SYSTEM ANALYZER

Rate My PC: Intel Core Ultra 5 235T + NVIDIA GeForce RTX 5070

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

Build Configuration

Select your components

PROCESSOR (CPU)

▼

GRAPHICS CARD (GPU)

▼

94 Score

Top 6% Ranking

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

92%

GPU

95%

PROCESSOR

Intel Core Ultra 5 235T

38,561 Benchmark Score

Top 8% Market Ranking

View Full Specs →

GRAPHICS CARD

NVIDIA GeForce RTX 5070

40,377 Benchmark Score

Top 5% 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

# Balance and Bottleneck

The Intel Core Ultra 5 235T and NVIDIA GeForce RTX 5070 pairing presents an interesting asymmetry in raw compute capability versus efficiency. The CPU sits at the 86th percentile among all processors, while the GPU holds the 82nd percentile among all graphics cards, and the combined system percentile lands at 84. This narrow spread suggests neither component dramatically overshadows the other in aggregate terms, but the workload-specific scores reveal where bottlenecks emerge.

The CPU’s Cinebench R23 multi-core score of 26,232 against a single-core score of 3,703 indicates strong scaling across its 14 cores, yet the PassMark multi-thread score of 30,918 versus single-thread of 4,339 shows a ratio of roughly 7:1, which is modest for a 14-thread part. This suggests that while the processor can handle parallel workloads competently, its thread count ceiling—14 threads with no hyper-threading—limits extreme multi-threaded scaling. The GPU, by contrast, delivers 30.87 TFLOPS FP32 performance, which is substantial for 1080p and 1440p gaming, but the CPU’s single-thread performance will become the limiting factor in lightly-threaded game engines that rely on a few fast cores.

The data implies that in gaming scenarios, the RTX 5070 will often wait on the Ultra 5 235T for frame pacing, particularly at lower resolutions where GPU load is lighter. PassMark integer math at 84,244 and floating-point math at 106,546 show balanced ALU and FPU capabilities, but the encryption score of 23,457 and data compression of 295,100 indicate that the CPU can saturate memory bandwidth in data-intensive tasks. The GPU’s 672.0 GB/s memory bandwidth is over six times the CPU’s 102.4 GB/s, meaning the graphics card is not constrained by system memory throughput in most rendering workloads. However, in physics simulations—where the CPU scores 2,157 on PassMark physics—the processor becomes the clear bottleneck, as the GPU cannot compensate for CPU-side collision detection or rigid body calculations.

The power draw disparity further clarifies the balance: the CPU’s 65W TDP against the GPU’s 250W TDP means the graphics card consumes nearly four times the power, and the suggested 600W PSU indicates the system is heavily GPU-dominated in terms of electrical load. This does not inherently create a bottleneck—GPU-bound workloads are typical for gaming—but it does mean that in CPU-intensive tasks like video encoding or software compilation, the processor may throttle or limit throughput despite its lower power draw. The 3nm process node for the CPU versus 5nm for the GPU also highlights different design priorities: the CPU maximizes efficiency per watt, while the GPU prioritizes raw throughput. Benchmark results indicate that for most gaming at high settings, the GPU is the primary driver, but for productivity tasks like data compression or encryption, the CPU’s 14 threads and 24MB L3 cache will define the pace.

# Usage Scenarios

High-refresh gaming: The RTX 5070’s 3DMark Steel Nomad DX12 score of 5,077 places it in the 82nd percentile of GPUs, which is capable of driving high frame rates at 1080p and likely 1440p with reduced settings. The CPU’s single-core Cinebench R23 score of 3,703 is strong enough to avoid severe frame drops, but the lack of hyper-threading may cause occasional stutter in games that spawn many background threads. Estimated FPS figures, since no measured data exists for this pair, suggest the GPU will be the primary limiter at 4K, while the CPU becomes more relevant at 1080p.

Streaming: The GPU’s 192 tensor cores and 48 RT cores provide hardware acceleration for encoding and AI-enhanced features, but the CPU’s PassMark extended instructions score of 23,212 indicates modest AVX-512 or similar workload capability. The 14 threads can handle a game plus encoding stream simultaneously, but the 65W TDP may limit sustained multi-threaded performance during long streams. The GPU’s 12GB GDDR7 memory offers enough headroom for game assets plus stream buffers.

Video editing: The CPU’s Cinebench R23 multi-core score of 26,232 is competitive for 4K timeline scrubbing and export tasks, while the GPU’s Geekbench OpenCL score of 172,660 accelerates effects rendering and color grading. The combined system percentile of 84 suggests this pairing can handle professional editing suites, with the CPU managing decode and the GPU handling compute-heavy filters. The 24MB L3 cache helps with large media files, but the 102.4 GB/s memory bandwidth may be a limitation for multi-stream 4K projects.

3D rendering: The GPU’s 30.87 TFLOPS FP32 performance and 201.0 GPixel/s pixel rate make it suitable for real-time viewport rendering, while the CPU’s 14 cores and 24MB cache contribute to ray tracing acceleration. The Cinebench R15 multi-core score of 2,644 indicates strong CPU rendering capability for Blender or similar software, but the GPU will dominate final frame renders due to its massive parallel throughput. The 12GB VRAM may be restrictive for very large scenes or high-resolution textures.

Software development: The CPU’s PassMark data compression score of 295,100 and integer math of 84,244 are excellent for code compilation and build servers, while the single-thread score of 4,339 ensures responsive IDE interaction. The 14 threads allow parallel builds, and the 65W TDP makes it efficient for always-on development machines. The GPU is largely idle in this scenario but can accelerate shader compilation and local ML model testing.

Student and office work: The CPU’s PassMark single-thread score of 4,339 ensures snappy web browsing and office suite performance, while the integrated Arc Xe-LPG Graphics with 24 EUs provides basic display output without needing the RTX 5070. The 65W TDP keeps electricity costs low, and the dual-channel DDR5 memory support at 102.4 GB/s is sufficient for multitasking. The GPU is overkill for this use case, but the system’s 84th combined percentile means it will never feel slow.

# Benchmark Performance

The Intel Core Ultra 5 235T achieves an average benchmark score of 38,561, placing it at the 86th percentile of all CPUs. Its nearest rival, the Intel Core i5-14500, scores 38,531—a delta of just 0.1%, meaning the two are functionally identical in aggregate performance. The Xeon w3-2525 is 0.4% ahead, while the Core 9 270H and Ultra 9 285H are 0.6% and 0.7% ahead respectively. This tight grouping shows the Ultra 5 235T is squarely in the mid-range desktop bracket, not a flagship.

Individual CPU benchmarks reveal strengths and weaknesses. The Cinebench R23 multi-core score of 26,232 is strong for a 65W part, while the single-core score of 3,703 is competitive with higher-TDP rivals. The PassMark multi-thread score of 30,918 and single-thread of 4,339 are consistent with the Cinebench results. The floating-point math score of 106,546 is notably high, suggesting good vector processing capability, while the prime number search score of 318 is low, indicating weaker branch prediction or memory latency characteristics.

The NVIDIA GeForce RTX 5070 posts an average benchmark score of 40,377, at the 82nd percentile of all GPUs. Its nearest rivals include the AMD Radeon Pro 580 at 40,318 (0.1% behind), the Radeon Pro WX 7100 at 40,063 (0.8% behind), and the RTX A500 Mobile at 39,568 (2% behind). The Radeon Pro 5300 is 1.2% ahead at 40,870. This indicates the RTX 5070 sits in a crowded mid-range GPU segment, with performance deltas of under 2% across its nearest competitors.

The GPU’s 3DMark Steel Nomad DX12 score of 5,077 is the key gaming metric, while Geekbench Vulkan at 178,923 and OpenCL at 172,660 show strong compute performance. The PassMark G3D score of 29,137 is substantial, but the DirectX 10 score of 180 and DirectX 9 score of 320 are legacy metrics that do not reflect modern workloads. The combined system picture is a CPU and GPU that are both mid-to-high tier in their respective categories, with the CPU slightly higher in percentile ranking.

# Who Should Build It

This pairing targets users who need a balanced desktop system for gaming and productivity without stepping into flagship territory. Gamers at 1080p and 1440p will find the RTX 5070’s 82nd GPU percentile sufficient for high to ultra settings, while the CPU’s 86th percentile ensures they are not leaving performance on the table in CPU-bound titles. The 12GB VRAM is adequate for current AAA games at those resolutions, though 4K enthusiasts may need to lower texture quality.

Content creators working in video editing or 3D rendering will appreciate the CPU’s Cinebench R23 multi-core score of 26,232 for export times, and the GPU’s 30.87 TFLOPS for GPU-accelerated effects. The 14 threads handle multitasking between preview and render, while the 24MB L3 cache reduces reload times. The system’s 84th combined percentile suggests it can handle professional workloads, but not at the level of dedicated workstation parts.

Software developers compiling large codebases will benefit from the PassMark integer math score of 84,244 and data compression of 295,100, which speed up build processes. The 65W TDP makes it suitable for always-on development servers, and the single-thread score of 4,339 ensures responsive tooling. Students and office workers will find the system overkill, but the 86th CPU percentile means no lag in any productivity application, and the integrated GPU provides a fallback if the discrete GPU is disabled for power savings. Small business workstations running database or virtualization workloads will use the CPU’s 14 threads effectively, while the GPU accelerates any CUDA-based analytics.

# CPU Analysis

The Intel Core Ultra 5 235T is a 14-core, 14-thread processor based on the Arrow Lake architecture, built on TSMC’s 3nm process node. It has a base clock of 2.20 GHz and a boost clock of 5.00 GHz, with a 65W TDP. The lack of hyper-threading means 14 threads is the maximum, which is a deliberate design choice to prioritize efficiency—the 3nm process and 65W envelope allow sustained boost clocks without excessive heat. The die size is 243 mm² with 17,800 million transistors, giving a transistor density that is lower than the GPU’s, reflecting the CPU’s focus on large, fast cores over dense parallel units.

The cache hierarchy includes 192 KB L1 per core, 3 MB L2 per core, and 24 MB shared L3. This 24MB L3 is critical for multi-threaded workloads that share data, as it reduces memory traffic to the dual-channel DDR5 interface. The memory bandwidth of 102.4 GB/s is modest for a modern desktop CPU, and it may be a limiting factor in workloads like data compression that benefit from high memory throughput. The PassMark data compression score of 295,100 is strong despite this, suggesting the L3 cache is effective at containing working sets.

Benchmark scores interpret to real workloads as follows: the Cinebench R23 multi-core score of 26,232 places it in the range of CPUs that can handle 4K video rendering in software, though not as fast as higher-core parts. The single-core score of 3,703 is sufficient for gaming and general desktop use, with PassMark single-thread of 4,339 confirming snappy responsiveness. The floating-point math score of 106,546 indicates good scientific computing capability, while the encryption score of 23,457 is moderate, meaning AES-NI is present but not top-tier. The processor is not unlocked (multiplier locked), so overclocking is not an option, but the 65W TDP and 3nm process mean it runs cool in stock configuration.

# Gaming Performance

No measured FPS data exists for the Intel Core Ultra 5 235T paired with the NVIDIA GeForce RTX 5070 — the FACT PACK contains zero entries in the measuredFpsUltraByGame field. As such, all frame rate figures here are estimates derived from the benchmark scores, not observed results. The GPU’s 3DMark Steel Nomad DX12 score of 5,077 and PassMark G3D of 29,137 indicate it is a strong performer for 1080p and 1440p gaming, while the CPU’s single-core Cinebench R23 score of 3,703 is adequate to feed it in most titles.

For 1080p ultra settings, the RTX 5070’s 30.87 TFLOPS FP32 performance suggests it can push well over 100 FPS in esports titles and 60-100 FPS in AAA games, depending on optimization. The CPU’s 14 threads are sufficient for modern game engines that use 8-12 threads, but the lack of hyper-threading may cause frame drops in heavily threaded titles like Cyberpunk 2077 or Starfield. At 1440p, the GPU becomes more of a bottleneck, and the CPU’s single-thread performance is less critical, so frame rates will scale with GPU load.

At 4K, the 12GB GDDR7 memory and 672.0 GB/s bandwidth will be the limiting factor, as texture-heavy games may exceed 12GB VRAM. Estimated frame rates at 4K ultra are likely 30-60 FPS for demanding titles, with the GPU at near 100% utilization and the CPU waiting. The lack of measured data is a significant gap, but the benchmark scores provide a consistent picture of a GPU that is 82nd percentile and a CPU that is 86th percentile, suggesting no dramatic mismatch. For high-refresh 1080p monitors (144Hz+), the CPU may hold back the GPU in lighter titles, but for 1440p 60Hz displays, the pairing is well-balanced.

# GPU Analysis

The NVIDIA GeForce RTX 5070 uses the GB205 chip based on Blackwell 2.0 architecture, built on TSMC’s 5nm process. It contains 31,100 million transistors on a 263 mm² die, with a transistor density of 118.3M per mm². The GPU has 12GB of GDDR7 memory on a 192-bit bus, delivering 672.0 GB/s of bandwidth—a substantial figure that supports high-resolution textures and compute workloads. The base clock is 2325 MHz with a boost of 2512 MHz, and memory runs at 1750 MHz (28 Gbps effective).

The GPU features 6,144 shading units, 192 texture mapping units, and 80 render output units. It has 48 RT cores and 192 tensor cores, which are dedicated to ray tracing and AI acceleration respectively. The FP32 performance is 30.87 TFLOPS, with FP16 at 30.87 TFLOPS (1:1 ratio), meaning there is no half-precision penalty. The pixel rate is 201.0 GPixel/s and texture rate is 482.3 GTexel/s. The 250W TDP is substantial, with a suggested 600W PSU, and it requires a single 16-pin power connector.

Benchmark results show a 3DMark Steel Nomad DX12 score of 5,077, which is a modern DX12 test, and Geekbench OpenCL of 172,660 and Vulkan of 178,923. The PassMark G3D score of 29,137 is competitive, but the DirectX 9 score of 320 and DirectX 10 of 180 are legacy metrics that do not reflect current performance. The GPU compute score of 15,787 indicates strong general-purpose compute. For rendering, the 48 RT cores accelerate ray tracing significantly, and the 192 tensor cores enable DLSS and AI denoising. The 12GB VRAM is sufficient for most games at 1440p, but may be tight for 4K with ultra textures or for large 3D scenes in rendering software. The 672.0 GB/s bandwidth is excellent for a mid-range GPU, minimizing memory bottlenecks in compute-heavy tasks.

# FAQ

Q: What is the combined performance percentile of this CPU and GPU pairing?

A: The combined system percentile is 84, based on the CPU’s 86th percentile and the GPU’s 82nd percentile.

Q: Does the CPU support overclocking?

A: No, the Intel Core Ultra 5 235T has a locked multiplier (multiplierUnlocked is false), so overclocking is not supported.

Q: What memory type does the CPU support?

A: The CPU supports DDR5 memory in dual-channel configuration, with a memory bandwidth of 102.4 GB/s.

Q: How much VRAM does the RTX 5070 have, and what type?

A: The GPU has 12GB of GDDR7 memory on a 192-bit bus, delivering 672.0 GB/s bandwidth.

Q: What is the power requirement for this system?

A: The CPU has a 65W TDP and the GPU has a 250W TDP, with a suggested PSU of 600W.

Q: Does the CPU have integrated graphics?

A: Yes, it includes Arc Xe-LPG Graphics with 24 execution units, which can handle basic display output.

Q: What is the release date for this CPU and GPU?

A: The CPU was released on January 6, 2025, and the GPU on March 3, 2025.

# Upgrade Path and Platform

The Intel Core Ultra 5 235T uses the Intel Socket 1851, which is the platform for Arrow Lake desktop processors. This is a new socket, so there is no upgrade path to older CPUs, but it supports the Core Ultra Series 2 family, meaning users could upgrade to a higher-tier part like the Core Ultra 9 285H if needed—though the 285H is a mobile part, so a desktop equivalent would be required. The socket supports DDR5 memory in dual-channel mode, and the CPU provides 20 PCIe Gen 5 lanes, which is sufficient for a single GPU plus one NVMe SSD at Gen 5 speeds.

The GPU uses PCIe 5.0 x16, which is fully compatible with the CPU’s Gen 5 lanes. The 65W CPU TDP and 250W GPU TDP together suggest a 600W PSU is adequate, leaving headroom for additional drives or peripherals. The power connectors are a single 16-pin for the GPU, and the CPU uses standard ATX power. The memory bus is dual-channel DDR5, so upgrading from 16GB to 32GB or 64GB requires two DIMMs for optimal bandwidth. The 24MB L3 cache is not upgradable, but the socket’s future compatibility with newer Arrow Lake derivatives may allow a CPU swap without changing the motherboard.

A sensible next upgrade would be to add more DDR5 memory if the user has 16GB, as the 102.4 GB/s bandwidth is shared, and more capacity helps with multitasking. For the GPU, the RTX 5070’s 12GB VRAM is the main limitation for 4K gaming, so a future GPU with 16GB or more would be a logical upgrade, but the 600W PSU would need to be checked for higher-TDP cards. The CPU’s 65W TDP means it is unlikely to be a power bottleneck, so the motherboard and PSU have room for a more powerful processor if the user wants more multi-threaded performance.

# Build Overview

This is a desktop build (buildClass is "desktop") combining the Intel Core Ultra 5 235T and NVIDIA GeForce RTX 5070. The system sits at the 84th combined percentile, meaning it outperforms the majority of desktop systems, but it is not in the top tier. The CPU is a 14-core, 14-thread Arrow Lake processor with a 65W TDP, and the GPU is a 12GB GDDR7 Blackwell 2.0 card with a 250W TDP.

The overall tier is upper-mid-range: the CPU’s 86th percentile and GPU’s 82nd percentile place both components in the high-mid bracket, with the CPU slightly stronger relative to its peers. This is a balanced pairing for 1080p and 1440p gaming, plus productive workloads like video editing and software development. The lack of measured FPS data means gaming expectations are estimated, but benchmark scores indicate no major bottleneck. The system is power-efficient for its class, with a combined TDP of 315W, and the 600W PSU provides adequate headroom. The launch MSRP for the CPU is $247 and for the GPU is 549 USD, though pricing is not a factor in performance analysis. This build is a solid choice for users who want strong all-around performance without stepping into enthusiast-tier pricing.

Rate My PC: Intel Core Ultra 5 235T + NVIDIA GeForce RTX 5070

Apex Performer

System Balance Analysis

Intel Core Ultra 5 235T

NVIDIA GeForce RTX 5070

Market Position

Game Performance Benchmarks

Cyberpunk 2077

Red Dead Redemption 2

Microsoft Flight Simulator

Control

Call of Duty: Warzone

Fortnite

Performance Insights

Optimal Performance

4K Gaming Ready

Performance Tiers Explained

Ultimate

High-End

Mid-Range

Entry Level

Legacy

4K Gaming Performance

Cyberpunk 2077

Red Dead Redemption 2

Microsoft Flight Simulator

Control

Call of Duty: Warzone

Fortnite

1440p Gaming Performance

Cyberpunk 2077

Red Dead Redemption 2

Microsoft Flight Simulator

Control

Call of Duty: Warzone

Fortnite

1080p Gaming Performance

Cyberpunk 2077

Red Dead Redemption 2

Microsoft Flight Simulator

Control

Call of Duty: Warzone

Fortnite

Related Processors

Related Graphics Cards

Compare Your CPU

Compare Your GPU

Related PC Builds

Same CPU with Different GPUs

Same GPU with Different CPUs

Other Popular Builds