The AMD Ryzen Z1 Extreme is a fast high-end laptop processor of the Phoenix series and fastest model of the U-series. It offers 8 cores (octa core) based on the Zen 4 architecture that supports SMT (hyperthreading , 16 threads). The cores clock from 3.3 (base) up to 5.1 GHz (single core boost). The processor includes 8 MB L2 cache and 16 MB L3 cache.
The performance of the R7 7840U is, depending on the workload, only slightly below the Ryzen 7 7840HS (higher base clock speed and TDP). Compared to the previous generation Ryzen 7 PRO 6850U, the 7840U is clearly faster thanks to the Zen 4 architecture and higher clock speeds. Compared to the similar Ryzen 7 7840U, the Z1 Extreme offers no AI engine.
The chip integrates a modern and fasts RDNA 3 graphics card (iGPU) called Radeon 780M with 12 CUs and up to 2.7 GHz clock speed. Furthermore, the Phoenix series include a video engine with AV1 de- and encoding, a new Xilinx FPGA based XDNA AI accelerator (Ryzen AI) that should be faster than the AI engine in the Apple M2 SOC and a dual-channel DDR5-5600 / LPDDR5x-7500 memory controller (with ECC support). The connectivity features includes 2 possible USB 4 (40 Gbps) ports and 20 PCIe 4.0 lanes for a GPU and SSDs.
The Phoenix series uses a single monolithic design (unlike the chiplet design of the 7045HX series) and is manufactured in the modern 4nm process at TSMC. The TDP can be configured between 15 and 30 Watt.
The Ryzen Z2 Extreme is a relatively powerful Strix Point family processor that debuted in early 2025. The hybrid architecture APU has 8 CPU cores onboard (a mix of three Zen 5 and five Zen 5c cores) running at 2.0 to 5.0 GHz, along with the 16 CU RDNA 3.5 Radeon 880M graphics adapter and the 50 TOPS XDNA 2 neural engine. Compared to the similar Ryzen AI 7 PRO 360, the Z2 Extreme does not offer the Ryzen AI NPU.
Architecture and Features
Strix Point family APUs are powered by Zen 5 and Zen 5c microarchitecture cores found in two separate clusters, the latter being a slightly slower, smaller and more energy-efficient version of the former. One of the differences between Zen 5 and Zen 5c is cache size; Zen 5 cores have larger caches to work with.
Either way, mobile Zen 5 implementation is reportedly (ChipsAndCheese) closer to desktop Zen 4 than to desktop Zen 5 due to differing cache sizes, vast differences in AVX-512 throughput and other factors.
Elsewhere, the Ryzen chip supports DDR5-5600 and LPDDR5x-8000 RAM, giving system designers a choice between lower latency and higher throughput respectively. The chip is natively compatible with USB 4 (and therefore Thunderbolt). It has PCIe 4.0 support for a throughput of 1.9 GB/s per lane, just like its 8000 series predecessors did. The integrated XDNA 2 NPU, which is a lot more complex than first-gen XDNA was, delivers up to 50 INT8 TOPS for accelerating various AI workloads.
As is usual for mobile CPUs, the Ryzen 7 AI PRO chip is not user-replaceable as it gets soldered down for good.
Performance
The performance should be similar to the Ryzen AI 7 PRO 360 due to the similar specs, but limited due to TDP and heat restrictions in handhelds.
Graphics
The Radeon 880M is the direct successor to the 780M. It packs several differences under the hood, such as faster caches; its 16 RDNA 3.5 architecture CUs/WGPs (768 unified shaders) run at a currently undisclosed clock speed. With well over 30 fps in both Once Human and The First Descendant (1080p - Low), this AMD iGPU is usually certainly fast enough for a fair bit of casual as well as competitive gaming.
Power consumption
The Z2 Extreme is supposed to have a low-term TDP power target of 28 W, with handheld makers free to set between 15 and 35 W.
The 4 nm TSMC process that these CPUs are built with makes for above average, as of late 2024, energy efficiency.
Average Benchmarks AMD Ryzen Z1 Extreme → 100%n=24
Average Benchmarks AMD Ryzen Z2 Extreme → 95%n=24
- Range of benchmark values for this graphics card - Average benchmark values for this graphics card * Smaller numbers mean a higher performance 1 This benchmark is not used for the average calculation
v1.35
log 04. 16:43:54
#0 ran 0s before starting gpusingle class +0s ... 0s
#1 checking url part for id 15017 +0s ... 0s
#2 checking url part for id 18735 +0s ... 0s
#3 redirected to Ajax server, took 1767541433s time from redirect:0 +0s ... 0s
#4 did not recreate cache, as it is less than 5 days old! Created at Sun, 04 Jan 2026 05:18:32 +0100 +0s ... 0s
#5 composed specs +0.007s ... 0.007s
#6 did output specs +0s ... 0.007s
#7 getting avg benchmarks for device 15017 +0.004s ... 0.011s
#8 got single benchmarks 15017 +0.008s ... 0.019s
#9 getting avg benchmarks for device 18735 +0.004s ... 0.023s
#10 got single benchmarks 18735 +0.006s ... 0.029s
#11 got avg benchmarks for devices +0s ... 0.029s
#12 min, max, avg, median took s +0.033s ... 0.062s