AMD GDDR6 Price Increase 2026: Nvidia Holds on GDDR7

AMD has told its graphics card board partners that GDDR6 memory prices are going up starting October 1, 2026, according to reports from Wccftech and TechPowerUp. The notice, which the outlets attribute to board channel sources, lands at an awkward moment for the add-in board (AIB) makers that build and sell Radeon graphics cards, since it arrives in the thick of a broader memory shortage that has already pushed DRAM and NAND prices higher across the PC industry. Nvidia, by contrast, has reportedly not sent its own partners a similar notice covering GDDR7, the memory standard that powers its GeForce RTX 50 series, as of the October 10, 2026 reporting date.

Neither AMD nor Nvidia has issued a public statement confirming the figures, and no exact dollar amount, percentage increase, or specific module pricing has been verified. What is confirmed, per the reporting, is the timing and the direction: AMD is raising GDDR6 costs for partners, and Nvidia has so far stayed quiet on GDDR7. That gap, however it plays out over the next few months, is the detail worth tracking, because memory pricing feeds directly into graphics card bills of materials and, eventually, into what gamers and system builders pay at retail.

Google · Preferred Sources

Don't miss new tech stories on Google

Add TrendinTech once in the Google app and our stories appear in your news suggestions.

Add Now

What AMD Reportedly Told Its Board Partners

According to the board-channel sourcing cited by Wccftech and TechPowerUp, AMD notified its graphics card partners of a GDDR6 price increase effective October 1, 2026. The notice reportedly went out to the AIB partners that assemble retail Radeon cards using AMD’s GPU dies, rather than to end consumers directly, which is standard practice when a component supplier adjusts pricing mid-cycle. Neither outlet published the specific percentage or per-module dollar figure AMD is charging, and no confirmed list of affected graphics card models has surfaced.

That lack of a hard number matters for how this story should be read. It is easy to assume a memory price hike from a major supplier automatically means a retail price hike of similar size, but that chain has several links in it: AIB margins, existing GDDR6 inventory already purchased at older prices, and how much headroom each partner has before a change becomes visible on a store shelf. AMD’s current Radeon RX 9000 series, including the RX 9070 XT compared in detail in our RTX 5080 vs RX 9070 XT benchmark breakdown, relies on GDDR6 memory, which makes this specific supplier a direct cost input for AMD’s current flagship consumer GPU lineup.

Nvidia’s Silence on GDDR7 Pricing

The other half of this story is what did not happen. As of the October 10, 2026 reporting window, Nvidia had not notified its own board partners of any GDDR7 price increase. Nvidia’s GeForce RTX 50 series uses GDDR7 memory, a newer standard that Nvidia adopted earlier and more aggressively than AMD did for its mainstream Radeon lineup. The absence of a notification is not the same as a guarantee that GDDR7 pricing will stay flat through the rest of 2026 — the reporting only confirms that no partner notice had gone out at the time these outlets published, not that Nvidia has made any long-term pricing commitment.

That distinction is worth sitting with, because it is tempting to read “Nvidia holds off” as a strategic masterstroke and “AMD hikes prices” as a misstep. The more accurate framing, based on what has actually been reported, is that the two companies currently sit in different positions on the memory-pricing timeline, for reasons that have not been publicly detailed by either supplier. Memory suppliers like Samsung, SK Hynix, and Micron set wafer and module pricing based on their own capacity allocation decisions, and GDDR6 and GDDR7 are not drawing from identical supply pools.

GDDR6 vs GDDR7: Why the Memory Standard Matters

To understand why AMD and Nvidia might be facing different cost pressures, it helps to know what separates GDDR6 from GDDR7 as memory standards. GDDR7 is the newer JEDEC-ratified specification, published as JESD239, and it represents a generational jump in how graphics memory moves data. According to Tom’s Hardware’s reporting on the JEDEC GDDR7 standard, the new specification roughly doubles per-device bandwidth over GDDR6, and it switches from the two-level NRZ signaling used in GDDR6 to a three-level PAM3 signaling scheme that packs more data into the same number of clock cycles.

GDDR6 remains a mature, widely manufactured standard that multiple memory makers have been producing in volume for years, which historically has made it cheaper and easier to source at scale. GDDR7 is newer, more complex to manufacture, and currently used in a narrower set of products, concentrated mostly in Nvidia’s current-generation flagship and upper-midrange GeForce cards. That maturity gap is part of why a supply or pricing shock can hit the two standards on different schedules: GDDR6’s broad, established production base can absorb demand shifts differently than GDDR7’s comparatively newer and more specialized manufacturing lines.

Signaling and Bandwidth Differences

The practical upshot for buyers is that GDDR7-equipped cards generally carry a bandwidth advantage per memory chip, which is one reason Nvidia leaned into the standard for its latest RTX 50 series flagships. AMD’s current mainstream Radeon RX 9000 series, by contrast, has stayed on GDDR6, a choice that keeps bill-of-materials costs more predictable in normal market conditions but also means AMD is more exposed when a GDDR6-specific price action like this one happens.

Attribute	GDDR6	GDDR7
JEDEC status	Established standard, in volume production for several years	Newer standard (JESD239), ramping production since late 2024
Signaling method	NRZ (two-level signaling)	PAM3 (three-level signaling)
Relative bandwidth per device	Baseline	Roughly double GDDR6, per JEDEC/Tom’s Hardware reporting
Primary current use	AMD Radeon RX 9000 series and mainstream GPUs industry-wide	Nvidia GeForce RTX 50 series flagships
Manufacturing maturity	Broad, multi-vendor, high volume	Narrower vendor base, still scaling
Reported October 2026 pricing action	AMD price increase to board partners, effective Oct. 1	No partner price notice reported as of Oct. 10

The AI-Driven Memory Squeeze Behind the Price Hike

This specific GDDR6 price move does not exist in a vacuum. It lands in the middle of what has become one of the most disruptive memory market cycles in years, driven largely by AI data center demand. Tom’s Hardware has reported that memory price increases, while beginning to cool slightly for some consumer categories, are still climbing for DRAM and NAND through the third quarter of 2026 as AI infrastructure buildouts continue absorbing a growing share of global chip output. JPMorgan’s research desk has similarly flagged the AI-driven memory shortage as a structural issue rather than a short-lived blip, noting that data center and AI accelerator demand is competing directly with consumer electronics for the same fabrication capacity at Samsung, SK Hynix, and Micron.

Graphics memory is not immune to that dynamic, even though GDDR6 and GDDR7 are distinct product lines from the DDR5 system memory most often cited in AI shortage coverage. Memory fabs allocate wafer capacity across product families based on margin and demand signals, and when AI customers are willing to pay a premium for high-bandwidth memory and server DRAM, that pressure can ripple into adjacent product lines, including the graphics memory that goes into gaming GPUs. AMD’s GDDR6 notice to partners is consistent with that broader pattern, even though neither AMD nor its memory suppliers have publicly tied this specific increase to AI demand.

Historical Context: Memory Shocks Have Hit GPU Pricing Before

Graphics card buyers have been through memory-driven price disruptions before, most memorably during the 2020-2021 cycle when GDDR6 scarcity combined with cryptocurrency mining demand and pandemic-era supply chain breakdowns to push GPU prices well above MSRP across the board. That episode eventually cooled once mining demand collapsed and foundries caught up on capacity, but it left a lasting lesson in the GPU market: component-level shortages do not stay contained to the component. They show up in finished-product pricing, often with a lag of one to two product cycles.

The current cycle differs in one important respect. The 2020-2021 shortage was driven by a demand spike that was expected to be temporary, tied to mining profitability that could evaporate overnight. The 2026 memory squeeze is being driven by AI infrastructure investment that major cloud providers have committed to for years, not quarters. That is part of why analysts covering the broader DRAM and NAND market have described this cycle as more structural than cyclical, which, if accurate, suggests graphics memory pricing pressure may not resolve itself as quickly as it did after the mining-driven shortage faded.

Who Uses GDDR6 and GDDR7 Across Current GPU Lineups

Understanding which products are exposed to which memory standard helps clarify who feels this price action first. The table below lays out the current split across the two companies’ mainstream and flagship desktop GPU stacks, based on publicly documented product specifications.

GPU Segment	Memory Standard Used	Vendor	Exposure to This Price Action
Radeon RX 9000 series (mainstream/flagship)	GDDR6	AMD	Direct — covered by the reported Oct. 1 increase
GeForce RTX 50 series (flagship/upper midrange)	GDDR7	Nvidia	None reported as of Oct. 10, 2026
Entry-level and prior-generation discrete GPUs	Mostly GDDR6	Both AMD and Nvidia	Potentially indirect, pending further notices
Handheld gaming PCs and APUs	Shared system memory, not GDDR6/7	AMD-based designs	Not directly affected by this specific notice

That last row matters for a growing segment of the market. Handheld gaming devices like the ones compared in our ROG Ally X20 vs Legion Go 2 vs AYN Thor handheld comparison and our broader Steam Deck and gaming handheld pricing roundup typically rely on unified system memory rather than dedicated GDDR6 or GDDR7 modules, which means this particular notice does not directly touch that category, even as those devices remain exposed to the broader DRAM squeeze described above.

Market Impact: What This Means for Buyers and Builders

For anyone shopping for a graphics card in the next few months, the honest answer is that nothing has definitively changed yet at retail. AMD’s notice to partners is a wholesale-level cost signal, not a confirmed consumer price increase. Board partners have several ways to absorb a supplier price bump before it reaches a shelf price tag, including existing inventory bought at pre-increase rates, margin compression on a given SKU, or spreading the added cost across a wider product line rather than concentrating it on one card.

Short-Term Buyers

Shoppers looking to buy a Radeon card in the immediate term are unlikely to see a direct price jump tied to this specific notice within days of October 1. Retail pricing changes on existing inventory typically lag supplier cost changes by weeks or months, and no AIB has publicly announced a resulting MSRP change as of this writing.

System Builders and OEMs

Builders assembling new systems around current-generation Radeon cards, and OEMs with existing supply agreements, are the group most likely to feel any downstream cost pass-through first, since they tend to be closer to the wholesale pricing layer than individual retail buyers.

AMD vs Nvidia: Diverging Memory Strategies

This episode highlights a quieter competitive dynamic between AMD and Nvidia that rarely gets attention outside of supply chain circles: the two companies are not buying the same memory, from the same vendors, at the same volumes. Nvidia’s early and aggressive move to GDDR7 for its GeForce RTX 50 series, detailed on Nvidia’s own GeForce 50 series product pages, gave it a bandwidth advantage on paper, but it also tied Nvidia’s flagship lineup to a newer, less mature supply chain with fewer established vendors. AMD’s decision to stay on GDDR6 for its current Radeon generation, examined in our Snapdragon X2 vs Intel Core Ultra chip pricing comparison coverage of the broader 2026 silicon pricing landscape, kept AMD on a cheaper, more established memory standard, right up until that standard’s own supplier decided to raise prices.

Neither approach is inherently safer. GDDR7’s narrower vendor base could just as easily see a price shock of its own once Nvidia’s suppliers decide capacity allocation warrants one, and GDDR6’s maturity did not insulate AMD’s partners from the increase reported this week. What this really demonstrates is that memory sourcing has become a genuine point of competitive differentiation between the two GPU makers, not just a background engineering decision.

The Board Partner Dilemma

AIBs sit in an uncomfortable middle position whenever a GPU vendor’s memory supplier raises prices. They do not set GDDR6 pricing, AMD does not fully control it either since AMD is itself buying from memory manufacturers like Samsung, SK Hynix, and Micron, and yet board partners are the ones who ultimately decide whether to raise retail prices, shrink margins, or quietly adjust specifications on lower-tier SKUs. Public filings from AMD, available through AMD’s investor relations site, do not currently break out GDDR6 cost pass-through as a separate disclosed line item, which is typical for component-level pricing changes that are considered commercially sensitive.

Analysts who track semiconductor supply chains, including research groups like TrendForce, have long noted that memory cost volatility is one of the hardest variables for AIBs to plan around, precisely because it can shift mid-quarter and apply unevenly across a product’s lifecycle. A card that launched with healthy margins six months ago can see those margins compressed by a supplier price change with little warning, which is exactly the position AMD’s partners now appear to be in.

Competitive Comparison: Reported Pricing Posture, October 2026

Putting the two companies side by side based strictly on what has been reported, rather than speculation, clarifies the current state of play heading into the end of 2026.

AMD: Reportedly notified board partners of a GDDR6 price increase effective October 1, 2026, per Wccftech and TechPowerUp. No confirmed dollar figure, percentage, or affected-model list has been published.
Nvidia: Had not notified partners of any GDDR7 price increase as of October 10, 2026, per the same reporting. No long-term pricing commitment has been confirmed either way.
Consumer pricing: No retail price increases have been confirmed as resulting from either company’s memory situation as of this report.
Underlying driver: Both companies are operating inside a broader DRAM and NAND market that Tom’s Hardware and JPMorgan have both described as under sustained pressure from AI data center demand through at least Q3 2026.

Predictions: Where GPU Memory Pricing Goes From Here

Based on the current reporting and the broader memory market backdrop, a handful of outcomes look more likely than others over the next two to three quarters, though none of these should be read as confirmed facts, since AMD, Nvidia, and their memory suppliers have not detailed their forward plans publicly.

AMD’s board partners will likely absorb at least part of the GDDR6 increase through margin compression before passing meaningful costs on to consumers, following the same playbook AIBs have used during prior component cost swings.
Nvidia’s GDDR7 pricing holiday, if it continues, probably will not last the full remainder of 2026 given how tight the broader memory market has become, though no specific trigger date has been reported.
Entry-level and previous-generation GDDR6-based cards from both vendors are the segment most likely to see retail price adjustments first, since those products run on thinner margins to begin with.
Expect more AIB-level commentary, rather than direct statements from AMD or Nvidia, to be the primary source of confirmation as this story develops, consistent with how memory pricing changes have surfaced in past cycles.
If the AI-driven DRAM and NAND squeeze persists into 2027 as some analysts expect, graphics memory of both standards could face renewed pricing pressure regardless of which company moves first this time.

What to Watch Next

The clearest next signal will be whether AMD’s board partners announce any retail pricing adjustments on Radeon RX 9000 series cards in the coming weeks, or whether they manage to hold current street prices steady through existing inventory. On the Nvidia side, the question is simply whether a similar GDDR7 notice eventually surfaces, and if it does, whether it comes with a larger or smaller adjustment than AMD’s GDDR6 move, given GDDR7’s newer and narrower supply base. Readers tracking GPU value across both ecosystems may also want to keep an eye on how this plays into next-generation planning, an area our Switch 2 vs ROG Xbox Ally X hardware comparison touches on from the handheld angle, where memory and silicon costs feed directly into device pricing.

Until either company issues an on-record statement, or an AIB partner confirms a specific number, the safest summary remains the one the original reporting supports: AMD has raised GDDR6 costs for its partners starting October 1, 2026, and Nvidia had not done the same for GDDR7 as of October 10, 2026. Everything beyond that, including the eventual size of any consumer-facing price change, remains unconfirmed.

FAQ

Did AMD officially confirm a GDDR6 price increase?
AMD has not issued a public statement confirming a GDDR6 price increase. The information comes from board-channel sourcing reported by Wccftech and TechPowerUp, dated around October 10, 2026.

What is the exact percentage or dollar increase AMD is charging?
No confirmed figure has been published. The available reporting does not include a specific dollar amount, percentage, or per-module price for the GDDR6 increase.

Will graphics card prices go up because of this?
That is not confirmed. AMD’s notice was reportedly sent to board partners at the wholesale level, and no retail price increases on Radeon graphics cards have been confirmed as a direct result.

Is Nvidia raising GDDR7 prices too?
As of October 10, 2026, Nvidia had reportedly not notified its board partners of any GDDR7 price increase. This does not guarantee GDDR7 pricing will stay unchanged going forward, only that no notice had been issued at the time of reporting.

Why is AMD on GDDR6 while Nvidia uses GDDR7?
AMD’s current Radeon RX 9000 series uses GDDR6, a more established and widely manufactured memory standard. Nvidia’s GeForce RTX 50 series uses GDDR7, a newer JEDEC standard with higher bandwidth per device, as detailed in JEDEC’s GDDR7 specification coverage.

Is this related to the broader AI memory shortage?
The AMD notice itself has not been publicly tied to AI demand by AMD or its suppliers. However, it arrives during a period that Tom’s Hardware and JPMorgan have both described as a sustained, AI-driven DRAM and NAND supply squeeze affecting the broader memory market through at least the third quarter of 2026.

Which graphics cards are affected by the GDDR6 price increase?
No confirmed list of affected models has been published. Any current AMD Radeon GPU that uses GDDR6 memory, which includes the RX 9000 series, could potentially see cost impacts at the board-partner level, but no specific SKUs have been named in the reporting.

When will we know the actual size of the price increase?
There is no confirmed timeline for when AMD, Nvidia, or their board partners will disclose specific figures. Historically, these details tend to surface through AIB partner statements or earnings calls rather than direct supplier announcements.

AMD Hikes GDDR6 Prices Oct. 1; Nvidia Holds on GDDR7 [2026]