SYSTEM ANALYZER

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

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 5 235A

48,201 Benchmark Score
Top 6% 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

CPU Analysis

The Intel Core Ultra 5 235A is a 14-core, 14-thread desktop processor built on Arrow Lake architecture, with a 3.40 GHz base clock and 5.00 GHz boost clock. It uses a 3 nm process from TSMC, containing 17,800 million transistors on a 243 mm² die. The cache hierarchy includes 192 KB L1 per core, 3 MB L2 per core, and 24 MB of shared L3 cache. Memory support is dual-channel DDR5 with 102.4 GB/s bandwidth, and the CPU provides 20 PCIe Gen 5 lanes. The integrated Arc Xe-LPG Graphics 24EU handles basic display output. The chip carries a 65 W TDP and is unlocked? No—the multiplier is locked, so overclocking is constrained.

Benchmark results place this CPU at the 90th percentile among all CPUs. The Cinebench R23 multi-core score of 32,633 and single-core score of 4,607 indicate strong all-around performance. In PassMark, the multi-thread score is 38,392, with single-thread at 4,557. The data shows a peculiar pattern: 14 cores with 14 threads means no Hyper-Threading, yet multi-core scores remain competitive because Arrow Lake's performance cores handle the load efficiently. The Cinebench R20 scores—13,705 multi-core and 1,934 single-core—suggest that heavily threaded workloads like video encoding or 3D rendering will scale well, though not as aggressively as CPUs with simultaneous multithreading.

The nearest rival comparison reveals tight clustering. The AMD Ryzen AI Max PRO 380 scores an average of 48,171, just 0.1% behind the Core Ultra 5 235A's average of 48,201. The Intel Core Ultra 5 245HX is 0.2% ahead at 48,287, while the AMD EPYC 4345P leads by 0.6% with 48,470. The AMD Ryzen 9 3900 trails by 0.6% at 47,918. This means the 235A sits in a performance band where no single rival dominates—differences are within measurement noise. For real workloads, this translates to near-identical performance in multi-core tasks versus these four competitors, with the 235A holding its own in both lightly and heavily threaded scenarios.

The PassMark data compression score of 393,800 and encryption score of 30,136 show strong data-handling capability. Floating-point math at 118,778 and integer math at 88,626 indicate balanced arithmetic performance, while extended instructions score 31,625. The physics score of 2,437 is notably lower, suggesting that some simulated physics workloads may not fully utilize the core layout. Random string sorting at 49,489 and prime number finding at 392 round out the compute profile. The average benchmark score of 48,201 confirms this is a top-decile processor, suitable for demanding productivity and content creation tasks.

GPU Analysis

The NVIDIA GeForce RTX 3070 uses the GA104 chip on Samsung's 8 nm process, packing 17,400 million transistors on a 392 mm² die. It features 8 GB of GDDR6 memory on a 256-bit bus, delivering 448.0 GB/s of bandwidth. The memory clock runs at 1750 MHz (14 Gbps effective). The GPU has 5,888 shading units, 184 texture mapping units, and 96 raster output units. For ray tracing and AI acceleration, it includes 46 RT cores and 184 tensor cores. The base clock is 1500 MHz with a boost of 1725 MHz. Pixel rate is 165.6 GPixel/s and texture rate is 317.4 GTexel/s. FP32 performance reaches 20.31 TFLOPS, with FP16 also at 20.31 TFLOPS (1:1 ratio).

The RTX 3070 sits at the 61st percentile among all GPUs, with an average benchmark score of 17,208. In 3DMark Steel Nomad DX12, it scores 3,162. GeekBench OpenCL yields 112,821, while Vulkan scores 21,022. PassMark G3D shows 22,214, with GPU compute at 11,195. DirectX 12 scores 85, DirectX 11 scores 182, DirectX 10 scores 150, and DirectX 9 scores 247. The G2D score is 1,001. These numbers indicate a GPU that is competent for 1080p and 1440p gaming, though it is not in the top tier of current graphics hardware.

Comparing to nearest rivals, the AMD Radeon RX 7600 XT averages 17,083, which is 0.7% behind the RTX 3070. The NVIDIA GeForce GTX 690 trails by 1% at 17,037, and the AMD Radeon HD 7970M is 1.1% behind at 17,019. The NVIDIA Tesla K40c is 1.5% ahead at 17,468. This clustering means the RTX 3070 performs nearly identically to these alternatives in synthetic benchmarks, with differences unlikely to be perceptible in real-world gaming. The RTX 3070's ray tracing and tensor cores give it a feature advantage over older or non-RT competitors, but the raw compute is comparable.

The GPU supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. Display outputs include 1x HDMI 2.1 and 3x DisplayPort 1.4a. The card is dual-slot, 242 mm long (9.5 inches), and 112 mm tall (4.4 inches), requiring a 1x 12-pin power connector. The suggested PSU is 550 W. The GPU has a TDP of 220 W. It is end-of-life, with the GeForce 40 series as its successor, and was released in August 2020. For rendering workloads, the 20.31 TFLOPS FP32 performance is solid for DCC applications, while the 184 tensor cores accelerate DLSS and AI-based features, though the 8 GB VRAM may limit very large scenes.

FAQ

Q: How does the Intel Core Ultra 5 235A compare to its closest rival in multi-core performance?

A: The data shows the 235A averages 48,201 in benchmark scores, which is 0.1% ahead of the AMD Ryzen AI Max PRO 380 (48,171) and 0.2% behind the Intel Core Ultra 5 245HX (48,287). The AMD EPYC 4345P is 0.6% ahead, while the AMD Ryzen 9 3900 trails by 0.6%.

Q: What is the RTX 3070's percentile ranking and what does it mean?

A: The RTX 3070 sits at the 61st percentile among all GPUs. This means it outperforms roughly 61% of all GPUs in the database, placing it in the mid-to-upper tier but not among the top performers. Its average score of 17,208 places it in a tight cluster with rivals like the RX 7600 XT (17,083, 0.7% behind).

Q: Does the Core Ultra 5 235A support overclocking?

A: No. The multiplier is locked, which means the boost clock of 5.00 GHz is the maximum frequency users can rely on without external clock manipulation. The 65 W TDP suggests it is designed for power efficiency rather than overclocking headroom.

Q: What memory type and bandwidth does the CPU support?

A: The CPU supports dual-channel DDR5 memory with a maximum bandwidth of 102.4 GB/s. It does not support ECC memory. This bandwidth is sufficient for most consumer workloads, though it is modest compared to high-end HEDT platforms.

Q: How much VRAM does the RTX 3070 have and is it enough?

A: The RTX 3070 has 8 GB of GDDR6 memory on a 256-bit bus, delivering 448.0 GB/s bandwidth. For 1080p and 1440p gaming, this is generally sufficient, but at 4K or with high-resolution textures, the capacity may become a limiting factor in some titles.

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

A: The combined percentile is 76, which means this pairing outperforms 76% of all CPU+GPU combinations in the database. This is driven by the CPU's 90th percentile rank, while the GPU's 61st percentile pulls the overall score down slightly.

Q: Is there measured FPS data for this combination?

A: No. The FACT PACK contains no measured FPS data for this exact CPU+GPU pairing. All FPS discussions must be treated as estimates derived from the benchmark scores, not observed gaming performance.

Balance and Bottleneck

The data reveals a significant imbalance between CPU and GPU performance tiers. The Core Ultra 5 235A ranks at the 90th percentile among all CPUs, while the RTX 3070 ranks at the 61st percentile among all GPUs. This 29-percentage-point gap suggests the CPU is substantially stronger relative to its peers than the GPU is relative to its own peers. In gaming workloads, this typically means the GPU becomes the limiting factor at lower resolutions, where the CPU can feed frames faster than the GPU can render them.

At 1080p, the CPU's high single-thread score of 4,607 in Cinebench R23 and PassMark single-thread of 4,557 indicate it can handle high frame rates in CPU-bound scenarios like esports titles. However, the RTX 3070's 61st percentile means it will likely cap frame rates in graphically demanding games. The GPU's PassMark G3D score of 22,214 and 3DMark Steel Nomad DX12 score of 3,162 show mid-tier rasterization performance that will struggle to push very high frame rates at 1440p in AAA titles.

For productivity workloads, the CPU's 90th percentile makes it the clear driver. The Cinebench R23 multi-core score of 32,633 and PassMark multi-thread of 38,392 indicate strong performance in video encoding, 3D rendering, and software compilation. The GPU's compute score of 11,195 in PassMark is less relevant here, but its 20.31 TFLOPS FP32 can accelerate GPU-accelerated rendering in applications like Blender or Premiere Pro. In mixed workloads, the CPU will rarely bottleneck; the GPU will be the constraint for graphics-heavy tasks.

The FPS scaling picture is estimated, not measured. Given the CPU's 90th percentile and the GPU's 61st percentile, the GPU is the more likely bottleneck in most gaming scenarios. At 4K, the GPU's 8 GB VRAM and 448 GB/s bandwidth may also cause texture streaming issues. The 0.7% performance gap between the RTX 3070 and the RX 7600 XT suggests that swapping to a rival GPU would not change the bottleneck situation meaningfully. The CPU's 0.1% lead over the Ryzen AI Max PRO 380 means CPU choice is not the primary constraint.

Benchmark Performance

The Intel Core Ultra 5 235A delivers an average benchmark score of 48,201, placing it at the 90th percentile of all CPUs. Key scores include Cinebench R23 multi-core at 32,633 and single-core at 4,607, Cinebench R20 multi-core at 13,705 and single-core at 1,934, and Cinebench R15 multi-core at 3,289 and single-core at 464. PassMark results show multi-thread at 38,392, single-thread at 4,557, integer math at 88,626, floating-point math at 118,778, and extended instructions at 31,625. Data compression scores 393,800, encryption scores 30,136, physics scores 2,437, and random string sorting scores 49,489.

The NVIDIA GeForce RTX 3070 averages 17,208 in its benchmarks, ranking at the 61st percentile among all GPUs. Its 3DMark Steel Nomad DX12 score is 3,162. GeekBench OpenCL scores 112,821 and Vulkan scores 21,022. PassMark G3D scores 22,214, GPU compute scores 11,195, G2D scores 1,001. DirectX tests show 85 for DX12, 182 for DX11, 150 for DX10, and 247 for DX9.

The combined picture places this pairing at the 76th percentile. The CPU's top-decile performance is the standout feature, while the GPU's mid-tier rank tempers the overall result. The nearest rival comparisons show the CPU is virtually tied with the Ryzen AI Max PRO 380 (0.1% ahead), Core Ultra 5 245HX (0.2% behind), EPYC 4345P (0.6% behind), and Ryzen 9 3900 (0.6% ahead). The GPU is similarly tied with the RX 7600 XT (0.7% ahead), GTX 690 (1% ahead), HD 7970M (1.1% ahead), and Tesla K40c (1.5% behind). When combined, the system's performance is driven by the CPU, with the GPU providing adequate but not exceptional graphics capability.

Usage Scenarios

High-refresh gaming: The CPU's 90th percentile and single-core score of 4,607 in Cinebench R23 support high frame rates in CPU-bound titles, but the RTX 3070's 61st percentile limits overall performance. At 1080p, the GPU can likely push 100+ FPS in esports titles, but AAA games at 1440p will see lower frame rates. The 8 GB VRAM may also constrain texture quality at higher resolutions.

Streaming: The CPU's 14 cores and strong multi-thread score of 38,392 in PassMark can handle software encoding while gaming, though the GPU's NVENC (via tensor cores) offers a more efficient alternative. The 65 W TDP means the CPU runs cool, leaving thermal headroom for sustained streaming sessions.

Video editing: The Cinebench R23 multi-core score of 32,633 and PassMark floating-point score of 118,778 indicate strong performance for timeline scrubbing, effects, and exports. The GPU's 20.31 TFLOPS FP32 accelerates GPU-accelerated effects and rendering in Premiere Pro and DaVinci Resolve, though 8 GB VRAM may limit complex compositions.

3D rendering: The CPU's multi-thread performance is excellent for CPU-based renderers, with Cinebench R20 multi-core at 13,705. The GPU's 20.31 TFLOPS handles GPU rendering well, but the 8 GB VRAM is a constraint for large scenes. The 46 RT cores provide hardware ray tracing acceleration, but the GPU's 61st percentile means it is not top-tier for this workload.

Software development: The CPU's high integer math score of 88,626 and data encryption score of 30,136 support fast compilation and secure development workflows. The 24 MB L3 cache and 102.4 GB/s memory bandwidth reduce compilation bottlenecks. The 14 threads without SMT may slightly impact heavily parallel build systems.

Student and office work: The CPU's 90th percentile ensures snappy responsiveness in productivity applications, while the GPU is overkill for basic office tasks. The integrated Arc Xe-LPG Graphics 24EU can handle display output, allowing the RTX 3070 to be reserved for demanding tasks. The 65 W TDP means low power draw for everyday use.

Who Should Build It

This pairing targets gamers who play at 1080p or 1440p and prioritize CPU-heavy workloads like streaming or content creation alongside gaming. The CPU's 90th percentile makes it ideal for users who need strong multi-thread performance for video editing or 3D rendering, while the GPU's 61st percentile handles mainstream gaming at respectable settings. The system is well-suited for developers who compile code frequently and occasionally game, given the CPU's integer math and encryption scores.

Content creators working with 1080p or 1440p footage will benefit from the CPU's Cinebench R23 multi-core score of 32,633 and the GPU's 20.31 TFLOPS for accelerated effects. Students in engineering or computer science programs will find the CPU's performance sufficient for simulations and coding, though the GPU is more than needed for typical student workloads. Small business workstations requiring fast data processing (compression score 393,800) and encryption (30,136) will perform well, though the GPU is excessive for office tasks.

The 76th combined percentile means this system outperforms most consumer builds, but it is not a top-tier enthusiast machine. Users seeking maximum gaming frame rates should consider a stronger GPU, while those focused on productivity could rely on the integrated graphics and save power. The 14-core, 14-thread configuration is a deliberate choice—no SMT means lower power and heat, but slightly reduced multi-thread throughput compared to 14-core/28-thread alternatives.

Build Overview

This is a desktop pairing of the Intel Core Ultra 5 235A and NVIDIA GeForce RTX 3070. The CPU is part of the Core Ultra Series 2, built on Arrow Lake architecture with a 3 nm process from TSMC. The GPU is from the GeForce 30-series, using the Ampere architecture on Samsung's 8 nm process. The CPU ranks at the 90th percentile among all CPUs, while the GPU ranks at the 61st percentile among all GPUs, resulting in a combined percentile of 76.

The system is a balanced mid-to-high-tier desktop build. The CPU's top-decile performance makes it a strong foundation for productivity, while the GPU's mid-tier rank provides competent gaming and graphics acceleration. The 65 W CPU TDP and 220 W GPU TDP mean the system is power-efficient for its performance class, requiring only a 550 W PSU per the GPU's suggested rating. The CPU's 14 cores and GPU's 5,888 shading units provide a capable combination for most consumer and prosumer workloads, though neither component is the absolute fastest in its class.

Upgrade Path and Platform

The Intel Core Ultra 5 235A uses the Intel Socket 1851, which supports DDR5 memory in a dual-channel configuration with 102.4 GB/s bandwidth. The CPU provides 20 PCIe Gen 5 lanes, while the GPU uses a PCIe 4.0 x16 interface—this means the GPU has sufficient bandwidth, but the CPU's PCIe Gen 5 lanes could accommodate a future GPU upgrade without a platform change. The socket is current for Arrow Lake, and the production status is active, so future BIOS updates may support newer CPUs on the same socket.

The GPU's suggested PSU is 550 W, and the CPU's TDP is 65 W, leaving headroom for upgrades. A reasonable next step would be a GPU with more VRAM or higher compute performance, as the RTX 3070's 8 GB VRAM and 61st percentile rank are the primary limits. The PCIe 4.0 x16 interface is backward-compatible with PCIe 5.0 cards, though performance gains from PCIe 5.0 are minimal for gaming. The CPU's 14 threads without SMT may become a bottleneck in heavily threaded workloads, so a future upgrade to a higher-core-count Arrow Lake CPU could improve multi-thread performance if the socket supports it.

The 550 W PSU headroom is sufficient for the current configuration but may need upgrading if a more power-hungry GPU is installed later. The CPU's locked multiplier limits overclocking, so performance gains must come from memory tuning or component upgrades. The DDR5 memory support means users can upgrade to higher-capacity or faster kits, though the 102.4 GB/s bandwidth is already respectable. The integrated Arc Xe-LPG Graphics 24EU provides a fallback if the discrete GPU fails or is removed.