Product1 distinct publisher2 min readPublished
AMD gained share while the market it was fighting over shrank by more than a fifth. Run the units and the share win looks like standing still while Intel falls.
The Product Desk · Product desk
Compiled by The Product DeskSomething wrong?How this is made
Set the share numbers against the volume numbers and the 6.5 points look thinner. Taking that figure as percentage points, Intel at 69.3% today [3] entered the year near 75.8%, with AMD's implied x86 share moving from about 24.2% to 30.7% [15][16]. Apply the reported contraction evenly across that base and Intel ships roughly 27% fewer x86 client parts than a year ago, while AMD ships about the same number it shipped before [17]. Six and a half points of share, and flat units to show for them.
The more useful gap is between two firms' counts. IDC put worldwide PC shipments down close to 5% in the second quarter of 2026 [9]; Mercury's annual CPU figure is down by more than four times that [18]. The windows differ, but the residual has to live somewhere, and there are only two plausible addresses: OEMs building from processor inventory they already bought, and the desktop channel that buys CPUs without ever registering as a PC shipment. Gizmodo's reading is that desktops are having a worse time of it than mobile [8], and AMD's gains came mostly from laptop silicon [7], which points the same way.
That is the part worth planning around. A 500% rise in DDR5 [5] is roughly a sixfold price, so a $100 kit is now nearer $600 [19]. Someone facing that does not step down a CPU tier; they stop building, and the CPU sale disappears along with everything else in the basket. Counterpoint has smartphones down 14% year over year on the same logic [10]. Memory-heavy consumer hardware is being deferred as whole units rather than value-engineered, which is a different problem from a bill of materials that got expensive. An inflated BOM shows up as thinner margin and cheaper configurations. A suppressed system market shows up as an absent transaction, and that is what a 20%-plus drop in processor units, against a much shallower drop in finished PCs, describes.
Nothing in the pricing outlook argues with it. Gizmodo's expectation is that neither next-generation PCs nor their components get cheaper, because the hyperscaler demand that emptied the memory market is the same demand both x86 vendors are best paid to serve [13]. The client buyer is competing for supply with the party that caused the shortage, and losing on price. AMD's response so far has been to sell around the problem rather than through it, and Intel's mid-range answer, the Core Ultra 7 270K Plus, has been outshone by AMD's latest parts on Gizmodo's account [20]. Neither is a fix for a market where the memory line dominates the budget.
Ranked by verification strength, evidence, and original report placement.
Recent CPU sales analysis from Mercury Research was cited in a Monday report from CNN detailing x86 market share between AMD and Intel in the PC space.
In the past year Intel lost x86 CPU market share to AMD, reported as 6.5% year over year.
Dean McCarron, president of Mercury Research, told CNN that total CPU shipments this past year were "best described as ugly", with a total downturn "in excess of 20%".
AMD has responded to the component price crisis by bringing back older gaming CPUs as a way to sell PCs built for DDR4 RAM and the aging AM4 socket.
July data from IDC suggested worldwide PC shipments fell by close to 5% in the second quarter of 2026.
Follow any of these and your For You feed starts watching them — no settings page required.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Named analyst numbers, all second-hand and single-publisher
Every quantitative anchor is attributed to a recognizable research firm (Mercury Research, IDC, Counterpoint) and one figure carries a named executive quote, which lifts the floor. But the cluster has exactly one publisher, no primary Mercury report or CNN link content, no vendor comment, an unattributed 500% DDR5 magnitude, a key rebound figure sourced to a social post, and an ambiguous 6.5% that the source never defines as points or relative change.
Real shipment and share data, thin on units and pricing basis
This story is about observed market behavior rather than a technology awaiting uptake, and the supplied material carries several concrete shipment and share disclosures across client CPUs, PCs, smartphones and servers, plus a product action (AMD's AM4/DDR4 revival) that shows vendors reacting to the cost shock. Adoption is not higher because no absolute unit volumes, vendor-level server splits or verifiable price series are supplied, and one client-shipment observation points opposite the others.
Share-win and collapse framing both run ahead of the units
Two overstatements push the same direction. AMD's 6.5-point gain is presented as momentum when applying the reported 20%+ contraction leaves its volume roughly flat and Intel's sharply down, and the 'sales are in the toilet' framing sits in the same article as an embedded report of Q2 client shipments up more than 10% sequentially and record server shipments up 20%. The gap is moderate rather than large because the underlying analyst figures themselves are real and mostly negative.
Visible vendor supply incentive plus consumer-outrage framing
The article itself identifies the central incentive: Intel and AMD are pulled toward AI datacenter customers, the same demand that tightened memory supply, which is why it expects no consumer price relief. Layered on top is the publisher's own incentive as a high-traffic consumer tech outlet, evident in the outrage framing of the headline and lede and in unattributed price magnitudes, plus the commercial interest of research firms whose numbers are being relayed through third parties.
Directionally credible, numerically shaky
The direction of travel (client CPU weakness, cost-driven demand destruction, datacenter strength) is consistent across four independent research firms named in the piece, so the shape of the story is likely right. Confidence stays below the midpoint because there is one publisher, all data is second-hand, the pivotal 6.5% figure is definitionally ambiguous, and the article carries an internal contradiction on client shipment direction that it never resolves.
build
Hugging Face's $13B process puts most teams' model pipeline under a single owner2 distinct publishers
product
Samsung's July foundry price rise moves the AI crunch into your bill of materials1 distinct publisher
invest
Arm starts selling the chip instead of the blueprint, and lands in its licensees' rack slot1 distinct publisher
build
Desktop x86 fell over 20%, and none of the causes Mercury names is a processor1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
1 article · August 25, 2026