SYSTEM ANALYZER

Rate My PC: Intel Core Ultra 9 285T + NVIDIA GeForce RTX 5090 SE

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

84 / 100
HIGH-END

Power Build

Top 16% of systems. Excellent for 1440p Ultra or 4K High gaming.

1440p Ultra4K High

System Balance Analysis

CPU vs GPU performance ratio
Well Balanced
CPU
94%
VS
GPU
74%
PROCESSOR

Intel Core Ultra 9 285T

51,310 Benchmark Score
Top 6% Market Ranking
View Full Specs →
GRAPHICS CARD

NVIDIA GeForce RTX 5090 SE

0 Benchmark Score
Top 26% 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

Strong Performance

Excellent for 1440p gaming. Most games will run at high/ultra settings smoothly.

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

This pairing combines Intel’s most power-efficient desktop flagship, the Core Ultra 9 285T, with NVIDIA’s top-tier GeForce RTX 5090 SE. The CPU is a 24-thread Arrow Lake-S part with a 35 W TDP, while the GPU is a 500 W Blackwell 2.0 behemoth with 24 GB of GDDR7 memory. Benchmark data shows the processor sits at the 91st percentile among all CPUs, but the GPU has no recorded benchmark scores, leaving its percentile at 50 with an average score of zero. The combined system percentile is 71, and there is no measured FPS data for this exact configuration. All gaming performance figures discussed below are estimates derived from the CPU’s and GPU’s architectural specifications and the CPU’s benchmark scores, not from direct testing of this pairing.

CPU Analysis

The Intel Core Ultra 9 285T is built on the Arrow Lake architecture, manufactured on TSMC’s 3 nm process with 17,800 million transistors on a 243 mm² die. It features 24 cores and 24 threads, meaning there is no hyper-threading—each core is a physical core. The base clock is 1.40 GHz, which is low, but the boost clock reaches 5.40 GHz, indicating significant headroom for burst workloads. This design prioritizes efficiency at idle and low-load states, given the 35 W TDP, while still providing high single-thread performance when needed.

The cache hierarchy includes 192 KB of L1 per core, 3 MB of L2 per core, and 36 MB of shared L3 cache. This is a substantial amount of L2, which can help with latency-sensitive workloads. The memory controller supports dual-channel DDR5 with a bandwidth of 102.4 GB/s, and ECC memory is supported, making this CPU suitable for error-sensitive tasks like scientific computing or long-running server-like workloads.

Benchmark results paint a clear picture of a processor that trades raw multi-threaded dominance for efficiency. In Cinebench R23, it scores 33,573 in multi-core and 4,739 in single-core. The multi-core score is competitive with older high-end desktop parts, but the single-core score is remarkably strong, indicating excellent per-thread performance. The PassMark multi-thread score is 39,931, which is high for a 35 W part. The data compression score of 384,140 and encryption score of 32,061 show that the CPU handles data-intensive tasks well, though the encryption result is modest compared to the compression figure.

Comparing to nearest rivals, the Core Ultra 9 285T has an average benchmark score of 51,310. It is 0.6% ahead of the Intel Core i9-14900T (51,015) and 1.1% ahead of the Intel Core i7-13850HX (50,761). It is 0.8% behind the Intel Core i9-13900F (51,730) and 1.2% ahead of the AMD Ryzen 9 5900XT (50,718). These deltas are all within a couple of percent, indicating that the 285T is not a performance outlier in either direction—it is a solid performer that achieves results comparable to established desktop CPUs while drawing far less power. The PassMark single-thread score of 4,576 reinforces that this CPU is not a slouch in lightly threaded applications.

For real workloads, the data suggests the 285T is a versatile processor. The high single-core score (4,739 in R23) means that tasks like web browsing, office applications, and light coding will feel responsive. The multi-core score of 33,573 in R23 indicates that video encoding, 3D rendering, and compilation can be handled, though not at the pace of a 125 W+ flagship. The floating-point math score of 137,923 and integer math score of 132,433 in PassMark show balanced arithmetic performance. The physics score of 2,842 is lower, which could hint at reduced sustained performance under heavy multi-threaded loads, likely due to the 35 W power limit.

Upgrade Path and Platform

The Core Ultra 9 285T uses the Intel Socket 1851, which is the platform for Arrow Lake-S desktop processors. The CPU supports PCIe Gen 5 with 20 lanes from the CPU itself, enabling fast NVMe storage and a high-bandwidth GPU connection. The integrated graphics is Arc Xe-LPG with 64 execution units, which is a capable fallback for basic display output or troubleshooting without a discrete GPU, but it is not intended for gaming.

Memory support is dual-channel DDR5 with ECC capability. The 102.4 GB/s bandwidth is standard for dual-channel DDR5, and for a 35 W CPU, this is sufficient for most workloads. Users looking to upgrade from an older platform will need a new motherboard with Socket 1851 and DDR5 memory; there is no compatibility with older DDR4 modules.

The CPU’s TDP is 35 W, which is exceptionally low for a 24-core processor. This means that most existing coolers, even a capable air cooler, will be more than sufficient to handle the thermal load. The power draw is so modest that the CPU is unlikely to be a bottleneck in system power consumption. The suggested PSU for the GPU is 900 W, but the CPU’s contribution to that total is minimal. For a sensible next upgrade, users on this platform should focus on the GPU and memory. The PCIe Gen 5 support means that future graphics cards and SSDs will be able to run at full bandwidth. The 20 CPU lanes allow for a single GPU at x16 and a Gen 5 NVMe drive at x4, which is a balanced configuration for a high-end build.

GPU Analysis

The NVIDIA GeForce RTX 5090 SE is built on the Blackwell 2.0 architecture, using the GB202 chip fabricated on TSMC’s 5 nm process. It contains 92,200 million transistors on a 750 mm² die, with a transistor density of 122.9 million per square millimeter. The GPU has 14,080 shading units, 440 texture mapping units, and 160 raster operation units. It also includes 110 ray tracing cores and 440 tensor cores, making it a fully featured RTX-class GPU with dedicated hardware for ray tracing and AI acceleration.

Memory consists of 24 GB of GDDR7 on a 384-bit bus, yielding a bandwidth of 1.34 TB/s. This is a massive amount of memory bandwidth, which is critical for high-resolution textures and data-heavy compute tasks. The memory clock is 1750 MHz, or 28 Gbps effective. The base clock is 1740 MHz with a boost clock of 2377 MHz. Pixel rate is 380.3 GPixel/s and texture rate is 1,045.9 GTexel/s. FP32 performance is 66.94 TFLOPS, with FP16 at the same rate (1:1), indicating that the GPU does not have a separate half-precision boost, but the raw FP32 throughput is excellent for both gaming and compute.

The TDP is 500 W, which requires a 900 W suggested PSU. The power connector is a single 16-pin, and the GPU is dual-slot with dimensions of 267 mm in length, 111 mm in height, and 40 mm in width. This is a compact design for a 500 W card, making it easier to fit in smaller cases. Display outputs include one HDMI 2.1b and three DisplayPort 2.1b, supporting high refresh rate monitors at 4K and above.

Because there are no benchmark scores for the RTX 5090 SE, the data cannot quantify its performance relative to other GPUs. However, the raw specifications suggest a very high-end part. The 66.94 TFLOPS of FP32 performance places it in the field of professional-grade compute cards, and the 1.34 TB/s memory bandwidth is more than sufficient for 4K gaming with ultra textures. The 110 RT cores indicate strong ray tracing capabilities, and the 440 tensor cores are designed for DLSS and AI workloads. For rendering, the combination of high FP32 throughput and large memory capacity means that complex scenes with heavy geometry and high-res textures will be handled with ease, though the lack of measured data means exact frame rates cannot be stated.

Usage Scenarios

High-refresh gaming: The CPU’s single-thread score of 4,739 in Cinebench R23 is strong, and the GPU’s 66.94 TFLOPS FP32 suggests high frame rates at 1080p and 1440p. However, the lack of measured FPS data means these are estimates. The CPU’s 24 cores will not bottleneck the GPU in most titles, but the 35 W TDP could limit sustained boost clocks in CPU-heavy games.

Streaming: The CPU has 24 threads and a PassMark multi-thread score of 39,931, which is sufficient for software encoding at moderate bitrates. The GPU’s NVENC encoder (implied by the 440 tensor cores and Blackwell architecture) can offload encoding, leaving CPU resources for the game. The data compression score of 384,140 indicates the CPU can handle stream overlays and chat processing.

Video editing: The CPU’s 33,573 score in Cinebench R23 multi-core suggests it can handle 4K timeline scrubbing and effects, though export times will be longer than a higher-TDP part. The GPU’s 24 GB VRAM and 1.34 TB/s bandwidth are excellent for GPU-accelerated effects and color grading. The FP32 performance of 66.94 TFLOPS accelerates render previews.

3D rendering: The CPU’s multi-thread score of 33,573 in R23 is adequate for CPU-based rendering in Blender or similar, but the GPU is the workhorse. The RTX 5090 SE’s 66.94 TFLOPS FP32 and 110 RT cores are well-suited for both raster and ray-traced rendering. The 24 GB VRAM allows loading large scenes without out-of-memory errors.

Software development: The CPU’s single-thread score of 4,576 in PassMark and 4,739 in R23 ensures fast compilation of individual files. The 24 threads handle parallel builds, with a PassMark integer math score of 132,433. The ECC memory support is a plus for long-running test suites. The GPU is overkill for development but can accelerate machine learning training with its 440 tensor cores.

Student and office work: The CPU’s low 35 W TDP means quiet and cool operation. The single-thread performance is excellent for spreadsheets and document processing. The integrated Arc graphics is sufficient for dual-monitor setups. The PassMark single-thread score of 4,576 ensures snappy application launches. This is a high-end workstation for students in engineering or data science, where the GPU can be used for compute projects.

FAQ

Q: What is the CPU’s TDP, and what does it mean for cooling?

A: The TDP is 35 W, which is very low for a 24-core processor. This means standard air coolers are more than adequate, and the system will run quiet.

Q: Does the CPU support ECC memory?

A: Yes, the Core Ultra 9 285T supports ECC memory, which is beneficial for error-sensitive workloads like scientific computing or long-running servers.

Q: What is the GPU’s memory bandwidth?

A: The RTX 5090 SE has a memory bandwidth of 1.34 TB/s, provided by 24 GB of GDDR7 on a 384-bit bus.

Q: How does the CPU compare to the Intel Core i9-14900T?

A: The Core Ultra 9 285T has an average benchmark score of 51,310, which is 0.6% higher than the i9-14900T’s 51,015.

Q: Is the GPU suitable for 4K gaming?

A: Based on its 66.94 TFLOPS FP32 performance and 24 GB VRAM, it is designed for high-resolution gaming, but no measured FPS data exists for this pairing.

Q: What is the suggested PSU wattage for this GPU?

A: The suggested PSU is 900 W, given the GPU’s 500 W TDP.

Q: Does the CPU have integrated graphics?

A: Yes, it has Arc Xe-LPG Graphics with 64 execution units, which is suitable for basic display and troubleshooting.

Who Should Build It

This pairing targets users who demand extreme graphics performance without sacrificing CPU efficiency. Gamers at 4K resolution will benefit from the GPU’s 24 GB VRAM and 1.34 TB/s bandwidth, which handle ultra textures and high-resolution assets. The CPU’s single-thread score of 4,739 in R23 ensures that even at high frame rates, the processor will not be the limiting factor in most titles. Content creators working with 8K video or complex 3D scenes will leverage the GPU’s 66.94 TFLOPS FP32 and 110 RT cores for accelerated rendering, while the CPU’s 24 threads handle background tasks.

Developers building AI models will find the 440 tensor cores useful for training and inference, and the CPU’s ECC memory support adds reliability. Students in engineering or data science fields get a dual-purpose machine: the CPU is efficient for daily work, and the GPU is a powerful compute accelerator for projects. Small business workstations that need to handle CAD or video editing will appreciate the quiet operation from the 35 W CPU and the GPU’s raw power, though the 500 W GPU TDP means the overall system is not energy-efficient. The 24 threads and 36 MB of L3 cache make this a capable multitasking machine for professionals who run multiple virtual machines or containers.

Balance and Bottleneck

The most striking imbalance is the power disparity: the CPU has a 35 W TDP, while the GPU has a 500 W TDP. This means the GPU is the dominant power consumer and likely the performance driver in graphics-intensive tasks. In gaming at lower resolutions like 1080p, the CPU’s 24 threads and high single-core score (4,739 in R23) may be sufficient to keep up, but the 35 W power limit could cause frequency drops in sustained loads, potentially bottlenecking the GPU in CPU-bound scenarios.

The CPU’s PassMark physics score of 2,842 is relatively low, which suggests that it may not sustain peak multi-threaded performance for extended periods. This could create a bottleneck in workloads that are both CPU-intensive and GPU-intensive, such as streaming while gaming. The GPU, with no benchmark scores, has an unknown percentile, but its raw specs indicate it is far more powerful than the CPU in absolute terms. The combined percentile of 71 reflects this imbalance: the CPU’s 91st percentile is dragged down by the GPU’s 50th percentile, which is an artifact of missing data rather than poor performance.

For FPS scaling, the data suggests that the GPU will be the limiting factor at 4K, where its 66.94 TFLOPS and 1.34 TB/s bandwidth are the primary drivers. At 1080p, the CPU’s single-thread performance becomes more relevant, and the 35 W TDP could cap frame rates in esports titles. The 24 GB VRAM means that memory capacity will rarely be a bottleneck, even with high-resolution texture packs.

Build Overview

This is a desktop-class build combining the Intel Core Ultra 9 285T with the NVIDIA GeForce RTX 5090 SE. The CPU is a 24-core, 24-thread Arrow Lake-S processor with a 35 W TDP, designed for efficiency. The GPU is a Blackwell 2.0 flagship with 24 GB GDDR7 and a 500 W TDP, designed for maximum performance. The combined percentile is 71, which places this system above the majority of all CPU+GPU pairings in the database. However, the GPU’s lack of benchmark data means this percentile is provisional. The CPU’s 91st percentile among all CPUs indicates it is a top-tier processor, while the GPU’s 50th percentile is not meaningful due to zero recorded scores. In terms of class, this is a high-end desktop workstation or gaming rig, with the CPU providing efficient multi-threading and the GPU providing uncompromised graphics performance.

Benchmark Performance

The CPU’s average benchmark score is 51,310, placing it at the 91st percentile among all CPUs. Its nearest rival, the Intel Core i9-14900T, scores 51,015, which is 0.6% lower. The Intel Core i9-13900F scores 51,730, which is 0.8% higher than the 285T. The Intel Core i7-13850HX scores 50,761, 1.1% lower, and the AMD Ryzen 9 5900XT scores 50,718, 1.2% lower. In Cinebench R23, the multi-core score is 33,573 and single-core is 4,739. PassMark multi-thread is 39,931 and single-thread is 4,576. The data compression test yields 384,140, and encryption yields 32,061.

The GPU has no benchmark scores, so its average score is 0, and its percentile is 50. The combined system percentile is 71, reflecting the CPU’s strong showing and the GPU’s unknown quantity. The pair rank by FPS is null, indicating no comparative data. The overall picture is a CPU that is measurably strong, paired with a GPU whose performance must be inferred from specifications. The CPU’s 91st percentile suggests it outperforms 91% of all CPUs, but the GPU’s lack of data means the combined percentile is not fully reliable.

Gaming Performance

There are no measured FPS rows for this exact CPU+GPU combination in the database. All gaming performance figures below are estimates based on the CPU’s benchmark scores and the GPU’s raw specifications, not from direct testing. The CPU’s single-core score of 4,739 in Cinebench R23 suggests that it can feed the GPU in most games, but the 35 W TDP may limit sustained boost clocks in long gaming sessions. The GPU’s 66.94 TFLOPS FP32 and 1.34 TB/s memory bandwidth indicate it is capable of high frame rates at 1440p and 4K with ultra settings.

For esports titles at 1080p, the CPU’s high single-thread performance should allow frame rates in the hundreds, but the power limit could cause occasional dips. At 1440p, the GPU’s raw performance becomes the primary driver, and the system should handle most games at ultra settings with high frame rates. At 4K, the GPU’s 24 GB VRAM and 1.34 TB/s bandwidth are well-suited for maximum detail, and frame rates will be limited by the GPU’s 500 W power envelope rather than the CPU. Ray tracing workloads will benefit from the 110 RT cores, and DLSS (implied by the 440 tensor cores) can boost performance further, but exact numbers cannot be provided.