Manufacturing Company (TSMC) and GlobalFoundries among the third-party foundries that make its wafers, alongside UMC and Samsung for programmable-logic parts, which is why a foundry's results carry information about AMD's costs.

Not every peer tells you the same thing. A supplier's results speak to cost and capacity, a direct rival's speak to share, and a custom-silicon designer's speak to whether AMD's customers plan to build their own chips instead of buying AMD's chips. The three answers can point in opposite directions on the same day.

The test case for how strong this coupling gets came on September 14, 2026. An essay published over the weekend by Anthropic's chief executive argued that AI labs should slow the pace of capability gains. AMD closed down 4.4% that day (source: Stock Analysis, AMD price history). Intel and Nvidia fell with it, and the selling ran across memory and chip names generally, even though the essay "doesn't name a single chipmaker" (source: 24/7 Wall St, September 14, 2026).

That is the single most useful habit in reading a move. Before looking for an AMD explanation, check whether the whole group moved. AMD against Nvidia and AMD against Intel each set out where the two businesses actually compete, which is what decides whether a rival's good news is AMD's bad news or its confirmation.

A handful of customers, and what they intend to deploy

Data-center revenue at this scale comes from a small number of very large buyers, and that concentration is itself a driver. When one customer's plans change, a meaningful share of a segment's forecast changes with it, so a single agreement can move the price more than a quarter of steady laptop sales.

AMD has signed supply agreements measured in gigawatts, a unit that describes the electrical power a data center draws. Its quarterly filing says OpenAI and Meta each intend to deploy up to 6 gigawatts of AMD chips. Each agreement came with a warrant, a right to buy AMD shares at a set price, here $0.01 a share. Each warrant could issue up to 160 million shares if purchase and share-price conditions are met (source: AMD Form 10-Q, Q2 2026). Anthropic added a third agreement, for up to 2 gigawatts of MI450 Series GPUs, the graphics processing units that run AI workloads (source: AMD and Anthropic strategic partnership, July 22, 2026).

None of these agreements is a booked order. Each names a ceiling the customer intends to reach, which is what makes one customer's news able to move a whole quarter's forecast.

Concentration cuts both ways, and a reader watching the price should hold both edges at once. The agreements give the data-center forecast named buyers and named sizes, and they make AMD's revenue sensitive to decisions made in a handful of rooms. A delay at one buyer is a revision to a number the market has already priced. The OpenAI agreement covers how the first of them is built, and the AI and data-center story traces how AMD reached that position.

Export licenses, and why the rules themselves are a driver

Advanced AI chips need export licenses, and the rules move. AMD's annual report describes a license requirement for exports of certain semiconductor products to China. That restriction hit its Instinct MI308 products and cost it about $800 million in inventory and related charges (source: AMD Form 10-K, fiscal 2025).

That charge, taken in the second quarter of 2025, is the clearest illustration of how a rule becomes a number. Chips built for one market and barred from it turn into inventory a company has to write down. The sequel matters as much. Once licenses were granted, AMD began shipping those products and reversed about $360 million of those charges in the fourth quarter. The net cost therefore landed nearer $440 million, and it moved across quarters instead of sitting in one.

The mechanism is worth understanding once, because the headlines recur. A license requirement does not always stop sales: AMD says it applied for, and was granted, some licenses allowing shipments to certain China-based customers. What it does is make revenue conditional on decisions the company cannot schedule. That is why the filing says an inability to predict when licenses will be granted "increases the difficulties of forecasting shipments".

That unpredictability is the driver, more than any single rule. AMD's own report notes that the Bureau of Industry and Security announced plans to rescind one rule and issue replacements, and warns the replacements may require new licenses or delay shipments. So a government announcement can change a forecast while every customer's appetite stays the same. The market reprices the forecast that day.

This class sits last of the five because it fires rarely, and it earns its place because of what it can do when it fires. One rule can remove a market, and no amount of demand replaces a shipment that cannot legally leave. China and export-control risk carries the regime as it stands, which is the part that dates fastest.

How AI sentiment moves AMD stock

Some of the sharpest moves in this stock have no AMD news behind them at all. A change in how the market feels about AI spending reprices every company selling into it. AMD is among the most sensitive, because so much of its profit now depends on that spending continuing.

Two 2026 examples make the class recognizable. From the June 30 close of $580.91, the highest at that point in the year, to the July closing low of $429.56 the stock fell 26%, a slide that ran across four weeks. Behind it sat a report that DeepSeek, a Chinese AI developer, was designing its own AI processor, and a wider worry that AI spending had outrun AI revenue (source: The Motley Fool, August 10, 2026). In September the pacing essay took 4.4% off in a session. Neither fall followed an AMD results release or a change to its guide. The July slide is the sharper lesson. AMD announced the Anthropic agreement on July 22, inside that falling window, and the stock closed up 1.45% that day at $552.33 before resuming its fall the next session (source: Stock Analysis, AMD price history). A 2-gigawatt customer bought one good day, and the market went back to repricing AI spending as a whole.

Telling a sentiment move from a substance move is mostly a matter of asking which number changed. A results day changes revenue, margin and a guide. A customer agreement changes a forecast. An essay changes none of them, and the price still moves, because the price was already carrying an assumption about demand that the essay called into question.

Sentiment moves are the hardest to act on and the easiest to misread, because the explanation arrives after the price. A reader who can place a move in this class has still learned something useful: that no number has changed, and that the next number is the thing to wait for. AMD's 2026 record high takes the run that followed those falls apart in detail, and AMD's price history and cycles shows how often the pattern repeats.

The calendar: Earnings, economic data and options expiry dates

A good part of any year's market news arrives on a calendar published in advance. Results land four times a year, macroeconomic figures land on fixed dates, and options expire monthly. Knowing which events are scheduled turns those days from surprises into appointments.

Rate decisions are the clearest example of a scheduled event with a published time. The Federal Reserve lists its meeting dates a year ahead (source: Federal Reserve, FOMC calendars).

The times matter as much as the dates once you hold the AMDUSDT Perpetual on BloFin, because BloFin publishes its own hours for stock futures. Outside regular weekday hours, and on weekends and market holidays, "stock futures trading may be limited to reduce-only orders" (source: BloFin, stock futures trading hours). Reduce-only means you can close or shrink a position but not open or add to one. Regular hours run 13:30 to 20:00 UTC during US daylight saving time and 14:30 to 21:00 UTC outside it.

The inflation report shows why the time zone matters. The Consumer Price Index is released at 08:30 in the morning on a schedule published months ahead (source: US Bureau of Labor Statistics, CPI release schedule). The monthly jobs report follows the same 08:30 convention on its own Bureau of Labor Statistics (BLS) schedule (source: US Bureau of Labor Statistics, Employment Situation schedule). Neither schedule prints a time zone. Read in Eastern Time, 08:30 is 12:30 UTC in daylight saving and 13:30 UTC in winter, an hour before BloFin's regular hours open in either season. The habit worth building is checking each release time against BloFin's regular hours before the day starts.

A rate decision's own release carries a time and a zone, which the macro schedules do not. The September 16, 2026 statement was marked "For release at 2:00 p.m. EDT", which is 18:00 UTC (source: Federal Reserve, FOMC statement of September 16, 2026). The January 28, 2026 statement reads 2:00 p.m. EST, which is 19:00 UTC, and both land inside BloFin's regular hours (source: Federal Reserve, FOMC statement of January 28, 2026).

Before a scheduled release, the contract's live price and next funding rate are on the AMDUSDT Perpetual page.

Options expiry and positioning, read as a signal

Options are contracts giving the right to buy or sell shares at a set price, and monthly contracts expire on the third Friday. The date is the useful part: it is fixed, published in advance, and it recurs, so a move on one can be placed in the calendar rather than treated as news.

The put/call ratio compares how many bearish contracts are trading against bullish ones, and a sharp change in it says the crowd has shifted. Read it as a crowd thermometer and nothing more. Sizing a position covers what a crowded market does to a leveraged one.

How to read a single day's move

Five checks answer almost any red or green day in AMD, and the order is the method. Each step rules out a whole class of cause, so the first one that fires is usually the answer, and a reader who runs them in sequence rarely needs a sixth.

One: did the sector move? Pull up Nvidia, Intel and a memory name for the same session. If they moved together, the cause is sector-wide and AMD's own news had little to do with it. That can be a rival's results, an export rule that hits every chipmaker, an economic data release or a shift in mood about AI demand.

Two: did AMD publish something that changes a number? Results, a guide revision, a customer agreement, or a filing that moves a forecast. Routine filings go out constantly and move nothing, so the test is whether the thing published changes what AMD is expected to earn. Company events explain many of the largest single-day moves, and they are timestamped, so the check is quick.

Three: was there a policy headline? An export rule, a license decision, or a tariff announcement affecting semiconductors. These arrive without warning and reprice the forecast before any customer changes an order.

Four: was it on the calendar? An inflation report, a rate decision, an options expiry, or an executive speaking at a scheduled investor conference, where a remark about demand can move the stock without any change to the guide.

Five: if none of those fired, it is flow. Index rebalancing, a large fund repositioning, or a crowd reacting to itself. Flow moves are real and hard to predict, and because no number changed behind them, there is nothing new for the next results day to confirm.

Take September 14, 2026 through the sequence. Step one fires immediately: Intel and Nvidia fell on the same day and the selling ran across chip names, so this was a sector move. Step two finds no results, no guide change and no customer agreement. An AMD insider did file a Form 4 that day, the routine disclosure an insider makes after buying or selling shares (source: AMD Form 4, filed September 14, 2026). It touches no forecast, which is exactly the distinction step two asks you to draw.

The Form 4 is still worth a second look, because an insider filing reads like news. It reports a transaction that has already happened, so an insider sale can look like a signal while saying nothing about the next quarter. That leaves step one's answer standing, and the answer also told a holder what to watch next, since a sentiment move leaves the company's own numbers untouched until the next results day.

AMD news that rarely moves the stock

Plenty of AMD news makes headlines and moves nothing: a product launch with no revenue attached, a benchmark win over a rival part, or an award. A keynote that repeats the published roadmap, a forum argument about which chip is better and an analyst note that repeats three others do the same. Each can fill a feed for a morning and leave the price flat.

The test is whether the news changes a number in a future quarter. A launch that a large customer has already committed to buy changes a forecast, and a launch nobody has ordered yet does not. Applying that one question filters most of the noise out.

What the drivers mean for a perpetual position

Holding AMD exposure through the AMDUSDT Perpetual on BloFin adds two considerations to every driver, and both change how a driver's news reaches your position. The first is that the contract keeps trading when the Nasdaq is shut. The second is that holding it costs or pays funding every eight hours.

Funding is a payment exchanged between the two sides of the market, and its sign tells you which side is crowded. Funding printed negative twice on September 23, 2026: the 00:00 and 08:00 UTC settlements each paid longs 0.0120%, after every settlement across September 21 and 22 had been positive, running from 0.0087% to 0.0619%. The 16:00 UTC settlement that day then came back to positive 0.0008%, which is how fast the sign turns (source: BloFin, AMDUSDT funding data). The venue's funding page shows those times in your own timezone, so check the clock against UTC before timing anything by them. Say you hold a $5,000 long: at 0.0120% that is $0.60 a settlement, or about $1.80 a day, paid to you while the rate stays negative and charged to you once it turns positive. A short pays the same amounts the other way. One settlement is a snapshot of crowding on the day, and the direction changes with the crowd. Funding on the AMD perpetual explains how the rate is set and what a sustained run of either sign costs.

The windows matter for the drivers that land outside trading hours. An inflation report arrives in the US morning, before the open. AMD's results arrive after the US close. Both land when orders may be limited to reducing a position. While that limit applies, you can close or shrink what you hold but cannot open or add to it until the window opens. Overnight and weekend gaps covers what a gap does to a leveraged position, and trading around earnings covers the results night specifically.

If you would rather follow a trader with a record, copy trading mirrors another account's positions in yours. AMDUSDT was missing from BloFin's copy-trading list on September 25, 2026, so none of the traders you could copy was able to hold AMD for you that day.

Order types, and how leverage is set on the contract, are the practical side of trading AMD with leverage.

Looking to trade the AMDUSDT Perpetual around AMD's next move? To get started, you'll need to first create a BloFin account, fund your account with cryptocurrency, and navigate to the AMDUSDT Perpetual trading page.

Frequently asked questions

Why is AMD stock down today?

Start with one quote for the whole chip sector, such as the PHLX Semiconductor Sector Index, which Nasdaq publishes from 30 chip companies (source: Nasdaq, PHLX Semiconductor Sector Index). If it fell by a similar amount in the same session, the cause is sector-wide and AMD's own news is secondary. If AMD fell alone, check its investor news page and its Securities and Exchange Commission (SEC) filings, which are both timestamped. An export or tariff headline is the next place to look.

Who are AMD's competitors?

Intel in nearly every segment, plus different rivals in each, named segment by segment in AMD's annual report. In discrete graphics its main rival is Nvidia, which the filing calls the market share leader, alongside Intel, which sells integrated graphics and gaming-focused discrete GPUs. In the Embedded segment the named rivals include the programmable-chip makers Altera, Lattice and Microsemi. The filing also lists chip vendors such as Broadcom, Marvell, Texas Instruments and Qualcomm, with Intel, the maker of central processing units (CPUs), named as "our main competitor for embedded CPUs" (source: AMD Form 10-K, fiscal 2025).

Does AMD move with Nvidia?

Often, and the coupling is tightest on news about demand for AI chips, which both companies sell into. It loosens on company-specific news: an Nvidia product delay can help AMD, and an AMD customer agreement can move AMD alone. The practical rule is that a move shared with Intel and Broadcom as well is a sector move, while a move AMD makes by itself usually has an AMD cause behind it.

When does AMD report results?

Four times a year, after the US market closes, with each quarter ending on a Saturday because AMD uses a 52- or 53-week fiscal year that ends on the last Saturday in December (source: AMD Form 10-K, fiscal 2025). The second quarter of 2026 ended on June 27 and the results were published on August 4. Exact dates are posted on AMD's investor relations calendar, and a results release after the close lands in a window where the perpetual may be limited to reduce-only orders.

Do interest rates move AMD?

Yes, indirectly, through two channels. Higher rates make distant profits worth less today, which pressures the most expensive stocks hardest, and AMD traded at about 160 times its trailing earnings at the September 24, 2026 close (source: Yahoo Finance, AMD). Rates also affect how freely AMD's customers borrow to build data centers. The next decision on the Fed's 2026 calendar is October 28 (source: Federal Reserve, FOMC calendars). That is before US clocks change on November 1, so a statement released at the usual 2:00 p.m. Eastern lands at 18:00 UTC, inside BloFin's regular hours.

Researched and written by the BloFin Academy editorial team with AI-assisted drafting. All facts independently verified. Primary sources include AMD's second-quarter 2026 results release, its Form 10-Q for the quarter ended June 27, 2026, its Form 10-K for fiscal 2025 and its Form 4 of September 14, 2026 filed with the SEC, the Stock Analysis and Yahoo Finance price records, The Motley Fool, Yahoo Finance and 24/7 Wall St reporting of August and September 2026, Nasdaq's PHLX Semiconductor Sector Index page, the US Bureau of Labor Statistics release schedules, the Federal Reserve's FOMC calendars and statements of January 28 and September 16, 2026, and BloFin's stock-futures trading notice and AMDUSDT funding data, current as of September 25, 2026.

Nothing in this article constitutes financial advice, and nothing in it is a recommendation to buy, sell or hold AMD in any form. The drivers described here explain how AMD's price has moved in the past, and the same catalyst can move it the other way next time. Results and many policy headlines land outside the Nasdaq session, when BloFin may accept only orders that reduce a position, so a leveraged AMDUSDT Perpetual position can meet the new price all at once and be liquidated by a move far smaller than the leverage multiple suggests. Past performance does not indicate future results. Do your own research and consider your risk tolerance before you trade on BloFin.