SYSTEM ANALYZER

Rate My PC: Intel Core Ultra 5 235T + NVIDIA GeForce RTX 3070

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

91 / 100
ULTIMATE READY

Apex Performer

Top 9% 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
90%
PROCESSOR

Intel Core Ultra 5 235T

38,561 Benchmark Score
Top 8% Market Ranking
View Full Specs →
GRAPHICS CARD

NVIDIA GeForce RTX 3070

17,208 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

# Intel Core Ultra 5 235T + NVIDIA GeForce RTX 3070: A Desktop Pairing Analysis

The Intel Core Ultra 5 235T and NVIDIA GeForce RTX 3070 form a desktop build that sits at the 74th combined percentile for all CPU+GPU pairings. The CPU, part of the Core Ultra Series 2 on the Arrow Lake architecture, delivers an average benchmark score of 38,561, placing it in the 86th percentile among all CPUs. The GPU, an Ampere-generation GeForce 30-series part, achieves an average benchmark score of 17,208, landing in the 61st percentile among all GPUs. This pairing combines a top-quartile CPU with a mid-range GPU, creating an interesting dynamic where the processor has substantially more compute headroom than the graphics card can fully utilize in most scenarios.

Benchmark Performance

The Intel Core Ultra 5 235T demonstrates strong multi-threaded performance with a Cinebench R23 multi-core score of 26,232 and a single-core score of 3,703. These numbers place the CPU in the 86th percentile overall, and the nearest rivals highlight just how tight the competition is at this performance tier. The Intel Core i5-14500 scores 38,531 on average, a mere 0.1% behind the 235T, while the Intel Xeon w3-2525 trails by 0.4% with an average score of 38,392. The Intel Core 9 270H and Intel Core Ultra 9 285H are also within 0.7% of the 235T's average, suggesting that this CPU sits at a performance plateau where several different architectures converge.

Single-thread performance is particularly noteworthy. The Cinebench R20 single-core score of 1,555 and the R15 single-core score of 373 indicate strong per-core efficiency, which is critical for lightly-threaded workloads like gaming and everyday applications. The PassMark single-thread score of 4,339 reinforces this picture. Multi-threaded scaling is also solid, with the PassMark multi-thread score of 30,918 and the Cinebench R23 multi-core result of 26,232 showing that the 14 cores and 14 threads can handle heavily parallel workloads effectively. The data compression score of 295,100 in PassMark suggests strong integer throughput, while the floating-point math score of 106,546 indicates robust compute capability for scientific and engineering applications.

On the GPU side, the RTX 3070 achieves a PassMark G3D score of 22,214 and a Geekbench OpenCL score of 112,821. The 3DMark Steel Nomad DX12 score of 3,162 provides a modern DirectX 12 gaming workload measurement. The GPU's percentile ranking of 61 places it in the upper-middle tier of all GPUs, and its nearest rivals show a competitive landscape: the AMD Radeon RX 7600 XT is just 0.7% ahead with an average score of 17,083, while the NVIDIA GeForce GTX 690 is 1% behind at 17,037. The AMD Radeon HD 7970M trails by 1.1%, and the NVIDIA Tesla K40c is 1.5% ahead — these are all very close margins, indicating that the RTX 3070's performance class has several viable alternatives.

The combined picture is one of asymmetry. The CPU outperforms the GPU in relative terms by a significant margin — 86th percentile versus 61st percentile. This means the system's overall capability is often constrained by the graphics card, particularly in GPU-bound workloads like high-resolution gaming and 3D rendering. However, in CPU-bound tasks such as software compilation, data processing, and productivity applications, the 235T can stretch its legs and deliver performance that rivals much more expensive processors.

Balance and Bottleneck

The data reveals a clear bottleneck profile: the RTX 3070 is the limiting component in graphics-intensive workloads, while the CPU has substantial headroom for compute-heavy tasks. The CPU's 86th percentile ranking versus the GPU's 61st percentile creates a delta of 25 percentile points, which is substantial. In gaming at 1080p or 1440p, the CPU's strong single-thread performance — evidenced by the Cinebench R23 single-core score of 3,703 — will rarely be the limiting factor. Instead, the GPU's compute throughput, measured at 20.31 TFLOPS FP32, will determine frame rates in most modern titles.

The GPU's memory bandwidth of 448.0 GB/s over a 256-bit bus is respectable for its class, but the 8 GB VRAM capacity may become a constraint in texture-heavy games at higher resolutions. The CPU's memory bandwidth of 102.4 GB/s on dual-channel DDR5 is adequate for feeding the 14 cores, but it is not exceptional. In CPU-bound scenarios, the 235T's 14 cores and 14 threads (note: no hyperthreading, so threads equal cores) will process data efficiently, but the lack of additional threads could limit performance in workloads that benefit from more than 14 concurrent threads.

The PassMark physics score of 2,157 and the find prime numbers score of 318 suggest that the CPU handles physics calculations and integer-heavy workloads well, but these are not the primary bottlenecks in a gaming system. The GPU's pixel rate of 165.6 GPixel/s and texture rate of 317.4 GTexel/s are the figures that matter most for rasterization-heavy games. In compute workloads like machine learning inference or video encoding, the GPU's tensor cores (184 of them) and RT cores (46) provide acceleration that the CPU cannot match, reinforcing the GPU as the performance ceiling for those tasks.

The FPS scaling picture, while not measured for this exact pairing, can be inferred from the percentile gap. In CPU-bound games (typically esports titles at low settings), the 235T's single-core strength will push frame rates high, but the GPU will still need to render frames, and at 1080p the RTX 3070 should keep pace. In GPU-bound games at 1440p or 4K, the CPU's advantage becomes irrelevant as the GPU saturates. The practical bottleneck is the GPU in most gaming scenarios, which is typical for a build with this percentile disparity.

Who Should Build It

This pairing targets users who need strong CPU performance for productivity but game at resolutions where the RTX 3070 can keep up. Gamers at 1080p will find the combination well-suited for high-refresh-rate play, leveraging the CPU's 86th percentile single-thread performance to drive high frame rates in less demanding titles. At 1440p, the RTX 3070's 61st percentile GPU performance will handle most games at high settings, though the 8 GB VRAM may require lowering textures in some modern titles.

Content creators who work with video editing, 3D rendering, or software development will benefit from the CPU's multi-threaded prowess. The Cinebench R23 multi-core score of 26,232 and the PassMark multi-thread score of 30,918 indicate that the 235T can handle video encoding, 3D scene compilation, and code builds efficiently. The GPU's CUDA cores (5,888 shading units) and tensor cores provide acceleration for GPU-accelerated rendering and AI workloads, making this a capable workstation for creators who use NVIDIA's ecosystem.

Students and small business users running office applications, web development, or light data analysis will find the CPU's performance more than adequate, with the integrated Arc Xe-LPG Graphics 24EU providing a fallback if the discrete GPU is not needed. The 65 W TDP of the CPU and 220 W TDP of the GPU mean the system can be built in a compact desktop with a 550 W suggested PSU. The CPU's 3 nm process node (TSMC) and the GPU's 8 nm node (Samsung) represent different manufacturing generations, but the combination works well for a mid-range desktop that prioritizes CPU compute over GPU extremes.

Usage Scenarios

High-refresh gaming: The CPU's strong single-thread score of 4,339 in PassMark and the GPU's 20.31 TFLOPS FP32 throughput suggest this build can drive 1080p displays at high refresh rates in most titles. The GPU's 165.6 GPixel/s pixel rate and 317.4 GTexel/s texture rate provide ample fill for 1080p rendering, while the CPU's 5.00 GHz boost clock ensures frame pacing remains consistent.

Streaming: The GPU's NVENC encoder (part of the Ampere architecture) offloads video encoding from the CPU, and the 14-core processor can handle game logic and streaming software simultaneously. The PassMark data encryption score of 23,457 indicates decent cryptography throughput for secure streaming protocols.

Video editing: The Cinebench R23 multi-core score of 26,232 and the PassMark floating-point math score of 106,546 suggest strong performance for video rendering and effects processing. The GPU's 184 tensor cores accelerate AI-based editing features in compatible software, and the 448.0 GB/s memory bandwidth helps move frame data quickly.

3D rendering: The GPU's 46 RT cores and 20.31 TFLOPS FP32 compute provide hardware-accelerated ray tracing and rasterization for real-time viewports. For final-frame CPU rendering, the 14-core 235T with 24 MB of L3 cache will handle most scenes, though render times will be longer than with a higher-core workstation CPU.

Software development: The PassMark integer math score of 84,244 and the data compression score of 295,100 indicate strong performance for code compilation, data serialization, and build processes. The 3 MB L2 cache per core helps keep frequently accessed data close to the execution units.

Student and office work: The CPU's 86th percentile ranking ensures smooth multitasking across office suites, web browsers, and communication tools. The integrated Arc Xe-LPG Graphics provides a low-power fallback for basic display tasks, and the 65 W TDP keeps the system quiet and cool under light loads.

GPU Analysis

The NVIDIA GeForce RTX 3070 is built on the GA104 chip using Samsung's 8 nm process, with 17,400 million transistors on a 392 mm² die. The memory subsystem consists of 8 GB of GDDR6 on a 256-bit bus, delivering 448.0 GB/s of bandwidth. The memory clock runs at 1750 MHz, which translates to 14 Gbps effective. The GPU's base clock is 1500 MHz with a boost clock of 1725 MHz, and it draws a TDP of 220 W with a suggested PSU of 550 W.

Compute resources include 5,888 shading units, 184 texture mapping units, and 96 raster output units. The 46 RT cores provide hardware-accelerated ray tracing, while the 184 tensor cores handle AI and deep learning workloads. The FP32 performance of 20.31 TFLOPS is identical to the FP16 performance (20.31 TFLOPS at 1:1 ratio), which is notable for a GPU of this generation. The pixel rate of 165.6 GPixel/s and texture rate of 317.4 GTexel/s are solid figures for 1080p and 1440p rendering.

The benchmark results place the GPU in the 61st percentile, with a PassMark G3D score of 22,214 and a 3DMark Steel Nomad DX12 score of 3,162. The Geekbench Vulkan score of 21,022 and OpenCL score of 112,821 indicate strong compute performance for non-gaming workloads. The nearest rivals — the AMD Radeon RX 7600 XT at +0.7%, the NVIDIA GeForce GTX 690 at +1%, and the AMD Radeon HD 7970M at +1.1% — all perform within a couple of percent, suggesting that the RTX 3070 is in a tightly contested performance tier.

The 8 GB VRAM is the most significant limitation. At 4K resolution or with high-resolution texture packs, this capacity may be insufficient, causing texture streaming or quality reductions. However, for 1080p and 1440p gaming, 8 GB is generally adequate for most titles at high settings. The GPU supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, ensuring compatibility with modern game APIs.

Gaming Performance

There are no measured FPS rows for this exact CPU+GPU combination in the FACT PACK. All FPS figures below are estimates derived from the benchmark scores and performance percentiles.

At 1080p ultra settings, the pairing should deliver playable frame rates in most games. The CPU's 86th percentile single-thread performance, evidenced by the Cinebench R23 single-core score of 3,703, ensures that the processor will not bottleneck the GPU in most titles. The RTX 3070's 20.31 TFLOPS FP32 throughput and 165.6 GPixel/s pixel rate suggest that frame rates will be GPU-limited, and the 61st percentile GPU ranking indicates performance slightly above the median for all GPUs.

At 1440p, the GPU's 448.0 GB/s memory bandwidth and 8 GB VRAM will be more heavily taxed. The 3DMark Steel Nomad DX12 score of 3,162 suggests that modern DirectX 12 titles will run at high settings with acceptable frame rates, though the 8 GB VRAM may require texture quality reductions in the most demanding games. The GPU's 46 RT cores enable ray tracing at 1440p, though enabling this feature will reduce frame rates due to the compute overhead.

At 4K, the RTX 3070 is likely to struggle with ultra settings in demanding titles. The pixel rate of 165.6 GPixel/s is sufficient for 4K output, but the 8 GB VRAM and 20.31 TFLOPS compute are below what is typically needed for consistent 60 FPS at 4K ultra in AAA games. Users targeting 4K should consider lowering settings or using DLSS (supported via the tensor cores) to maintain playable frame rates.

The GPU's PassMark G3D score of 22,214, compared to rivals within 1.5%, indicates that frame rates will be similar to systems using the AMD Radeon RX 7600 XT or the older NVIDIA GTX 690. The PassMark DirectX 12 score of 85 and DirectX 11 score of 182 provide relative performance indicators across API generations, with the DX11 score being notably higher, suggesting that older titles may run particularly well.

CPU Analysis

The Intel Core Ultra 5 235T features 14 cores and 14 threads — there is no hyperthreading, so the thread count equals the core count. The base clock is 2.20 GHz with a boost clock of 5.00 GHz, and the TDP is rated at 65 W. The processor is built on the Arrow Lake architecture (Arrow Lake-S codename) using TSMC's 3 nm process node, with 17,800 million transistors on a 243 mm² die. The cache hierarchy consists of 192 KB L1 per core, 3 MB L2 per core, and 24 MB of shared L3 cache.

Single-thread performance is strong, with a Cinebench R23 single-core score of 3,703 and a PassMark single-thread score of 4,339. The boost clock of 5.00 GHz contributes to this, and the 3 nm process helps maintain power efficiency. Multi-threaded performance is also impressive, with a Cinebench R23 multi-core score of 26,232 and a PassMark multi-thread score of 30,918. The PassMark data compression score of 295,100 and integer math score of 84,244 indicate strong throughput for data-heavy workloads.

The CPU's average benchmark score of 38,561 places it in the 86th percentile, and the nearest rivals — the Intel Core i5-14500 (0.1% behind), Intel Xeon w3-2525 (0.4% behind), Intel Core 9 270H (0.6% behind), and Intel Core Ultra 9 285H (0.7% behind) — all cluster within a percentage point. This suggests that the 235T is near the top of its performance class, but the competition is extremely tight. The PassMark physics score of 2,157 and extended instructions score of 23,212 round out the CPU's compute profile.

The CPU supports DDR5 memory in dual-channel configuration with a bandwidth of 102.4 GB/s, and it does not support ECC memory. It has 20 PCIe Gen 5 lanes (CPU only), which provides ample bandwidth for a modern GPU and NVMe storage. The integrated Arc Xe-LPG Graphics with 24 execution units provides a basic display output and can handle light graphics tasks, though the discrete RTX 3070 will be the primary rendering solution. The CPU's 3 nm process node is notably more advanced than the GPU's 8 nm node, reflecting a generational difference in manufacturing technology.

FAQ

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

A: The combined percentile is 74, placing it in the upper-middle range of all desktop CPU+GPU pairings.

Q: How does the Intel Core Ultra 5 235T compare to its nearest rival, the Intel Core i5-14500?

A: The 235T has an average benchmark score of 38,561, which is 0.1% higher than the Core i5-14500's score of 38,531.

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

A: The RTX 3070 has 8 GB of GDDR6 memory on a 256-bit bus, delivering 448.0 GB/s of bandwidth.

Q: Does the CPU support ECC memory?

A: No, the Intel Core Ultra 5 235T does not support ECC memory; it supports dual-channel DDR5 with a bandwidth of 102.4 GB/s.

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

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

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

A: The suggested PSU for the RTX 3070 is 550 W, and the GPU has a TDP of 220 W.

Q: What is the GPU's DirectX support and API compatibility?

A: The GPU supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Upgrade Path and Platform

The Intel Core Ultra 5 235T uses the Intel Socket 1851, which is the platform for Core Ultra Series 2 processors. The CPU supports DDR5 memory in dual-channel configuration, and the memory bandwidth is rated at 102.4 GB/s. The platform provides 20 PCIe Gen 5 lanes (CPU only), which is sufficient for a high-bandwidth GPU and modern NVMe storage. The CPU's TDP of 65 W is modest, allowing for compact cooling solutions, while the GPU's TDP of 220 W with a suggested PSU of 550 W means the overall system power draw is manageable.

A sensible next upgrade for this system would be a more powerful GPU, as the CPU's 86th percentile performance provides headroom for a faster graphics card. The PCIe 4.0 x16 bus interface of the RTX 3070 is compatible with the CPU's PCIe Gen 5 lanes, and a newer GPU would drop into the same slot. The 8 GB VRAM of the RTX 3070 is the most likely bottleneck in gaming scenarios, so upgrading to a GPU with more memory would be the most impactful change.

The CPU's socket and memory support are current for the Core Ultra Series 2 generation, meaning that a CPU upgrade within the same generation would require a new motherboard. The 65 W TDP allows for a wide range of cooling solutions, from stock coolers to mid-range air coolers. The integrated Arc Xe-LPG Graphics provides a fallback if the discrete GPU is removed or fails, making the system more resilient. The platform's DDR5 support and PCIe Gen 5 lanes ensure that the system is not immediately outdated, and the 3 nm process node suggests that power efficiency is a priority for this generation.

Build Overview

This desktop build pairs the Intel Core Ultra 5 235T — a 14-core Arrow Lake processor on TSMC's 3 nm node — with the NVIDIA GeForce RTX 3070, an Ampere-generation GPU on Samsung's 8 nm node. The CPU's 86th percentile ranking among all CPUs contrasts with the GPU's 61st percentile ranking, creating a system that is CPU-forward in its performance profile. The combined percentile of 74 places this build in the upper-middle tier of all desktop pairings.

The CPU's strengths are single-thread speed (5.00 GHz boost clock, 3,703 Cinebench R23 single-core) and multi-threaded throughput (26,232 Cinebench R23 multi-core), making it suitable for both gaming and productivity. The GPU's 20.31 TFLOPS FP32 compute, 46 RT cores, and 184 tensor cores provide solid rasterization, ray tracing, and AI acceleration, respectively. The 8 GB VRAM and 448.0 GB/s bandwidth are adequate for 1080p and 1440p gaming but may limit 4K performance.

The system's overall tier, reflected in the 74th combined percentile, suggests a build that can handle most modern workloads with good performance, though it is not at the extreme high end. The CPU has room to support a more powerful GPU in the future, making this a balanced starting point for a desktop that can be upgraded over time. The 65 W CPU TDP and 220 W GPU TDP with a 550 W suggested PSU indicate a system that is power-efficient relative to its performance, and the desktop form factor allows for a wide range of case and cooling options.