SYSTEM ANALYZER

Rate My PC: Intel Core Ultra 9 290K Plus + NVIDIA GeForce RTX 5090 D V2

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
98%
VS
GPU
90%
PROCESSOR

Intel Core Ultra 9 290K Plus

84,003 Benchmark Score
Top 2% Market Ranking
View Full Specs →
GRAPHICS CARD

NVIDIA GeForce RTX 5090 D V2

16,504 Benchmark Score
Top 10% 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

# CPU Analysis

The Intel Core Ultra 9 290K Plus is a 24-core, 24-thread desktop processor built on the Arrow Lake Refresh architecture. It represents the top of Intel's Core Ultra Series 2 lineup, fabricated on TSMC's 3 nm process node with 17,800 million transistors packed into a 243 mm² die. The chip runs at a 3.70 GHz base clock and boosts up to 5.80 GHz, with a 125 W TDP that positions it as a high-end desktop part rather than an extreme workstation chip.

The core configuration is notable for what it lacks: 24 threads on 24 cores means no hyperthreading. This is a deliberate design choice that prioritizes raw core count and efficiency over simultaneous multithreading. The cache hierarchy is substantial — each core gets 192 KB of L1 and 3 MB of L2, with 36 MB of shared L3 cache. This arrangement favors workloads that can scale across many physical cores without relying on thread-level parallelism from SMT. For heavily threaded rendering, compilation, or scientific computing, the 24 physical cores provide strong scaling potential.

The integrated Arc Xe-LPG Graphics with 64 execution units means the chip can handle basic display output and light GPU-accelerated tasks without a discrete card, though in this build it is paired with a flagship discrete GPU anyway. The memory controller supports DDR5 in dual-channel configuration, delivering 115.2 GB/s of bandwidth, and ECC memory is supported — a feature that matters for workstation reliability. The CPU provides 20 PCIe Gen 5 lanes directly, which is sufficient for a single high-end GPU plus one or two Gen 5 NVMe drives.

Benchmark results confirm the CPU's position near the top of the desktop heap. In Cinebench R23, the chip scores 51,731 points in multi-core and 7,303 in single-core. The multi-core figure is the headline: it places the 290K Plus in the 96th percentile of all CPUs in the database, meaning it outperforms roughly 96% of all processors tracked. The single-core score of 7,303 is equally impressive for lightly threaded tasks like web browsing, office productivity, or legacy applications that only use one or two threads. The Cinebench R20 results — 21,727 multi-core and 3,067 single-core — follow the same pattern, and the older R15 test shows 5,214 multi-core and 736 single-core.

PassMark results paint a more detailed picture of real-world compute. The multi-thread score of 60,860 and single-thread score of 4,823 both sit at elite levels. Data compression scores 698,346, which is exceptional for archiving and file-server workloads. Floating-point math hits 214,760, indicating strong scientific and engineering compute capability. Integer math scores 166,194, and extended instructions (SIMD workloads) reach 52,338. Data encryption at 52,563 shows the chip handles AES and other cryptographic tasks efficiently. The physics score of 3,315 is modest relative to the CPU's other results, suggesting that legacy physics simulation workloads that rely on specific instruction paths may not scale as well as modern multi-threaded engines. Random string sorting at 79,744 and prime number finding at 503 round out the picture — the latter is low because it is a single-threaded, latency-bound test.

The nearest rivals in the database show just how tight the competition is at this tier. The Intel Core Ultra 9 285K scores 83,807 average versus 84,003 for the 290K Plus, a 0.2% delta in favor of the newer chip. The AMD EPYC 4584PX and EPYC 9135 trail by 1.1% and 1.2% respectively, while the EPYC 7F72 actually leads by 1.3%. These are all server-class or flagship desktop parts, and the margins are within noise for most workloads. The practical takeaway: the 290K Plus is not dramatically faster than its direct predecessor, but it holds its own against enterprise silicon in multi-threaded tasks.

# Benchmark Performance

The CPU's average benchmark score is 84,003, placing it in the 96th percentile of all CPUs. The GPU, an NVIDIA GeForce RTX 5090 D V2, scores 16,504 in 3DMark Steel Nomad DX12, which puts it in the 59th percentile of all GPUs — a solid mid-to-upper tier result, though not the absolute top. The combined build percentile is 78, indicating that the pairing of this CPU and GPU is better than 78% of all recorded system configurations.

The GPU's nearest rivals are revealing. The NVIDIA T400 scores 16,508, just 0.0% higher, and the AMD Radeon PRO W7500 scores 16,415, a 0.5% deficit. The NVIDIA RTX PRO 6000 Blackwell scores 16,408 (0.6% behind), and the AMD Radeon RX 5700 XT scores 16,361 (0.9% behind). This clustering is suspicious — the RTX 5090 D V2 is a 24 GB GDDR7 monster with 104.8 TFLOPS of FP32 compute, yet it benchmarks nearly identically to a workstation card and a last-gen midrange GPU. The data suggests that the 3DMark Steel Nomad test may not be stressing the GPU's full capabilities, or there is a driver or thermal limitation in the recorded result. Regardless, the percentile ranking of 59 reflects the measured score, not the hardware's theoretical potential.

The benchmark picture is complicated by the absence of measured FPS data. No measuredFpsUltraByGame entries exist for this exact CPU+GPU combination. All gaming performance discussion below is therefore estimated from the CPU and GPU benchmark scores, using the relative strengths of each component to infer likely frame rates. The CPU's 96th percentile and the GPU's 59th percentile suggest that the processor will not be the limiting factor in most gaming scenarios; the GPU's measured score is more modest, but its raw specs — 24 GB of VRAM, 1.34 TB/s bandwidth, and 170 RT cores — imply it has headroom that the 3DMark result does not capture.

# Upgrade Path and Platform

The Intel Core Ultra 9 290K Plus uses Socket 1851, Intel's current desktop platform for Arrow Lake and Arrow Lake Refresh processors. The chip supports DDR5 memory in dual-channel configuration, so any upgrade should retain DDR5 DIMMs. ECC memory support is present, which is a differentiator for users who need error-correcting memory in a desktop platform. The CPU provides 20 PCIe Gen 5 lanes, which is enough for one x16 GPU at full bandwidth plus a Gen 5 NVMe drive. The integrated Arc Xe-LPG Graphics with 64 EUs provides a fallback display output if the discrete GPU is removed or fails.

The GPU requires a 950 W suggested PSU and draws 575 W under load, so the power supply is a critical component in this build. The CPU's 125 W TDP is modest for a 24-core part, meaning the total system draw is dominated by the GPU. When upgrading, the PSU must be sized for the GPU's transient spikes; the 950 W recommendation should be treated as a floor, not a suggestion. The GPU uses a single 16-pin power connector, so the PSU must have that connector or an adapter.

For a sensible next upgrade, the CPU is already near the top of its platform. Socket 1851 will likely not see a new generation after Arrow Lake Refresh, so a CPU upgrade would require a new motherboard. The more practical upgrade path is the GPU: if the RTX 5090 D V2's 59th percentile result is a driver or test artifact, a future driver update could unlock more performance. Alternatively, a different GPU in the same or lower power class could be swapped in without changing the platform. The 24 GB of VRAM on the current GPU is already generous for 4K gaming and AI workloads, so a GPU upgrade would only make sense if compute requirements grow beyond what Blackwell 2.0 provides.

Memory bandwidth is 115.2 GB/s, which is standard for dual-channel DDR5 at typical speeds. For memory-intensive workloads like large dataset processing or virtual machines, this could be a bottleneck, but for gaming and most productivity tasks it is sufficient. The 36 MB of L3 cache helps mitigate memory latency in many scenarios.

# Who Should Build It

This build targets users who need maximum CPU throughput and are willing to accept a GPU that benchmarks in the mid-tier range. The CPU's 96th percentile ranking makes it ideal for content creators who render video, compile code, or run complex simulations. The 24 physical cores and 51,731 Cinebench R23 multi-core score will chew through multi-threaded rendering jobs — think Blender, Premiere Pro exports, or software compilation — faster than nearly any other desktop CPU. The 698,346 data compression score also suits users who work with large archives or backup systems.

Developers will benefit from the single-thread score of 4,823 in PassMark and 7,303 in Cinebench R23, which ensure fast compilation of single-threaded build steps and responsive IDE performance. The 52,563 encryption score helps with secure development workflows and VPN endpoints. For data scientists, the 214,760 floating-point math score and 166,194 integer math score provide strong compute for numerical analysis, though the GPU would need to carry heavier AI workloads.

Gamers at 4K resolution will find the CPU more than sufficient — the 96th percentile means it will not bottleneck any current GPU. However, the GPU's 59th percentile suggests that at 1080p or 1440p, frame rates may be limited by the graphics card rather than the processor. Students and small business workstations would be over-served by this CPU in most cases — a lower-tier chip would suffice for document editing and web browsing — but for a small business running local servers, VMs, or batch processing, the 24 cores provide headroom for years of use.

The build is less suitable for competitive esports gamers who prioritize maximum FPS at 1080p with low settings, because the GPU's mid-tier benchmark score becomes the limiting factor. It is better suited to high-resolution gaming with maximum detail settings, where the GPU's 24 GB VRAM and 1.34 TB/s bandwidth can shine even if raw FPS is moderate.

# FAQ

Q: How does the Intel Core Ultra 9 290K Plus compare to the Core Ultra 9 285K?

A: The 290K Plus scores 84,003 on average versus 83,807 for the 285K, a 0.2% advantage. This is within measurement noise, so the two chips are effectively tied in most workloads.

Q: Does the CPU support ECC memory?

A: Yes, the memory controller supports ECC memory, which is valuable for workstation reliability. The chip supports DDR5 in dual-channel configuration with 115.2 GB/s bandwidth.

Q: What is the GPU's VRAM capacity and memory type?

A: The NVIDIA GeForce RTX 5090 D V2 has 24 GB of GDDR7 memory on a 384-bit bus, delivering 1.34 TB/s of bandwidth.

Q: Is the CPU good for gaming?

A: The CPU is in the 96th percentile of all CPUs, so it will not bottleneck any GPU in gaming. The single-thread score of 7,303 in Cinebench R23 ensures strong performance in games that rely on one or two cores.

Q: What power supply is recommended for this build?

A: The GPU's suggested PSU is 950 W, and the GPU itself draws 575 W. The CPU has a 125 W TDP, so the PSU must be sized for the GPU's demands.

Q: What PCIe version and lane count does the CPU provide?

A: The CPU provides 20 PCIe Gen 5 lanes, which supports a full x16 GPU slot plus additional Gen 5 devices like NVMe SSDs.

Q: What is the GPU's architecture and process node?

A: The GPU is based on the Blackwell 2.0 architecture with the GB202 chip, fabricated on a 5 nm process by TSMC with 92,200 million transistors.

# Gaming Performance

No measured FPS data exists for this exact CPU+GPU combination. The FACT PACK contains no FPS rows for any game at any resolution, so all frame rate figures below are estimates derived from the CPU's and GPU's benchmark scores. Treat these as directional guidance, not guaranteed results.

The CPU's 96th percentile ranking and strong single-thread score (7,303 in Cinebench R23) suggest it can feed any GPU in gaming workloads. The GPU's 3DMark Steel Nomad score of 16,504, which places it in the 59th percentile, is the limiting factor. At 1080p with ultra settings, this build should deliver high frame rates in most titles — likely well above 60 FPS in esports and competitive games, and above 60 FPS in most AAA titles, though the exact numbers depend on the game's engine and optimization.

At 1440p ultra, the GPU's mid-tier score means frame rates will drop compared to 1080p, but the 24 GB VRAM ensures texture quality and asset streaming will not be compromised. Expect playable frame rates in the 60-100 FPS range for most modern games, with dips below 60 in the most demanding titles. At 4K ultra, the GPU's 59th percentile becomes a real constraint — the RTX 5090 D V2 is not among the top GPUs in the database, so 4K gaming at maximum settings will likely require DLSS or reduced detail settings to maintain 60 FPS. The 170 RT cores and 680 tensor cores provide hardware support for ray tracing and DLSS, which can offset raw rasterization limitations.

The CPU's 36 MB of L3 cache and high single-thread performance help with frame pacing and minimum FPS in CPU-bound scenarios, but the GPU is the bottleneck in most gaming situations. For competitive gamers who play at 1080p with low settings, the GPU's 59th percentile will limit maximum FPS to what a mid-tier card can achieve, which is still likely in the 144-240 FPS range for esports titles, but not the 300+ FPS that top-tier GPUs deliver.

# Balance and Bottleneck

The balance between CPU and GPU in this build is heavily skewed toward the processor. The CPU sits in the 96th percentile of all CPUs, while the GPU sits in the 59th percentile of all GPUs. This is an unusual pairing — most high-end builds pair a top-tier CPU with a top-tier GPU, but here the GPU is mid-tier by benchmark score despite its flagship-class specs. The combined percentile of 78 reflects this imbalance: the system is better than 78% of all builds, but the GPU drags the overall ranking down from what the CPU alone would suggest.

In CPU-bound workloads — rendering, compilation, scientific computing, data compression — the CPU is the clear driver and the GPU is irrelevant. The 24 cores and 51,731 Cinebench R23 multi-core score will dominate these tasks. In gaming, the GPU becomes the bottleneck at lower resolutions where the CPU could otherwise push higher frame rates. The 3DMark Steel Nomad score of 16,504 suggests the GPU is the limiting factor in GPU-bound scenarios, and the FPS scaling from 1080p to 4K will follow the GPU's performance curve, not the CPU's.

The GPU's 24 GB VRAM and 1.34 TB/s bandwidth are the saving grace. Even though its compute score is mid-tier, the memory subsystem can handle high-resolution textures and large datasets without stuttering. For AI inference or machine learning workloads that fit within 24 GB, the GPU's tensor cores (680) provide acceleration that the CPU cannot match. The practical bottleneck depends on the workload: CPU for multi-threaded compute, GPU for graphics rendering, and GPU memory for data-heavy tasks.

# Build Overview

This is a desktop-class build pairing the Intel Core Ultra 9 290K Plus with the NVIDIA GeForce RTX 5090 D V2. The CPU is a 24-core Arrow Lake Refresh part in the 96th percentile of all CPUs, while the GPU is a Blackwell 2.0 card in the 59th percentile. The combined percentile of 78 places this system in the upper quarter of all recorded builds, but the GPU is the weaker link.

The CPU is a top-tier desktop processor with no significant weaknesses — it leads its direct predecessor by 0.2% and trades blows with server-class EPYC chips. The GPU is a peculiar case: its hardware specs are flagship-class (24 GB GDDR7, 104.8 TFLOPS, 170 RT cores), but its benchmark score places it near workstation and midrange cards like the NVIDIA T400 and AMD Radeon RX 5700 XT. The build is best understood as a CPU-first system with a capable, but not class-leading, GPU.

# GPU Analysis

The NVIDIA GeForce RTX 5090 D V2 is built on the Blackwell 2.0 architecture with the GB202 chip, fabricated on TSMC's 5 nm process. It contains 92,200 million transistors on a 750 mm² die, giving a transistor density of 122.9 million per mm². The GPU has 21,760 shading units, 680 texture mapping units, and 176 ROPs, with 170 RT cores and 680 tensor cores. The base clock is 2017 MHz with a boost of 2407 MHz, and the memory runs at 1750 MHz, effective 28 Gbps.

The memory subsystem is the standout feature: 24 GB of GDDR7 on a 384-bit bus delivers 1.34 TB/s of bandwidth. This is a massive amount of VRAM and bandwidth, sufficient for 4K textures, large AI models, and professional rendering workloads. The pixel rate is 423.6 GPixel/s, and the texture rate is 1,636.8 GTexel/s. Compute performance is rated at 104.8 TFLOPS for both FP32 and FP16 (1:1 ratio), which is exceptional on paper.

The 3DMark Steel Nomad DX12 score of 16,504 is the only GPU benchmark in the data, and it places the card in the 59th percentile. The nearest rivals — NVIDIA T400, AMD Radeon PRO W7500, NVIDIA RTX PRO 6000 Blackwell, and AMD Radeon RX 5700 XT — all score within 0.9% of it. This is a paradoxical result for a card with 24 GB of VRAM and 104.8 TFLOPS of compute; it suggests either the 3DMark test is not representative of the card's capabilities, or there is a performance anomaly in the recorded data. For rendering workloads that leverage the RT cores and tensor cores, the card's theoretical throughput is far higher than the Steel Nomad score implies. The 680 tensor cores support DLSS and AI acceleration, and the 170 RT cores handle ray tracing. The 575 W TDP and 950 W suggested PSU are the practical costs of this hardware.