SYSTEM ANALYZER

Rate My PC: AMD Ryzen AI Max 390 + NVIDIA GeForce RTX 4070 Mobile

Get a comprehensive performance analysis of your gaming rig with detailed benchmarks, bottleneck detection, and upgrade recommendations

94 / 100
ULTIMATE READY

Apex Performer

Top 6% of systems. Capable of 4K Ultra gaming and advanced rendering.

4K 60+ FPSVR ReadyRay Tracing

System Balance Analysis

CPU vs GPU performance ratio
Well Balanced
CPU
95%
VS
GPU
93%
PROCESSOR

AMD Ryzen AI Max 390

56,273 Benchmark Score
Top 5% Market Ranking
View Full Specs →
GRAPHICS CARD

NVIDIA GeForce RTX 4070 Mobile

27,435 Benchmark Score
Top 7% Market Ranking
View Full Specs →

Market Position

How your build compares to others
Budget
0-30
Mid-Range
30-60
High-End
60-85
Enthusiast
85-100
Your Build

Game Performance Benchmarks

Real-world 4K FPS in popular titles
View All Games →

Performance Insights

Tips to maximize your system

Optimal Performance

Your system is in the top tier. You can run any modern game at maximum settings.

4K Gaming Ready

Consider a 4K 144Hz monitor to fully utilize your hardware capabilities.

Compatible Games See what you can play Compare CPUs Find upgrades Compare GPUs Find upgrades

Performance Tiers Explained

90-100

Ultimate

4K Ultra gaming, VR ready, ray tracing enabled, professional workloads

4K 60+ FPS VR Ready
70-89

High-End

1440p Ultra or 4K High settings, excellent for modern AAA titles

1440p Ultra 4K High
50-69

Mid-Range

1080p Ultra or 1440p Medium, great value for most gamers

1080p Ultra 1440p Med
30-49

Entry Level

1080p Medium settings, suitable for eSports and older titles

1080p Med eSports
0-29

Legacy

Basic gaming, older titles, consider upgrading for modern games

720p-1080p Low Older Games

# AMD Ryzen AI Max 390 + NVIDIA GeForce RTX 4070 Mobile

This pairing combines AMD’s 12-core Zen 5 Strix Halo processor with NVIDIA’s Ada Lovelace mobile GPU in a laptop configuration. The CPU sits at the 91st percentile among all processors, while the GPU reaches the 73rd percentile among all graphics cards, yielding a combined percentile of 82. The data here is drawn purely from synthetic benchmarks, as no measured FPS rows exist for this exact CPU+GPU combination; all gaming discussion is therefore framed as estimates derived from those scores.

CPU Analysis

The AMD Ryzen AI Max 390 is built on the Zen 5 architecture with the Strix Halo codename, manufactured on TSMC’s 4 nm process with a die composed of two 70.6 mm² chiplets. It offers 12 cores and 24 threads, a base clock of 3.20 GHz, and a boost clock of 5.00 GHz. The cache hierarchy includes 80 KB of L1 per core, 1 MB of L2 per core, and a substantial 64 MB of shared L3 cache — a configuration that favors workloads with large working sets, such as code compilation or complex data processing.

The processor’s Cinebench scores are revealing. In Cinebench R23, it scores 36,064 multi-core and 5,091 single-core. The multi-core result places it within striking distance of its nearest rivals: it trails the AMD Ryzen AI 9 HX PRO 470 by just 0.1% (average score 56,306 vs. 56,273 for the Ryzen AI Max 390 on the aggregate benchmark), while edging out the Intel Core i9-14900HX by 0.5%. The single-core R23 score of 5,091 suggests strong per-thread performance, which matters for lightly threaded applications like legacy software or certain game engines that rely on one or two threads.

PassMark results add texture. The multi-thread score of 41,737 and single-thread score of 4,028 align with a high-end mobile CPU. Data compression reaches 487,145, while encryption hits 25,097 — the latter indicating robust AES-NI and related instruction throughput. Integer math (146,519) outpaces floating-point math (90,594), which is typical for a Zen 5 part and suggests the CPU handles logic-heavy tasks (spreadsheets, databases, scripting) more readily than heavy scientific floating-point workloads. Extended instructions score 38,716, and random string sorting reaches 53,113, both pointing to solid memory subsystem performance enabled by the 256.0 GB/s quad-channel LPDDR5X bandwidth.

The 55 W TDP is modest for a 12-core part, yet the benchmark scores rival desktop-class chips like the AMD Ryzen Threadripper PRO 3955WX (which the Ryzen AI Max 390 trails by just 0.5% on aggregate). This efficiency comes from the 4 nm process and the unified memory architecture. For real workloads, expect strong multi-tasking, quick code builds, and smooth 4K video editing timelines — all within a laptop thermal envelope.

Benchmark Performance

The CPU’s average benchmark score is 56,273, placing it at the 91st percentile of all CPUs. Its nearest rival, the AMD Ryzen AI 9 HX PRO 470, scores 56,306 (0.1% higher), while the Intel Core i9-14900HX scores 56,004 (0.5% lower). This means the Ryzen AI Max 390 sits in a tight cluster of top-tier mobile and workstation processors — any difference is within noise, but the data shows it holds its own against Intel’s flagship mobile HX part.

On the GPU side, the NVIDIA GeForce RTX 4070 Mobile achieves an average benchmark score of 27,435, landing at the 73rd percentile of all GPUs. Its nearest rival is the AMD Radeon RX 6700 XT with a score of 27,425 (0% delta), followed by the NVIDIA GeForce RTX 3090 at 27,565 (0.5% higher) and the NVIDIA RTX PRO 4000 Blackwell at 27,135 (1.1% lower). The RTX 4070 Mobile thus performs comparably to a desktop RTX 3090 in aggregate synthetic terms — a remarkable result for a laptop GPU, though the 3090’s larger memory pool (not in this data) would favor it in VRAM-bound scenarios.

The GPU’s individual benchmark scores show a mixed profile. PassMark G3D scores 19,587, which is strong for mobile hardware. DirectX 12 performance scores 85, while DirectX 11 scores 179 and DirectX 9 scores 223 — the higher scores on older APIs are typical of how these tests scale. Compute performance via PassMark GPU Compute is 8,399, while Geekbench OpenCL reaches 109,197 and Vulkan 108,367. The close OpenCL and Vulkan numbers suggest balanced compute capability across APIs, which bodes well for cross-platform rendering and machine learning inference.

Combined, the CPU’s 91st percentile and GPU’s 73rd percentile yield a system-level combined percentile of 82. This indicates the CPU is the stronger component relative to its peers, while the GPU, though capable, is less exceptional. The data implies a pairing where the processor can feed the graphics card without bottlenecking in most CPU-bound scenarios, but the GPU will cap frame rates in GPU-bound titles.

Balance and Bottleneck

The percentile gap — 91 for CPU versus 73 for GPU — is the first clue to a CPU-heavy balance. In synthetic aggregate scores, the CPU is 18 percentile points stronger. This suggests that in mixed workloads, the processor will often wait on the graphics card, particularly at higher resolutions where pixel throughput becomes the limiting factor.

The GPU’s PassMark G3D score of 19,587, while solid, is not in the same league as the CPU’s aggregate percentile position. The CPU’s multi-thread score of 41,737 and Cinebench R23 multi-core of 36,064 indicate it can handle physics, AI, and simulation tasks in games without breaking a sweat. Conversely, the GPU’s FP32 throughput of 15.62 TFLOPS and 8 GB of GDDR6 memory with 256.0 GB/s bandwidth will constrain frame rates in modern titles at high settings.

The memory architecture further shapes this balance. Both the CPU and GPU share the same 256.0 GB/s bandwidth figure — the CPU via quad-channel LPDDR5X and the GPU via its 128-bit GDDR6 bus. This symmetry means neither component has an inherent memory bandwidth advantage, but the GPU’s 8 GB capacity is a potential bottleneck for 1440p or 4K textures. The CPU’s 64 MB L3 cache helps offset memory latency for the processor, but the GPU cannot cache its way out of a VRAM shortage.

In FPS terms (estimated, not measured), expect the GPU to be the limiting factor at 1440p and above in graphically demanding titles. At 1080p with lower settings, the CPU’s single-core strength (5,091 in R23 single-core) could allow higher frame rates, but the GPU’s 73rd percentile standing will cap maximum FPS well below what the CPU could theoretically support. The data suggests a system where CPU-bound tasks (compilation, spreadsheet crunching) far outpace GPU-bound tasks (gaming at high detail, rendering) in relative performance.

Usage Scenarios

High-refresh gaming: The GPU’s 73rd percentile and 15.62 TFLOPS FP32 suggest 1080p high-refresh gaming is feasible, with estimated frame rates in esports titles exceeding 100 FPS at medium settings. However, the 8 GB VRAM and 256.0 GB/s bandwidth will limit 1440p ultra settings, where the CPU’s 91st percentile is wasted.

Streaming: The CPU’s 24 threads and strong multi-core scores (41,737 PassMark multi-thread) handle encoding and game logic simultaneously. The GPU’s 144 tensor cores and 36 RT cores support DLSS and ray tracing, but the 8 GB VRAM may constrain simultaneous game rendering and stream encoding at high bitrates.

Video editing: Cinebench R23 multi-core of 36,064 and PassMark data compression of 487,145 indicate fast export times and responsive timeline scrubbing. The GPU’s 109,197 OpenCL score accelerates effects and color grading, while the 256.0 GB/s shared bandwidth moves frames quickly.

3D rendering: The CPU’s 90,594 floating-point math score handles physics simulations and scene logic, while the GPU’s 8,399 PassMark compute score provides moderate rendering acceleration. Expect slower GPU renders than dedicated desktop workstation cards, but the CPU’s 12 cores keep viewport interaction fluid.

Software development: The CPU’s 146,519 integer math score and 53,113 random string sorting score make compilation and code analysis fast. The 64 MB L3 cache reduces context-switching overhead, and 24 threads allow parallel builds. The GPU is irrelevant here, but the CPU’s 91st percentile justifies this pairing for developers.

Student and office work: The CPU’s 4,028 single-thread score ensures snappy application launches, while the 55 W TDP keeps the laptop cool and quiet. The GPU is overkill for documents and browsing, but the Radeon 8050S integrated graphics (part of the CPU) could handle light tasks while the RTX 4070 Mobile remains idle, saving power.

FAQ

Q: How does the Ryzen AI Max 390 compare to the Intel Core i9-14900HX?

A: The Ryzen AI Max 390 has an average benchmark score of 56,273, which is 0.5% higher than the Core i9-14900HX’s 56,004. Both are top-tier mobile CPUs, with the AMD part holding a slight edge in aggregate performance.

Q: Is the RTX 4070 Mobile faster than a desktop RTX 3090?

A: In aggregate benchmarks, the RTX 4070 Mobile scores 27,435, while the RTX 3090 scores 27,565 — a 0.5% difference. They are statistically equivalent in synthetic tests, though the RTX 3090’s larger VRAM (not specified here) would likely win in memory-heavy tasks.

Q: What is the CPU’s single-thread performance?

A: The Ryzen AI Max 390 scores 5,091 in Cinebench R23 single-core and 4,028 in PassMark single-thread, placing it well above average for mobile CPUs and competitive with desktop parts.

Q: Does the GPU support ray tracing?

A: Yes, the RTX 4070 Mobile has 36 RT cores and 144 tensor cores, supporting DirectX 12 Ultimate (12_2), which includes ray tracing and DLSS features.

Q: What is the memory bandwidth available to the CPU and GPU?

A: Both the CPU’s quad-channel LPDDR5X and the GPU’s 128-bit GDDR6 provide 256.0 GB/s of bandwidth, creating a balanced memory subsystem.

Q: How does the CPU compare to a workstation processor?

A: The Ryzen AI Max 390 trails the AMD Ryzen Threadripper PRO 3955WX by only 0.5% in aggregate score (56,273 vs. 56,555), making it nearly as fast as a desktop workstation chip in synthetic tests.

Q: Is this a good pairing for 4K gaming?

A: The GPU’s 8 GB VRAM and 73rd percentile standing suggest 4K gaming would require reduced settings, as the CPU’s 91st percentile cannot compensate for GPU memory and throughput limits at that resolution.

GPU Analysis

The NVIDIA GeForce RTX 4070 Mobile is built on the Ada Lovelace architecture (AD106 chip) using TSMC’s 5 nm process, with 22,900 million transistors on a 188 mm² die. It features 4,608 shading units, 144 texture mapping units, and 48 raster operation processors. Clock speeds are 1,395 MHz base and 1,695 MHz boost, with memory running at 2,000 MHz (16 Gbps effective). The 8 GB GDDR6 memory sits on a 128-bit bus, delivering 256.0 GB/s bandwidth — identical to the CPU’s system memory bandwidth, which is notable for a unified memory architecture.

The GPU’s 36 RT cores and 144 tensor cores enable hardware-accelerated ray tracing and AI-based features like DLSS. Its FP32 throughput is 15.62 TFLOPS, with FP16 at the same rate (1:1), indicating strong compute for a mobile part. Pixel rate is 81.36 GPixel/s, and texture rate is 244.1 GTexel/s, both respectable for a 115 W TDP mobile GPU. The bus interface is PCIe 4.0 x8, and the slot width is listed as IGP (integrated graphics package), meaning it is soldered to the motherboard rather than a replaceable MXM module.

Benchmark results place this GPU at the 73rd percentile, with an average score of 27,435. Its nearest rivals include the AMD Radeon RX 6700 XT (27,425, 0% delta) and the NVIDIA GeForce RTX 3090 (27,565, 0.5% higher). This means the mobile 4070 performs like a mid-range desktop card from a previous generation, which is impressive for a laptop. However, the 8 GB VRAM is a limiting factor for modern titles at high resolutions or with heavy texture packs. The GPU’s Geekbench OpenCL score of 109,197 and Vulkan score of 108,367 show strong cross-API compute, while PassMark G3D of 19,587 confirms solid rasterization performance. The DirectX 12 score of 85 is lower than DirectX 11’s 179, which may indicate driver overhead or architectural differences in how the tests exercise the hardware.

For rendering workloads, the GPU’s compute scores (8,399 PassMark GPU Compute) suggest it can accelerate Blender cycles or Adobe Premiere exports, but it is not a dedicated workstation card. The 36 RT cores provide real-time ray tracing in supported games, though the 8 GB VRAM may force lower quality settings when RT is enabled. Overall, this is a capable mobile GPU that punches above its class in synthetic tests but has memory capacity constraints.

Upgrade Path and Platform

The CPU uses AMD Socket FP11, which is a mobile-specific socket, meaning the processor is not user-upgradable in most laptops. The platform supports LPDDR5X memory in quad-channel configuration, with ECC memory support — a feature typically found in workstation systems. The CPU provides 16 PCIe Gen 4 lanes, while the GPU uses PCIe 4.0 x8, leaving headroom for additional NVMe storage or other peripherals.

The CPU’s TDP is 55 W, and the GPU’s TDP is 115 W, totaling 170 W for the two main components. This suggests a laptop power adapter in the 200 W range would be necessary (though not specified in the data), and the system likely requires robust cooling given the combined thermal load. The GPU has no power connectors (listed as “None”), confirming it is soldered and draws power through the motherboard.

For a sensible next upgrade, the data points to memory capacity rather than component swaps. The 8 GB GPU VRAM is the most likely bottleneck in gaming and rendering scenarios. Since the GPU is not replaceable (IGP slot width), the only memory upgrade path is system RAM, which could alleviate some pressure by allowing larger page files or shared memory allocations, though this does not increase GPU VRAM. The CPU’s 64 MB L3 cache and 256.0 GB/s bandwidth are already generous, so further CPU-side upgrades are unlikely to yield meaningful gains. The platform’s PCIe Gen 4 lanes support fast NVMe SSDs, which would improve load times and data-heavy workflows.

Build Overview

This is a laptop-class build (buildClass: “laptop”) pairing AMD’s top-tier Ryzen AI Max 390 processor with NVIDIA’s mid-high-range GeForce RTX 4070 Mobile GPU. The CPU is the standout component, sitting at the 91st percentile among all CPUs, while the GPU at the 73rd percentile is solid but not exceptional. The combined percentile of 82 reflects a system that excels in CPU-bound tasks and is competent but not class-leading in GPU-bound ones.

The CPU’s aggregate benchmark score of 56,273 places it in the same league as the Intel Core i9-14900HX (56,004) and AMD Ryzen Threadripper PRO 3955WX (56,555), making this a workstation-class processor in a mobile form factor. The GPU’s 27,435 average score matches the desktop Radeon RX 6700 XT (27,425) and RTX 3090 (27,565), delivering desktop-equivalent graphics performance in a laptop.

This pairing is unusual for a laptop — most mobile systems pair a lower-tier CPU with a higher-tier GPU. Here, the balance is inverted, prioritizing multi-threaded productivity over raw gaming frame rates. The 55 W CPU TDP allows sustained all-core workloads, while the 115 W GPU TDP is moderate for a mobile 40-series card. The system is best described as a portable workstation with gaming capabilities, not a dedicated gaming laptop.

Who Should Build It

The primary audience is professionals who need serious multi-threaded CPU performance in a portable package. The 91st percentile CPU handles software compilation, video encoding, and scientific simulations with ease, as evidenced by Cinebench R23 multi-core of 36,064 and PassMark multi-thread of 41,737. Developers will appreciate the 146,519 integer math score and 53,113 random string sorting score, which speed up code builds and data processing.

Content creators targeting 1080p or 1440p video editing will find the CPU’s 487,145 data compression score and the GPU’s 109,197 OpenCL score sufficient for smooth timelines and reasonable export times. Students in engineering or data science programs will benefit from the CPU’s 24 threads for MATLAB or Python workloads, while the GPU’s 8,399 compute score handles basic CUDA tasks.

Gamers at 1080p with mid-to-high settings will be satisfied, but those seeking 1440p ultra or 4K should look elsewhere — the 8 GB VRAM and 73rd percentile GPU will bottleneck. Small business workstations handling databases, accounting software, or virtualization will run smoothly given the CPU’s 91st percentile and ECC memory support. This is a build for users who prioritize processing power over raw graphics, with gaming as a secondary use case.

Gaming Performance

No measured FPS rows exist for this exact CPU+GPU combination in the data, so all frame rate figures below are estimates derived from the benchmark scores. The GPU’s 73rd percentile and PassMark G3D score of 19,587 suggest 1080p gaming at high settings is achievable in most titles, with esports games likely exceeding 100 FPS given the CPU’s strong single-thread score of 5,091 in Cinebench R23. At 1440p, expect 60-90 FPS in optimized titles, but the 8 GB VRAM will cause stuttering in games with large texture packs.

The CPU’s 91st percentile ensures it will not bottleneck the GPU at 1080p, but at 1440p and above, the GPU becomes the limiting factor. The GPU’s 36 RT cores enable ray tracing, but enabling it will drop frame rates significantly — the 15.62 TFLOPS FP32 is modest for RT workloads. DLSS via the 144 tensor cores could recover some performance, but the 8 GB VRAM limits the quality preset that can be used with DLSS enabled.

In GPU-bound titles, the RTX 4070 Mobile’s performance relative to the desktop RTX 3090 (0.5% slower in aggregate) suggests it can handle most games at 1080p ultra or 1440p high. However, the mobile GPU’s 115 W TDP means sustained clocks may drop under load, unlike a desktop card. For competitive gaming at 1080p with low settings, the CPU’s 91st percentile and the GPU’s 73rd percentile combine to deliver very high frame rates, likely exceeding 200 FPS in lighter titles. For AAA single-player games at 1440p, estimate 60-80 FPS at high settings, with drops to 40-50 FPS if ray tracing is enabled. At 4K, the 8 GB VRAM will force medium or low settings, and frame rates will hover around 30-50 FPS depending on the title.