SYSTEM ANALYZER

Rate My PC: Intel Core Ultra 5 245T + NVIDIA GeForce RTX 5090 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
92%
VS
GPU
98%
PROCESSOR

Intel Core Ultra 5 245T

38,194 Benchmark Score
Top 8% Market Ranking
View Full Specs →
GRAPHICS CARD

NVIDIA GeForce RTX 5090 D

77,712 Benchmark Score
Top 2% 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

# Intel Core Ultra 5 245T + NVIDIA GeForce RTX 5090 D: A Desktop Powerhouse Analysis

This pairing combines Intel's 14-core Arrow Lake desktop processor with NVIDIA's flagship Blackwell-architecture GPU, creating a desktop build that sits at the 89th percentile overall among all tracked configurations. The CPU, a 65-watt part with a 5.10 GHz boost clock, pairs with a 575-watt GPU that delivers 104.8 TFLOPS of FP32 compute, making this a system with substantial compute capability across both CPU-bound and GPU-bound workloads. The data indicates a desktop-class configuration where the GPU overwhelmingly dominates the performance envelope, though the CPU's efficiency-oriented design raises interesting questions about balance in various application scenarios.

CPU Analysis

The Intel Core Ultra 5 245T is a 14-core, 14-thread desktop processor built on the Arrow Lake architecture, manufactured on a 3 nm process by TSMC. The chip contains 17,800 million transistors on a 243 mm² die, with a base clock of 2.20 GHz and a boost clock of 5.10 GHz. The 65-watt TDP is notably modest for a desktop CPU with this core count and boost capability, suggesting a design focused on power efficiency while maintaining high single-thread performance. This is not a hyperthreaded part — 14 cores and 14 threads means each core handles one thread, which impacts heavily threaded workloads compared to CPUs with simultaneous multithreading.

The cache hierarchy consists of 192 KB of L1 per core, 3 MB of L2 per core, and 24 MB of shared L3 cache. The L3 figure is moderate by modern desktop standards, and the per-core L2 allocation is generous, which helps with data locality in single-threaded tasks. Memory support is DDR5 with dual-channel configuration, providing 102.4 GB/s of bandwidth, and ECC memory is supported — a feature relevant for workstation and data-integrity-sensitive use cases. The CPU provides 20 PCIe Gen 5 lanes, enabling high-bandwidth connectivity for modern GPUs and NVMe storage. Integrated graphics are present as Arc Xe-LPG with 64 execution units, providing basic display output and media capabilities without a discrete GPU.

Benchmark results show a processor that excels in single-thread performance relative to its multi-thread scaling. The Cinebench R23 single-core score of 3699 is strong, while the multi-core score of 26208 indicates a roughly 7x scaling factor across the 14 cores — typical for a design without SMT. In PassMark tests, the single-thread score of 4367 is competitive, while the multi-thread score of 30833 reflects the absence of extra threads. Data compression achieves 283812, encryption scores 23656, and extended instructions hit 22071. Integer math scores 88676, floating-point math reaches 108499, and prime number finding (a test sensitive to both clock speed and core efficiency) scores 324.

The CPU's 86th percentile ranking among all CPUs places it in the upper tier of desktop processors. Its average benchmark score of 38194 puts it nearly dead-even with its nearest rivals: the AMD Ryzen 7 250 (38221, deltaPct -0.1), the Intel Core i5-14490F (38149, deltaPct 0.1), the Intel Core i5-13600HX (38261, deltaPct -0.2), and the Intel Core i5-13600KF (38103, deltaPct 0.2). These sub-0.5% deltas indicate the Ultra 5 245T sits at a performance equilibrium with these mid-range desktop and high-end mobile parts, meaning real-world differences would be imperceptible. The 65-watt TDP is notably lower than typical desktop rivals in this performance class, suggesting the efficiency advantage is real.

Benchmark Performance

The combined benchmark picture shows a system with a powerful but not class-leading CPU paired with a GPU that sits near the top of the performance hierarchy. The CPU's average benchmark score of 38194 places it at the 86th percentile, while the GPU's average score of 77712 puts it at the 92nd percentile. The combined build percentile is 89th, indicating that the CPU slightly lags the GPU in relative standing.

The GPU benchmarks are dominated by the RTX 5090 D's compute capabilities. The 3DMark Steel Nomad DX12 score of 14326 reflects strong rasterization performance in modern APIs. Geekbench scores show a substantial gap between OpenCL (310674) and Vulkan (376915), suggesting the Vulkan path is particularly well-optimized. PassMark G3D scores 44065, while GPU compute scores 28396 — the gap between these two indicates that the card's compute performance is strong but not proportionally as dominant as its graphics rendering capabilities. DirectX 10, 11, and 12 scores (231, 371, and 219 respectively) are notably lower in PassMark's legacy tests, likely due to driver overhead in these older APIs.

The GPU's nearest rivals, based on average benchmark scores, present an interesting comparison. The AMD Radeon RX 6650M XT (76904, deltaPct 1.1) and RX 6850M XT (78940, deltaPct -1.6) are mobile parts that score within 2% of the RTX 5090 D, as are the NVIDIA Tesla P100 PCIe 12GB (79396, deltaPct -2.1) and P100 PCIe 16GB (79605, deltaPct -2.4). These deltas are small, but the RTX 5090 D's 32 GB of GDDR7 memory and 1.79 TB/s bandwidth place it in a different memory-capability class — the benchmark scores don't capture memory-bound workloads effectively. The GPU's 92nd percentile ranking means it outperforms the vast majority of tracked GPUs, though the nearest rivals being mobile and datacenter parts suggests the desktop GPU landscape is competitive at this level.

Gaming Performance

No measured FPS data exists for this exact CPU+GPU combination — the FACT PACK contains no measuredFps rows. All gaming frame-rate discussion is therefore estimated from the benchmark scores and should be treated as an approximation. The dataIsMeasured flag is false, so these are projections, not observed results.

Based on the CPU's 86th percentile and the GPU's 92nd percentile, this system should deliver extremely high frame rates at 1080p and 1440p in most titles, with the GPU being the primary driver of performance. The CPU's strong single-thread score of 3699 in Cinebench R23 and 4367 in PassMark single-thread suggests it can feed a high-end GPU without bottlenecking in most games. At 4K resolution, the GPU's immense compute (104.8 TFLOPS FP32) and 32 GB of VRAM would allow maximum settings with ray tracing enabled, and the 1.79 TB/s memory bandwidth provides ample headroom for high-resolution textures.

The RTX 5090 D's 170 RT cores and 680 tensor cores suggest exceptional ray tracing and DLSS performance, though no specific FPS figures are available. Given the GPU's 3DMark Steel Nomad score of 14326, this is clearly a top-tier gaming GPU for rasterization as well. The CPU's 14 cores without SMT may show occasional stutter in highly threaded games that benefit from more threads, but the high boost clock mitigates this concern. For competitive esports titles, the system should easily exceed 144 Hz at 1080p, while AAA titles at 4K with max settings would see 60+ FPS in most cases — these are estimates based on the percentile rankings and benchmark scores, not measured data.

Usage Scenarios

High-refresh gaming: The GPU's 92nd percentile and the CPU's 86th percentile combine to support very high frame rates at 1080p and 1440p. The GPU's 104.8 TFLOPS FP32 performance and 1.79 TB/s bandwidth provide ample compute for 144 Hz and above in most titles, though the CPU's lack of SMT could limit frame pacing in some multi-threaded game engines.

Streaming: With 14 cores at 65 watts, the CPU has enough headroom for encoding overhead while gaming, and the GPU's NVENC-capable hardware (implied by the 680 tensor cores and modern Blackwell architecture) can offload encoding entirely. The system's combined 89th percentile ranking supports smooth streaming at high bitrates, though specific encoding performance is not benchmarked.

Video editing: The PassMark data compression score of 283812 and floating-point math score of 108499 indicate strong CPU performance for video processing tasks. The GPU's 32 GB of VRAM is exceptional for timeline rendering and effects work, and the 1.79 TB/s bandwidth accelerates 4K and 8K timeline scrubbing. The CPU's 24 MB L3 cache and dual-channel DDR5 at 102.4 GB/s provide adequate memory bandwidth for most editing workloads.

3D rendering: The GPU's 104.8 TFLOPS FP32 and 170 RT cores make this an excellent choice for GPU-accelerated rendering in Blender, Octane, or V-Ray. The CPU's Cinebench R23 multi-core score of 26208 is respectable for CPU-based rendering, though it won't match high-core-count workstation chips. The 92nd GPU percentile ensures render times are short, and 32 GB VRAM handles large scenes without out-of-core swapping.

Software development: The CPU's strong single-thread performance (3699 Cinebench R23 single-core) accelerates compilation of single translation units, while the 14 cores handle parallel builds. The PassMark extended instructions score of 22071 suggests good SIMD performance for scientific computing. ECC memory support is a plus for long-running builds and data-intensive applications.

Student and office work: This system is massively overprovisioned for office tasks. The CPU's 65-watt TDP keeps power draw modest during light loads, and the integrated Arc Xe-LPG graphics would suffice for office work without the discrete GPU. The 86th CPU percentile is far beyond what word processing or spreadsheet tasks require, making this more of a luxury than a necessity for this use case.

Upgrade Path and Platform

The CPU uses Intel Socket 1851, which is the LGA 1851 platform for Arrow Lake-S desktop processors. Memory support is DDR5 dual-channel with 102.4 GB/s bandwidth, and ECC memory is supported — a feature more common in workstation platforms. The CPU provides 20 PCIe Gen 5 lanes, which is sufficient for one full-bandwidth GPU and one Gen 5 NVMe drive. The 65-watt TDP means the CPU has significant thermal and power headroom in most motherboards, allowing for future upgrades to higher-TDP parts on the same socket if the platform supports them.

The GPU requires a 950-watt suggested PSU, which is substantial headroom above the CPU's 65 watts and the GPU's 575-watt TDP. The system's total power draw with both components at full load would be around 640 watts plus system components, so the 950-watt recommendation provides about 50% headroom. The GPU uses a 1x 16-pin power connector and is dual-slot, 304 mm long, 137 mm tall, and 48 mm wide — fitting most mid-tower and full-tower cases. PCIe 5.0 x16 interface ensures compatibility with current and future motherboards.

A sensible next upgrade path would focus on the CPU side, as the GPU is already at the 92nd percentile. Moving to a higher-core-count Arrow Lake or future socket-1851 compatible CPU would improve multi-threaded workloads without changing the platform. Alternatively, adding more DDR5 memory could benefit memory-sensitive applications, though the dual-channel configuration means the 102.4 GB/s bandwidth is fixed by the memory bus. The 20 PCIe Gen 5 lanes also allow for adding a second Gen 5 NVMe drive for storage-heavy workloads.

FAQ

Q: Does the CPU support ECC memory?

A: Yes, the Intel Core Ultra 5 245T supports ECC memory, which is valuable for data-integrity-sensitive workloads such as scientific computing or long-running server tasks.

Q: What is the launch MSRP of the CPU and GPU?

A: The CPU has a launch MSRP of $270, and the GPU has a launch MSRP of 2,299 USD.

Q: How many PCIe lanes does the CPU provide?

A: The CPU provides 20 PCIe Gen 5 lanes, which supports a full-bandwidth GPU and one Gen 5 NVMe drive, or other Gen 5 devices.

Q: What is the GPU's memory configuration?

A: The RTX 5090 D has 32 GB of GDDR7 memory on a 512-bit bus, providing 1.79 TB/s of memory bandwidth.

Q: Is the CPU's multiplier unlocked for overclocking?

A: No, the multiplier is locked, meaning overclocking is not supported through standard unlocked multiplier methods.

Q: What is the CPU's process node and foundry?

A: The CPU is manufactured on a 3 nm process by TSMC, with 17,800 million transistors on a 243 mm² die.

Q: What does the 89th combined percentile mean?

A: The combined percentile of 89 indicates that this CPU+GPU pairing outperforms 89% of all tracked desktop configurations in the benchmark database, based on the CPU's 86th and GPU's 92nd percentile rankings.

Who Should Build It

This system targets enthusiasts and professionals who need maximum GPU compute with a power-efficient CPU. Gamers at 4K resolution will benefit most, as the GPU's 92nd percentile and 32 GB VRAM handle high-resolution textures and ray tracing without compromise. Content creators working with 8K video or large 3D scenes will find the 32 GB VRAM and 1.79 TB/s bandwidth transformative for their workflow. Developers working on parallel codebases will appreciate the 14-core CPU with strong single-thread performance (4367 PassMark single-thread) and ECC memory support. Students in GPU-intensive fields like machine learning or computational graphics would benefit from the compute capabilities, though the cost is substantial. Small business workstations requiring ECC memory and high-throughput PCIe Gen 5 connectivity could leverage this platform for data processing and rendering tasks. The 65-watt CPU keeps idle and light-load power draw low, making this viable for always-on workstations.

Build Overview

This is a desktop-class build combining the Intel Core Ultra 5 245T, a 14-core Arrow Lake processor with a 65-watt TDP and 86th percentile CPU ranking, with the NVIDIA GeForce RTX 5090 D, a Blackwell-architecture GPU with 32 GB GDDR7 memory, 104.8 TFLOPS FP32 performance, and a 92nd percentile GPU ranking. The combined system ranks at the 89th percentile overall, placing it in the upper tier of desktop configurations. This pairing represents an efficiency-focused CPU matched with a raw-performance GPU, creating a system where the GPU is the clear performance leader. The CPU's 86th percentile and the GPU's 92nd percentile indicate a performance gap that will manifest in GPU-bound workloads and may create a bottleneck in certain CPU-heavy scenarios.

Balance and Bottleneck

The data reveals a significant performance asymmetry: the GPU's 92nd percentile sits well above the CPU's 86th percentile, creating a system where the GPU is the dominant component. In GPU-bound workloads like 4K gaming or GPU rendering, the RTX 5090 D will operate near its full potential, with the CPU providing sufficient single-thread performance (3699 Cinebench R23 single-core) to avoid major frame pacing issues. However, in CPU-bound scenarios — such as 1080p gaming at high frame rates or multi-threaded productivity tasks — the CPU's 14 threads without SMT may limit overall performance.

The PassMark multi-thread score of 30833 versus the GPU's compute score of 28396 suggests that in mixed workloads, the GPU's compute advantage (relative to its own class) is less pronounced than its graphics advantage. The CPU's nearest rival deltas (all within 0.2%) indicate it's well-matched against mid-range competitors, but the GPU's nearest rivals are mobile and datacenter parts with smaller deltas (1.1% to -2.4%), suggesting the GPU is more competitive within its peer group. In gaming, the FPS scaling would likely show a CPU bottleneck at 1080p in esports titles where frame rates exceed 200 FPS, while at 4K the GPU becomes the limiting factor. The 575-watt GPU TDP versus the 65-watt CPU TDP underscores this imbalance — the system is designed for GPU-heavy workloads, and the CPU is sufficient but not outstanding for the GPU's class.

GPU Analysis

The NVIDIA GeForce RTX 5090 D is a Blackwell 2.0 architecture GPU built on a 5 nm TSMC process, containing 92,200 million transistors on a 750 mm² die with a transistor density of 122.9M per mm². The GPU operates at a base clock of 2017 MHz and boosts to 2407 MHz, with memory clocked at 1750 MHz (28 Gbps effective). The 32 GB of GDDR7 memory on a 512-bit bus delivers 1.79 TB/s of bandwidth — an exceptional figure that supports high-resolution textures and large datasets.

The compute configuration includes 21760 shading units, 680 TMUs, 176 ROPs, 170 RT cores, and 680 tensor cores. This translates to 104.8 TFLOPS of FP32 throughput and 104.8 TFLOPS of FP16 (1:1 ratio), with pixel rate of 423.6 GPixel/s and texture rate of 1,636.8 GTexel/s. The 170 RT cores and 680 tensor cores indicate robust ray tracing and AI acceleration capabilities, though specific RT benchmark scores are not provided. The GPU's 92nd percentile ranking and average benchmark score of 77712 place it among the top GPUs, with nearest rivals including the AMD Radeon RX 6650M XT (76904, +1.1%) and RX 6850M XT (78940, -1.6%), as well as NVIDIA Tesla P100 variants. The 3DMark Steel Nomad score of 14326 is strong for DX12 gaming, while Geekbench Vulkan (376915) outperforms OpenCL (310674), suggesting good API optimization. The 575-watt TDP with a 950-watt suggested PSU indicates substantial power requirements, and the dual-slot design (304mm x 137mm x 48mm) fits most full-tower cases. Display outputs include 1x HDMI 2.1b and 3x DisplayPort 2.1b, with support for DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4.