SYSTEM ANALYZER

Rate My PC: Intel Core Ultra 9 285T + NVIDIA GeForce RTX 5090 D V2

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

92 / 100
ULTIMATE READY

Apex Performer

Top 8% 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
94%
VS
GPU
90%
PROCESSOR

Intel Core Ultra 9 285T

51,310 Benchmark Score
Top 6% 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

This pairing combines Intel’s power-sipping 24-core Arrow Lake processor with NVIDIA’s flagship 24 GB Blackwell GPU, creating a desktop build that targets maximum compute throughput in a surprisingly restrained power envelope. The data shows a CPU that trades blows with last-generation high-core-count parts while sipping power, paired with a GPU whose raw specifications are enormous but whose benchmark standing is more nuanced. Because the FACT PACK contains no measured FPS rows for this exact combination, all gaming performance figures discussed here are estimates derived from the individual component benchmark scores, not direct measurements.

Usage Scenarios

High-refresh gaming: The RTX 5090 D V2’s 104.8 TFLOPS of FP32 compute and 1.34 TB/s memory bandwidth provide the theoretical muscle for extreme frame rates, but the GPU’s 59th percentile ranking among all GPUs suggests that at 1080p or 1440p, the CPU’s single-core performance (4739 in Cinebench R23 single-core) will be the limiting factor for pushing past 200 FPS in esports titles. At 4K, the GPU’s massive memory subsystem becomes the dominant factor, and the CPU’s 91st percentile standing ensures it will not bottleneck the GPU’s rendering throughput.

Streaming: The CPU’s 24 threads and 36 MB of shared L3 cache provide ample headroom for concurrent game encoding and broadcast software. The Passmark data compression score of 384140 indicates strong throughput for real-time video encoding tasks, while the GPU’s dedicated tensor cores (680) can offload AI-based encoding and enhancement workloads. The 35 W TDP of the CPU means the streaming PC runs cool and quiet, leaving thermal headroom for the GPU’s 575 W draw under sustained load.

Video editing: The combination of 24 CPU threads and 24 GB of GDDR7 VRAM creates a formidable editing workstation. Cinebench R23 multi-core score of 33573 places this CPU in the top 10% of all processors, which translates to smooth timeline scrubbing and fast export times in applications that leverage multi-threaded encoding. The GPU’s 24 GB frame buffer allows for complex multi-layer compositions and 8K timelines without spilling to system memory, and the 1.34 TB/s bandwidth moves raw footage quickly.

3D rendering: The CPU’s 14100 Cinebench R20 multi-core score and the GPU’s 16504 3DMark Steel Nomad DX12 score both contribute to rendering performance, but they serve different rendering paradigms. CPU-based renderers will see strong performance from the 24 cores, while GPU-based renderers (V-Ray GPU, Octane, Blender Cycles) will leverage the 21760 shading units and 170 RT cores. The 104.8 TFLOPS FP32 throughput is the key metric for unbiased GPU rendering, and the 24 GB VRAM accommodates large scene geometry and high-resolution textures.

Software development: The Passmark integer math score of 132433 and extended instructions score of 27477 indicate strong performance for code compilation, static analysis, and test execution. The 24 threads handle parallel build systems efficiently, while the 5.40 GHz boost clock keeps single-threaded tooling responsive. The 35 W TDP means developers can run this system for extended periods without high fan noise, and the ECC memory support (a rarity for desktop CPUs) provides data integrity for long-running test suites.

Student and office work: This configuration is massively over-provisioned for typical productivity workloads. The Passmark single-thread score of 4576 ensures snappy response in document editing and web browsing, but the cost and power requirements of the GPU make this a poor fit for a standard office environment. The CPU alone, with its 35 W TDP and integrated Arc Xe-LPG Graphics 64EU, would handle office tasks with negligible power draw, but the GPU’s presence fundamentally changes the system’s character.

GPU Analysis

The RTX 5090 D V2 is built on the GB202 chip, a 5 nm TSMC design containing 92,200 million transistors on a 750 mm² die. This is the largest consumer GPU die in the data, and it houses 21760 shading units, 680 texture mapping units, and 176 ROPs. The memory subsystem is a 384-bit bus carrying 24 GB of GDDR7 at 1750 MHz, yielding 1.34 TB/s of bandwidth — a figure that exceeds what most mid-range systems have for total system memory throughput.

The clock behavior is notable: a 2017 MHz base clock and 2407 MHz boost clock. The boost clock is relatively modest for a flagship, which hints that this "D V2" variant is tuned for power efficiency rather than raw clock speed. The pixel rate is 423.6 GPixel/s and the texture rate is 1,636.8 GTexel/s, both of which are class-leading figures. The FP32 throughput of 104.8 TFLOPS (with matching FP16 at a 1:1 ratio) is the headline compute specification, and the 680 tensor cores provide dedicated hardware for AI workloads.

The 3DMark Steel Nomad DX12 score of 16504 is the sole GPU benchmark in the data, and it presents a confusing picture. The GPU sits at the 59th percentile of all GPUs, with nearest rivals including the NVIDIA T400 (16508, 0% delta), AMD Radeon PRO W7500 (16415, 0.5% slower), and NVIDIA RTX PRO 6000 Blackwell (16408, 0.6% slower). This suggests that either the benchmark is not fully exercising the GPU’s capabilities, or this variant is significantly cut down from the full GB202 implementation. The 24 GB VRAM and 1.34 TB/s bandwidth are top-tier specifications, but the benchmark score places it in the same performance tier as workstation and mid-range gaming GPUs rather than at the absolute top.

For rendering workloads, the 104.8 TFLOPS FP32 throughput is the decisive metric. This is not a gaming-first GPU; it is a compute-first GPU that also happens to have display outputs (1x HDMI 2.1b, 3x DisplayPort 2.1b). The dual-slot form factor and 304 mm length are manageable for most ATX cases, but the single 16-pin power connector and 950 W suggested PSU indicate significant power delivery requirements.

Benchmark Performance

The CPU’s average benchmark score is 51310, placing it at the 91st percentile of all CPUs. Its nearest rivals are tightly clustered: the Intel Core i9-14900T (51015, 0.6% higher), Intel Core i9-13900F (51730, 0.8% lower), Intel Core i7-13850HX (50761, 1.1% higher), and AMD Ryzen 9 5900XT (50718, 1.2% higher). This means the 285T is effectively performance-equivalent to a high-end desktop chip from the previous generation, despite having a 35 W TDP versus those rivals’ higher power envelopes.

The Cinebench R23 results show a multi-core score of 33573 and single-core score of 4739. The multi-core score is strong for a 35 W part, leveraging 24 cores at modest clocks to beat many higher-TDP processors. The single-core score of 4739 is respectable but not class-leading, indicating that the 5.40 GHz boost clock is achieved only briefly and under favorable thermal conditions. The Cinebench R20 results (14100 multi-core, 1990 single-core) and R15 results (3384 multi-core, 477 single-core) follow the same pattern.

Passmark results reveal workload-specific strengths: data compression at 384140 is exceptional, data encryption at 32061 is strong, floating-point math at 137923 and integer math at 132433 are both high, and the multithread score of 39931 confirms strong parallel performance. The physics score of 2842 is moderate, and the single-thread score of 4576 aligns with the Cinebench single-core data. The find prime numbers score of 345 is notably low, suggesting this CPU’s architecture prioritizes throughput over latency in certain integer workloads.

The GPU’s 3DMark Steel Nomad score of 16504 places it at the 59th percentile, with an average benchmark score of 16504. The combined percentile for this CPU+GPU pair is 75th. The combined picture is one of a top-tier CPU paired with a GPU that benchmarks in the mid-range, which creates an imbalance for gaming but a potential advantage for compute-heavy workloads where the CPU’s 24 threads and the GPU’s 104.8 TFLOPS can work in parallel on different tasks.

Balance and Bottleneck

The data reveals a significant performance asymmetry. The CPU sits at the 91st percentile while the GPU sits at the 59th percentile. In gaming workloads, this means the CPU will rarely be the limiting factor; the GPU’s rendering throughput will cap frame rates well before the CPU’s 24 threads are saturated. At 1080p, the CPU’s strong single-core performance (4739 Cinebench R23) could push frame rates high enough that the GPU becomes the bottleneck even at lower resolutions, but the GPU’s modest benchmark ranking suggests it will not deliver the extreme frame rates its specifications imply.

In compute workloads, the balance shifts. The CPU’s 24 threads and the GPU’s 21760 shading units can be utilized simultaneously in heterogeneous workflows (e.g., physics simulation on the CPU while rendering on the GPU). The Passmark multithread score of 39931 and the GPU’s 104.8 TFLOPS FP32 throughput provide complementary compute resources. The 35 W TDP of the CPU means the total system power draw is dominated by the GPU’s 575 W TDP, so the PSU requirement (950 W suggested) is driven almost entirely by the GPU.

The FPS scaling picture, estimated from benchmark scores, shows that increasing resolution from 1080p to 4K will shift the bottleneck decisively toward the GPU. At 1080p, the CPU’s single-thread performance may limit frame rates in lightly-threaded games to around what a 59th-percentile GPU can produce. At 4K, the GPU’s 1.34 TB/s bandwidth and 24 GB VRAM become the dominant factors, and the CPU’s 91st percentile standing ensures it will not interfere with GPU-bound frame rates.

CPU Analysis

The Intel Core Ultra 9 285T is a 24-core, 24-thread processor (no hyperthreading) based on the Arrow Lake architecture, built on TSMC’s 3 nm process. The die contains 17,800 million transistors on a 243 mm² area. The base clock is 1.40 GHz with a boost clock of 5.40 GHz, and the TDP is just 35 W — a remarkable figure for a 24-core desktop part. The cache hierarchy includes 192 KB L1 per core, 3 MB L2 per core, and 36 MB of shared L3 cache.

The architecture is Arrow Lake-S, which represents Intel’s tiled design approach, with the CPU cores, GPU, and I/O on separate chiplets. The process node advantage (3 nm) explains how 24 cores can operate within a 35 W envelope. The integrated graphics are Arc Xe-LPG with 64 execution units, which provides basic display output and hardware acceleration for media tasks without requiring the discrete GPU.

Memory support is DDR5 in dual-channel configuration, with a memory bandwidth of 102.4 GB/s. ECC memory is supported, which is unusual for a consumer desktop chip and suggests workstation-oriented design. The PCIe implementation is Gen 5 with 20 lanes from the CPU, providing substantial I/O bandwidth for the GPU and NVMe storage.

The benchmark data shows a processor that punches above its power class. The Cinebench R23 multi-core score of 33573 is within 0.6% of the i9-14900T and 1.1% of the i7-13850HX, both of which have significantly higher power limits. The single-core score of 4739 trails those rivals slightly but remains competitive. The Passmark data encryption score of 32061 indicates strong AES-NI performance, and the extended instructions score of 27477 confirms robust SIMD throughput for AVX-512-class workloads.

For real workloads, this CPU excels in power-constrained environments where sustained multi-core performance matters more than peak clocks. The 35 W TDP means it can maintain high all-core performance without thermal throttling in well-ventilated cases, and it can be cooled by modest air coolers. The lack of an unlocked multiplier (multiplierUnlocked: false) means no overclocking headroom, but the architecture’s efficiency makes overclocking largely unnecessary.

Upgrade Path and Platform

The CPU uses the Intel Socket 1851 platform, which is the LGA1851 socket designed for Arrow Lake processors. The platform supports DDR5 memory in dual-channel configuration, with ECC support enabled on this CPU. The PCIe Gen 5 x16 slot for the GPU and 20 total CPU lanes provide modern I/O bandwidth, and the 35 W TDP of the CPU leaves substantial power budget for upgrades.

The GPU requires a 950 W suggested PSU, and the 575 W TDP dominates the system’s power requirements. The single 16-pin power connector is the modern standard, but it means the PSU must have native 16-pin support or an appropriate adapter. The GPU is dual-slot, 304 mm long, and 137 mm tall, so case compatibility requires checking CPU cooler height and front fan clearance.

The upgrade path for this system is limited by the platform’s maturity. Socket 1851 is the current generation, so CPU upgrades would be limited to other Arrow Lake parts with higher TDPs (though the 285T already sits near the top of the stack). The GPU is the more likely upgrade candidate, but the 24 GB VRAM and 1.34 TB/s bandwidth are already top-tier specifications. The PCIe Gen 5 x16 interface ensures the GPU slot will not be a bottleneck for future graphics cards. The memory support for DDR5 with ECC provides a path to higher-capacity modules without sacrificing data integrity for workstation use.

The PSU headroom is the most constrained aspect. With the GPU’s 575 W TDP and the CPU’s 35 W TDP, the rest of the system (motherboard, drives, fans) likely draws 50-100 W, placing the total system draw near or above 700 W under peak load. The 950 W suggested PSU provides approximately 35% headroom, which is adequate for transient GPU power spikes but does not leave room for a more power-hungry CPU upgrade.

Gaming Performance

The FACT PACK contains no measured FPS rows for this exact CPU+GPU combination, so all gaming figures presented here are estimates derived from the benchmark scores. The CPU’s 91st percentile standing and the GPU’s 59th percentile standing inform the following expectations.

At 1080p with ultra settings, the CPU’s single-core performance (4739 Cinebench R23) and the GPU’s mid-range benchmark score suggest frame rates that are CPU-limited in esports titles but GPU-limited in AAA games. The GPU’s 3DMark Steel Nomad score of 16504, which is comparable to the AMD Radeon RX 5700 XT (16361, 0.9% slower), suggests that 1080p ultra performance will be similar to that GPU — playable but not exceptional for a flagship-class card.

At 1440p, the GPU’s 24 GB VRAM and 1.34 TB/s bandwidth become more relevant, but the benchmark score still places it in the same tier as the NVIDIA T400 and Radeon PRO W7500. These are workstation-class GPUs, and their gaming performance is typically lower than gaming-focused counterparts at the same price point. Expect 1440p ultra frame rates that are playable (60+ FPS) in most titles but not competitive with top-tier gaming GPUs.

At 4K, the GPU’s memory subsystem is the saving grace. The 24 GB frame buffer prevents VRAM overflow in the most demanding titles, and the 1.34 TB/s bandwidth provides sufficient throughput for 4K textures. However, the compute throughput of 104.8 TFLOPS is not the same as rasterization performance, and the benchmark score suggests 4K ultra frame rates will be in the 40-60 FPS range for AAA titles, with lighter games reaching 80+ FPS.

The CPU’s gaming performance is solid but not exceptional. The single-core score of 4739 is sufficient for high-refresh gaming at 1080p in most titles, but the 24-thread configuration without hyperthreading means lightly-threaded games may see slightly lower performance than chips with higher single-core scores. The 36 MB L3 cache helps with gaming workloads that benefit from large shared cache.

FAQ

Q: How does the Intel Core Ultra 9 285T compare to the Intel Core i9-14900T?

A: The 285T has an average benchmark score of 51310, which is 0.6% lower than the i9-14900T’s 51015. The two are effectively performance-equivalent in average workloads, though the 285T achieves this with a 35 W TDP versus the i9-14900T’s higher power envelope.

Q: What is the GPU’s benchmark percentile, and why does it seem low for a flagship?

A: The RTX 5090 D V2 has a 3DMark Steel Nomad score of 16504, placing it at the 59th percentile of all GPUs. Its nearest rivals include the NVIDIA T400 (16508) and AMD Radeon PRO W7500 (16415), which are workstation-class GPUs. This suggests the benchmark may not fully capture the GPU’s compute-oriented capabilities.

Q: Does the CPU support ECC memory?

A: Yes, the Intel Core Ultra 9 285T supports ECC memory. This is a notable feature for a desktop processor and makes the platform suitable for workstation use where data integrity is critical.

Q: What is the memory bandwidth of the CPU, and how does it compare to the GPU’s bandwidth?

A: The CPU supports dual-channel DDR5 with a memory bandwidth of 102.4 GB/s. The GPU has a 384-bit bus with GDDR7 memory providing 1.34 TB/s of bandwidth — over 13 times the CPU’s memory bandwidth.

Q: What is the combined percentile of this CPU+GPU pair?

A: The combined percentile is 75th, which is lower than the CPU’s individual 91st percentile but higher than the GPU’s 59th percentile. This reflects the performance imbalance between the two components.

Q: Is the CPU multiplier unlocked for overclocking?

A: No, the multiplier is locked (multiplierUnlocked: false). The 35 W TDP and locked multiplier mean the CPU is designed for efficient operation at stock settings, not for enthusiast overclocking.

Q: What PSU wattage is suggested for this GPU?

A: The suggested PSU for the RTX 5090 D V2 is 950 W. The GPU has a 575 W TDP and requires a single 16-pin power connector.

Who Should Build It

This configuration targets a narrow but specific user profile: someone who needs a powerful 24-core CPU for multi-threaded productivity workloads but who also requires the GPU’s 24 GB VRAM for large dataset processing, AI inference, or GPU-accelerated rendering. The CPU’s 35 W TDP makes this an excellent choice for quiet workstations or builds where power efficiency is paramount, and the 91st percentile CPU performance ensures no compute workload will be CPU-bound.

Gamers at 4K who prioritize VRAM capacity over raw frame rates will find the 24 GB frame buffer and 1.34 TB/s bandwidth appealing, though the 59th percentile GPU benchmark score suggests this is not the fastest gaming GPU available. Content creators working with 8K video or complex 3D scenes will benefit from the combination of CPU compute and GPU VRAM, but they should verify their specific software leverages the GPU’s compute capabilities rather than just its rasterization performance.

Software developers building large parallel applications will appreciate the 24 threads and strong Passmark integer math score (132433), while the ECC memory support provides data integrity for long-running builds. Students and office workers are poorly served by this build — the GPU’s power requirements and cost are unjustified for productivity tasks, and the CPU’s integrated Arc graphics are sufficient for those workloads without the discrete GPU.

Small business workstations running CPU-bound simulation, financial modeling, or data analysis software will see strong performance from the 285T, and the GPU can be repurposed for CUDA-accelerated compute when needed. The upgrade path is limited by the platform’s current-generation status, but the existing configuration is well-balanced for compute-heavy workloads, with the CPU’s efficiency and the GPU’s memory capacity providing a solid foundation for multi-year use.