SYSTEM ANALYZER

Rate My PC: Intel Core Ultra 5 225F + NVIDIA GeForce RTX 4090 D

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

95 / 100
ULTIMATE READY

Apex Performer

Top 5% 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
91%
VS
GPU
99%
PROCESSOR

Intel Core Ultra 5 225F

37,313 Benchmark Score
Top 9% Market Ranking
View Full Specs →
GRAPHICS CARD

NVIDIA GeForce RTX 4090 D

178,050 Benchmark Score
Top 1% 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

The Intel Core Ultra 5 225F paired with the NVIDIA GeForce RTX 4090 D is a deliberately lopsided desktop build: a mid-tier Arrow Lake processor sitting in the 85th percentile of all CPUs, matched with a flagship-tier Ada Lovelace graphics card in the 98th percentile of all GPUs. The combined system lands in the 92nd percentile overall, which tells you immediately where the performance story lives — in the graphics card, not the CPU. This page examines what that imbalance means for gaming, rendering, and real-world workloads.

FAQ

Q: What CPU and GPU are in this build?

A: The Intel Core Ultra 5 225F, a 10-core, 10-thread Arrow Lake desktop processor on Intel Socket 1851, paired with the NVIDIA GeForce RTX 4090 D, an AD102-based Ada Lovelace graphics card built on TSMC's 5 nm process.

Q: Is this a desktop or laptop configuration?

A: Desktop. Both components carry the desktop market segment — the CPU is an active Socket 1851 part, and the GPU is a triple-slot PCIe card drawing power from a single 16-pin connector.

Q: Is the GPU still in production?

A: No. The RTX 4090 D is listed as end-of-life, with the GeForce 50 series listed as its successor. The CPU, by contrast, is in active production status.

Q: How much power does this build need?

A: The RTX 4090 D carries a 425 W TDP with NVIDIA suggesting an 800 W power supply. The CPU adds a modest 65 W TDP on top, so the GPU dominates the power budget entirely.

Q: Does the CPU have integrated graphics?

A: No. The "F" suffix part ships with no integrated graphics, so the RTX 4090 D is the only display output path — the card provides one HDMI 2.1 and three DisplayPort 1.4a connections.

Q: Where do the components rank against the broader market?

A: The CPU sits in the 85th percentile versus all CPUs; the GPU sits in the 98th percentile versus all GPUs. The combined build ranks in the 92nd percentile.

Q: Can the CPU be overclocked?

A: No. The Core Ultra 5 225F ships with a locked multiplier, so the boost clock of 4.90 GHz is effectively the ceiling.

Balance and Bottleneck

This is a GPU-forward pairing, and the percentile data makes that unambiguous. An 85th-percentile CPU feeding a 98th-percentile GPU creates a configuration where the graphics card will rarely be the limiting factor at high resolutions — the CPU will be, whenever a workload depends on per-frame processor throughput.

The single-thread results explain why. A Cinebench R23 single-core score of 1893 and a PassMark single-thread score of 4397 are solid mainstream numbers, and the CPU's overall average benchmark score of 37313 places it within a fraction of a percent of rivals like the Intel Core i9-13900HK (within 0.3 percent), the AMD Ryzen 7 7735H (0.4 percent ahead of it), the Intel Core i7-13700 (0.5 percent ahead), and the AMD Ryzen 7 160 (0.5 percent ahead). Those are tightly clustered comparisons against parts that are not flagship desktop chips. Feeding a card that trades blows with the NVIDIA A100 SXM4 80 GB and the RTX PRO 5000 Blackwell means the processor is the subordinate partner.

The practical consequence: at 4K and other high resolutions, where GPU load scales steeply, this pairing will perform close to its graphics-card potential. At lower resolutions, where the CPU drives frame delivery, the 10-core, 10-thread Arrow Lake part becomes the frame-rate governor. With only ten threads and no hyper-threading-style doubling, heavily threaded background tasks — streaming, encoding, compilation — will compete directly with game threads in a way they wouldn't on a higher-tier processor.

It's worth noting there are no measured FPS rows for this exact combination in the database — the FACT PACK contains no measured per-game frame rates, and dataIsMeasured is flagged false. All frame-rate discussion below is therefore estimated from the benchmark scores rather than observed.

Benchmark Performance

The CPU's benchmark sheet covers Cinebench R15, R20, and R23, plus a broad PassMark suite. In Cinebench R23 multi-core the chip scores 16467, with a single-core score of 1893. The R20 figures are 11059 multi-core and 1561 single-core; the R15 figures are 2660 multi-core and 287 single-core. PassMark multithread lands at 31004, with single-thread at 4397. Individual PassMark subscores include 310843 in data compression, 22648 in encryption, 28027 in extended instructions, 92554 in floating point math, 66417 in integer math, 2430 in physics, 37325 in string sorting, and 352 in finding prime numbers.

The GPU's recorded results are a 3DMark Steel Nomad DX12 score of 8587, a Geekbench OpenCL score of 278621, and a Geekbench Vulkan score of 246941, contributing to an average benchmark score of 178050. That average sits within striking distance of some serious professional hardware: the RTX PRO 5000 Blackwell is only 2.2 percent ahead of it (deltaPct -2.2 from the 4090 D's perspective), the A100 SXM4 80 GB is 3.1 percent ahead, the RTX 5000 Ada Generation is 3.6 percent ahead, and the A100 SXM4 40 GB is 4.9 percent ahead. In other words, this consumer card benchmarks within roughly five percent of dedicated data-center and workstation parts.

The combined picture is a 92nd-percentile system whose GPU benchmark muscle does the heavy lifting. The compute scores — nearly 280 thousand points in OpenCL — indicate serious throughput for GPU-accelerated work, while the CPU's PassMark physics score of 2430 marks the ceiling for simulation-heavy tasks that can't be offloaded to the graphics card.

Who Should Build It

4K and high-refresh high-resolution gamers. With a GPU in the 98th percentile, this build targets players who want to push maxed-out settings at high resolutions, where the CPU's 85th-percentile standing matters far less. Esports players chasing extremely high frame rates at lower resolutions will bump into the processor's ten-thread ceiling sooner.

Content creators who lean on GPU acceleration. Video editors, 3D renderers, and anyone working in GPU-accelerated pipelines benefit most from this configuration. The OpenCL and Vulkan scores indicate the RTX 4090 D can absorb the majority of rendering and export workloads, letting the CPU act primarily as a dispatcher.

Developers running GPU compute workloads. Machine-learning experimentation, CUDA-based tooling, and local model inference are realistic use cases given the card's proximity to A100-class benchmark averages and its 456 tensor cores.

Students and small-business workstations. The 85th-percentile CPU is more than capable for general productivity, development compilation, and moderate multitasking, while the GPU provides headroom for visualization, CAD, and occasional rendering work — though buyers needing maximum CPU thread counts for heavy local compilation or batch processing may find ten threads limiting.

Builders who prioritize graphics headroom over processor tier. This pairing makes sense when the workload is GPU-bound by design. Anyone whose workload is CPU-bound — heavy code compilation, CPU rendering, simulation — would be better served rebalancing the budget between the two sockets.

GPU Analysis

The RTX 4090 D is built on the AD102 die, fabricated on TSMC's 5 nm process, packing 76,300 million transistors into a 609 mm² die — a transistor density of 125.3 million per square millimetre. It belongs to the Ada Lovelace architecture, sits in the GeForce 40 generation, and carries a launch MSRP of 1,599 USD.

The memory subsystem is the card's backbone for both gaming and rendering: 24 GB of GDDR6X on a 384-bit bus, delivering 1.01 TB/s of bandwidth with memory running at 1313 MHz (21 Gbps effective). For rendering workloads, that capacity and bandwidth mean large scenes, high-resolution textures, and substantial GPU-resident datasets can be held on-card without spilling to system memory — though the CPU's dual-channel DDR5 platform, with 102.4 GB/s of memory bandwidth, becomes a narrow pipe whenever data does have to cross it.

Compute resources are extensive: 14592 shading units, 456 texture mapping units, 176 render outputs, 114 RT cores, and 456 tensor cores. The FP32 figure is 73.54 TFLOPS, with FP16 at the same 73.54 TFLOPS on a 1:1 ratio. Pixel fill rate is 443.5 GPixel/s and texture fill rate is 1149.1 GTexel/s. The RT core and tensor core counts underpin ray-traced rendering and AI-accelerated upscaling respectively — the DirectX 12 Ultimate (12_2) API support confirms hardware ray tracing is a first-class feature, alongside Vulkan 1.4 and OpenGL 4.6.

Clocks run from a 2280 MHz base to a 2520 MHz boost. What the benchmark scores mean for rendering is straightforward: the Geekbench OpenCL result of 278621 and Vulkan result of 246941, combined with an average score within five percent of A100-class and RTX PRO-class hardware, indicate this card renders and computes at a near-workstation tier despite its consumer positioning. Note the bus interface is PCIe 4.0 x16, while the CPU offers PCIe Gen 5 with 20 lanes — the card will not use the platform's newest generation, though the interface is not a practical bottleneck at these performance levels.

Power and physical requirements are significant: a 425 W TDP, one 16-pin power connector, a triple-slot width, and dimensions of 304 mm length, 137 mm height, and 61 mm width. NVIDIA suggests an 800 W power supply for the card alone.

Gaming Performance

To restate clearly: the database contains no measured FPS data for this exact CPU and GPU combination — the measuredFpsUltraByGame field is empty and dataIsMeasured is false. The following frame-rate expectations are estimates derived from the benchmark scores, not recorded gameplay results.

The starting point is the 3DMark Steel Nomad DX12 score of 8587, a demanding DirectX 12 benchmark whose results correlate with modern AAA gaming performance. A score in this range, from a card whose average benchmark score sits within two to five percent of the RTX PRO 5000 Blackwell, the RTX 5000 Ada Generation, and both A100 SXM4 variants, places the 4090 D at the very top of consumer gaming performance.

Estimated expectations follow the resolution scaling logic established above. At 4K with ultra settings, the GPU's 24 GB frame buffer, 1.01 TB/s of bandwidth, and top-percentile compute should deliver high, fluid frame rates in current AAA titles — this is the resolution class the card was built for. At 1440p, frame rates should climb substantially, but the CPU's influence grows; the PassMark physics score of 2430 and single-thread score of 4397 define where processor-bound ceilings appear in simulation-heavy engines. At 1080p, many modern games will be almost entirely CPU-limited with this GPU installed, and the ten-thread Arrow Lake part — scoring 16467 in Cinebench R23 multi-core — becomes the frame governor.

Ray-traced gaming is a genuine use case given the 114 RT cores and DirectX 12 Ultimate support, and the tensor cores enable AI-based frame generation and upscaling in supported titles. Any frame-rate figures cited for this pairing elsewhere should be treated as estimates unless the database gains measured rows for this combination.

Upgrade Path and Platform

The CPU platform is Intel Socket 1851, supporting DDR5 memory in a dual-channel configuration with 102.4 GB/s of memory bandwidth. ECC is not supported. CPU-provided PCIe is Gen 5 with 20 lanes, so a future graphics card that fully exploits Gen 5 will find the platform ready — even though the current card operates on PCIe 4.0 x16. The CPU is a locked part: multiplierUnlocked is false, so tuning headroom on the processor side is essentially nil.

Power headroom is generous if the builder follows the GPU's 800 W suggested PSU: with the card's 425 W TDP and the CPU's 65 W TDP, a quality unit at that rating leaves substantial margin for transients and future component swaps. Physical clearance matters more than wattage here — the card's 304 mm length and triple-slot, 61 mm-wide footprint demand a spacious case.

The sensible next upgrade is the processor. Moving from the 85th-percentile Core Ultra 5 to a higher-tier Socket 1851 chip would directly address the build's principal imbalance — the CPU bottleneck at lower resolutions and in threaded workloads — while retaining the motherboard and memory. Swapping the GPU is the harder path: the RTX 4090 D is end-of-life with the GeForce 50 series as successor, and any replacement that meaningfully exceeds its 98th-percentile standing would be a step into the next generation. Builders should also note that the "F" CPU has no integrated graphics, so the system cannot fall back to iGPU output during a graphics card swap or failure — a spare card or careful planning is required.

Build Overview

This is a desktop-class build — both components are desktop parts, and the combined system ranks in the 92nd percentile. The Intel Core Ultra 5 225F is an Arrow Lake-S processor manufactured by TSMC on a 3 nm process, with 17,800 million transistors and a 243 mm² die, released in early January 2025 with a launch MSRP of $231. It carries 10 cores and 10 threads, a 3.30 GHz base clock and 4.90 GHz boost, and a cache hierarchy of 192 KB L1 per core, 3 MB L2 per core, and 20 MB of shared L3.

The NVIDIA GeForce RTX 4090 D, released in late December 2023, is a flagship-class Ada Lovelace card whose production status has since moved to end-of-life. Its average benchmark score of 178050 and 98th-percentile ranking frame it as near the top of what the database records.

As a pairing, this is a graphics-first system: a competent, efficient mainstream processor — its 65 W TDP is modest against the GPU's 425 W — paired with a card that benchmarks alongside professional hardware. The tier is high-end overall, with the caveat that the CPU half of the equation defines the build's ceiling for anything that isn't GPU-bound.

CPU Analysis

The Core Ultra 5 225F is an Arrow Lake-S part from Intel's Core Ultra Series 2, fabricated on TSMC's 3 nm node. It has 17,800 million transistors on a 243 mm² die, runs at a 3.30 GHz base and 4.90 GHz boost, and carries a 65 W TDP. The ten cores each get 192 KB of L1 and 3 MB of L2, with 20 MB of shared L3 behind them; with ten threads across ten cores, there is no SMT-style doubling. The memory controller supports DDR5 in dual-channel at 102.4 GB/s, and the platform exposes PCIe Gen 5 with 20 lanes from the CPU.

What the benchmark scores mean in practice: the Cinebench R23 multi-core score of 16467 — and the R20 score of 11059 and R15 score of 2660 — describes a solid mainstream multi-threaded performer, not a workstation chip. Its nearest rivals by average score are the Core i9-13900HK (37425, just 0.3 percent ahead), the Ryzen 7 7735H (37161, 0.4 percent behind it), the Core i7-13700 (37135, 0.5 percent behind), and the Ryzen 7 160 (37117, 0.5 percent behind). Landing in the 85th percentile against all CPUs means it outruns the vast majority of the market while sitting well below halo silicon.

Single-thread strength is the more relevant characteristic in this pairing. The Cinebench R23 single-core score of 1893, R20 single-core of 1561, R15 single-core of 287, and PassMark single-thread of 4397 describe a processor with responsive per-core throughput — which matters most when gaming, since the ten threads must carry game logic, drivers, and background tasks simultaneously.

The PassMark suite rounds out the picture for real workloads: 310843 in data compression and 22648 in encryption speak to strong throughput for file archiving and cryptographic tasks; 92554 in floating point and 66417 in integer math indicate capable general computation; 37325 in string sorting and 28027 in extended instructions cover text processing and vectorized code paths; the 352 prime-finding result and 2430 physics score sit at the performance-focused end of the suite. For developers, the build compiles adequately; for CPU rendering farms, it is outclassed by its own GPU. In this system, the processor's job is to stay out of the RTX 4090 D's way — and at high resolutions, with the workload quantified by the 98th-percentile graphics scores above it, that is a job the 85th-percentile Arrow Lake chip largely accomplishes.