AMD’s Zen 6 “Olympic Ridge” Ryzen Desktop CPUs Reportedly Target 24-Core Peak, Spanning 6 to 24 Cores Across Seven SKUs

AMD is gearing up to refresh its Ryzen desktop lineup with a new generation of processors based on the Zen 6 CPU architecture. The chips are expected to arrive under the “Olympic Ridge” codename and are shaping up to be a meaningful upgrade for the AM5 platform, bringing a mix of architectural refinements, higher performance per clock (IPC), faster boost behavior, and more flexible core counts. Buyers can also expect continued momentum around AMD’s advanced 3D V-Cache (X3D) stacking strategy, which has become a major draw for gamers and performance-focused PC builders.

While AMD has been relatively quiet publicly about Zen 6 details on the consumer side, earlier information tied to its server roadmap has hinted at the direction the company is taking. On the data center front, Zen 6-based designs have been described with very high core-density options and larger cache configurations per CPU chiplet (CCD). Those enterprise-class numbers don’t translate directly to desktop parts, but they do suggest AMD is evolving its chiplet building blocks in ways that can benefit mainstream users too.

Now, new chatter around Ryzen “Olympic Ridge” points to a broader and more granular set of desktop SKUs than the current generation. Instead of sticking to the familiar 6-, 8-, 12-, and 16-core structure seen recently, Zen 6 desktop CPUs are rumored to span at least seven different core configurations split between single-CCD and dual-CCD designs.

Here’s what the rumored Zen 6 “Olympic Ridge” desktop core lineup looks like:

Single-CCD Zen 6 models
6 cores
8 cores
10 cores
12 cores

Dual-CCD Zen 6 models
16 cores (8+8)
20 cores (10+10)
24 cores (12+12)

If this holds true, the biggest headline is a potential 24-core Ryzen desktop flagship built from two 12-core CCDs. That would give AMD more room to tune pricing and performance across the stack, while also offering better “in between” options for people who want more than 16 cores without necessarily jumping to workstation-class platforms.

Another key detail behind these rumored configurations is the CCD itself. The Zen 6 CCD is expected to move to a 12-core layout paired with a larger L3 cache pool than prior mainstream CCDs. For context, AMD’s earlier chiplet generations generally centered on 8-core CCDs with 32MB of L3 cache, while the updated Zen 6 building block is rumored to step up to 12 cores and 48MB L3. That combination could improve multi-threaded throughput for creators and power users, while also opening the door for new X3D-enhanced variants that could boost gaming performance.

Competition is also heating up on the other side of the aisle. Intel’s next major desktop push is expected to arrive under the Nova Lake-S family, with reports suggesting the company will scale from single compute tile designs to more extreme dual compute tile parts that can push core counts far beyond today’s mainstream desktops. However, those top-end Intel configurations are also rumored to demand much higher power budgets, which makes direct comparisons tricky. In practical terms, it’s likely we’ll see the more typical Intel configurations positioned against AMD’s high-end 24-core offerings, while Intel’s ultra high-core-count chips could sit in a more expensive, higher-wattage tier.

From a buyer perspective, that’s an important distinction. Core counts matter, but so do total platform cost, cooling requirements, and power draw. If AMD can deliver 20- and 24-core options within a familiar pricing and power envelope for AM5 systems, it could be an attractive upgrade path for anyone already invested in the socket—especially users who do heavy multitasking, content creation, compiling, or other multi-threaded workloads.

Assuming these roadmaps land as expected, the second half of 2026 could be a particularly interesting window for desktop PC upgrades. New CPU architectures, higher core count options, and fresh performance features are all on the table—now the big questions will be final clock speeds, cache variants (including X3D models), real-world power behavior, and where pricing settles as competition intensifies.