SYSTEM ANALYZER

Rate My PC: Intel Core Ultra 5 235A + NVIDIA GeForce RTX 5090 D

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

96 / 100
ULTIMATE READY

Apex Performer

Top 4% 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
98%
PROCESSOR

Intel Core Ultra 5 235A

48,201 Benchmark Score
Top 6% 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

# FAQ

Q: What is the Intel Core Ultra 5 235A and where does it sit in the CPU hierarchy?

A: The Intel Core Ultra 5 235A is a desktop processor from the Core Ultra Series 2, built on the Arrow Lake architecture (Arrow Lake-S) using a 3 nm process from TSMC. It scores in the 90th percentile against all CPUs, with an average benchmark score of 48,201. Its nearest rival is the AMD Ryzen AI Max PRO 380, which scores 48,171 — a delta of just 0.1%, meaning these two processors are statistically tied in overall performance.

Q: What are the core and thread counts of this CPU, and does it use hyper-threading?

A: The Core Ultra 5 235A has 14 cores and 14 threads, which means it does not employ simultaneous multithreading (SMT) — each core handles exactly one thread. This is significant because many competing 14-core processors offer 20 or more threads. The base clock is 3.40 GHz with a boost clock of 5.00 GHz, and the CPU has a 65 W TDP.

Q: How does the RTX 5090 D compare to its nearest GPU rivals?

A: The NVIDIA GeForce RTX 5090 D sits in the 92nd percentile among all GPUs, with an average benchmark score of 77,712. Its closest competitor is the AMD Radeon RX 6650M XT, which scores 76,904 — just 1.1% behind. The RX 6850M XT is 1.6% ahead at 78,940, placing the 5090 D in a tightly contested mid-upper tier band despite its massive 32 GB memory configuration.

Q: What memory and PCIe support does the platform offer?

A: The CPU supports dual-channel DDR5 memory with a bandwidth of 102.4 GB/s, and it does not support ECC memory. The processor provides 20 PCIe Gen 5 lanes from the CPU itself. The GPU connects via PCIe 5.0 x16, fully compatible with the CPU's lane allocation.

Q: Is the RTX 5090 D a desktop or laptop GPU?

A: The RTX 5090 D is a desktop-class graphics card, as confirmed by the build class of this analysis being "desktop." The card is dual-slot, measures 304 mm in length, 137 mm in height, and 48 mm in width, and uses a single 16-pin power connector.

Q: What is the integrated graphics capability of the CPU?

A: The Core Ultra 5 235A includes an integrated GPU: Arc Xe-LPG Graphics with 24 execution units. This provides basic display output and light graphical tasks, but for any demanding gaming or rendering workload, the discrete RTX 5090 D is the primary graphics solution.

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

A: The Intel Core Ultra 5 235A has a launch MSRP of $269, and the NVIDIA GeForce RTX 5090 D has a launch MSRP of 2,299 USD. Both products are currently marked as "Active" in production status.

---

Benchmark Performance

The combined picture for this CPU+GPU pairing places it at the 91st percentile among all desktop builds, indicating a very high-tier system overall. The CPU alone achieves a 90th percentile rank among all CPUs with an average benchmark score of 48,201, while the GPU reaches the 92nd percentile among all GPUs with an average score of 77,712.

The CPU's Cinebench results reveal strong multi-core capability: it scores 3,289 in Cinebench R15 multi-core, 13,705 in R20 multi-core, and 32,633 in R23 multi-core. Single-core performance is equally robust, with scores of 464 in R15, 1,934 in R20, and 4,607 in R23. These numbers place the 235A in a competitive position against its nearest rivals, all of which cluster within a narrow band: the AMD Ryzen AI Max PRO 380 (48,171, +0.1% delta), the Intel Core Ultra 5 245HX (48,287, -0.2%), the AMD EPYC 4345P (48,470, -0.6%), and the AMD Ryzen 9 3900 (47,918, +0.6%). The data shows that the 235A is essentially performance-equivalent to these four alternatives, with deltas under 1% in every case.

The GPU side is dominated by the RTX 5090 D's sheer compute throughput. In 3DMark Steel Nomad DX12, it scores 14,326. Geekbench OpenCL returns 310,674 and Vulkan 376,915. Passmark G3D yields 44,065 with GPU compute at 28,396. These figures indicate a graphics card that excels in both rasterization and compute-heavy workloads. The nearest GPU rivals — RX 6650M XT (76,904, +1.1%), RX 6850M XT (78,940, -1.6%), Tesla P100 12 GB (79,396, -2.1%), and Tesla P100 16 GB (79,605, -2.4%) — are all within 2.4% of the 5090 D's average score, suggesting that despite its massive memory and modern architecture, the 5090 D performs in a similar aggregate tier to those alternatives.

Combined, the 91st percentile build rank indicates that this system will handle almost any software workload with headroom. The CPU and GPU are both upper-decile performers, and their pairing creates a balanced high-end platform.

---

CPU Analysis

The Intel Core Ultra 5 235A is built on Arrow Lake-S, a 3 nm TSMC process with 17,800 million transistors on a 243 mm² die. The architecture is monolithic, with 14 cores and 14 threads — no hyper-threading. Each core has 192 KB of L1 cache and 3 MB of L2 cache, with a shared 24 MB L3 cache. The base clock of 3.40 GHz boosts to 5.00 GHz, and the multiplier is locked, meaning no overclocking via multiplier adjustment.

The Cinebench R23 multi-core score of 32,633 is 0.2% behind the Intel Core Ultra 5 245HX's average benchmark score of 48,287, but in raw Cinebench terms, the 235A delivers strong multi-threaded throughput. The Passmark multi-thread score of 38,392, combined with single-thread at 4,557, shows that the 235A handles both heavily parallel and latency-sensitive workloads well. Data compression in Passmark scores 393,800, encryption 30,136, extended instructions 31,625, and integer math 88,626. These numbers indicate a processor that is well-suited for database operations, scientific computing, and encryption-heavy tasks.

The 14-thread configuration is a deliberate design choice: Arrow Lake's performance cores are designed for high IPC, and the 5.00 GHz boost clock compensates for the lack of SMT in many workloads. In single-threaded tasks, the 4,607 R23 score puts it in the upper echelon of desktop CPUs. For real-world workloads, this means snappy application launches, responsive compile times, and strong performance in lightly-threaded software like office suites and web browsers. The 90th percentile ranking confirms that this CPU outperforms 90% of all processors in the benchmark database, making it a high-tier choice for desktop builds.

The memory bandwidth of 102.4 GB/s over dual-channel DDR5 is adequate for the 14-core design but could bottleneck memory-intensive workloads like large database queries or 3D scene rendering with massive texture sets. The lack of ECC support limits its use in mission-critical server roles, but for a desktop workstation, this is rarely a concern.

---

Gaming Performance

No measured FPS rows exist for this exact CPU+GPU combination. The FACT PACK contains no measuredFps data for the Core Ultra 5 235A paired with the RTX 5090 D. Therefore, all FPS figures discussed here are estimated from the benchmark scores and should be treated as approximate expectations, not empirically measured results.

The RTX 5090 D's Passmark G3D score of 44,065 and 3DMark Steel Nomad DX12 score of 14,326 suggest very high rasterization performance. Given the GPU's 104.8 TFLOPS FP32 throughput and 32 GB of GDDR7 memory with 1.79 TB/s bandwidth, frame rates at 1080p and 1440p should be extremely high — likely exceeding 144 FPS in most modern titles at ultra settings, based on the percentile rank of 92nd among all GPUs. At 4K, the 5090 D has the memory capacity and bandwidth to maintain high frame rates, though the CPU may become a limiting factor in lightly-threaded games due to its 14-thread ceiling.

The CPU's single-thread score of 4,607 in Cinebench R23 indicates strong gaming IPC, which is critical for frame pacing. The 5.00 GHz boost clock further supports high-refresh gaming. However, in games that scale well beyond 14 threads, the lack of SMT could cap performance — modern titles like Cyberpunk 2077 or Starfield can use 16+ threads, and the 235A may see diminishing returns compared to 20-thread rivals. The 91st combined percentile suggests that this pairing will deliver excellent gaming performance, but the GPU is likely the dominant factor in most scenarios, with the CPU providing sufficient headroom for 60+ FPS at 4K ultra in most games.

For esports titles at 1080p, the estimated FPS could exceed 200 in games like CS2 or Valorant, given the GPU's raw throughput and the CPU's high single-core clocks. At 1440p, expect 144+ FPS in competitive shooters and 60-100 FPS in AAA single-player games at ultra settings. At 4K, the 32 GB VRAM ensures no texture-related stutters, but frame rates will depend on the game's engine and CPU scaling.

---

Who Should Build It

The target user for this Core Ultra 5 235A + RTX 5090 D build is a desktop enthusiast seeking high-end performance across multiple domains. The 90th percentile CPU and 92nd percentile GPU make this system suitable for:

  • Gamers at 1440p and 4K: The GPU's 32 GB VRAM and 1.79 TB/s bandwidth handle high-resolution textures and ray tracing without memory pressure. The CPU's single-core strength ensures high frame rates in CPU-bound titles.
  • Content creators: The CPU's Cinebench R23 multi-core of 32,633 and Passmark multi-thread of 38,392 support video encoding, 3D rendering, and batch processing. The GPU's 104.8 TFLOPS FP32 accelerates CUDA-based renders.
  • Software developers: With 14 threads and strong integer math (88,626 in Passmark), the 235A compiles code efficiently. The 24 MB L3 cache aids in large project builds.
  • Students and researchers: The 90th percentile CPU handles simulation software, data analysis tools, and virtual machines, though the lack of ECC memory limits reliability-critical scientific computing.
  • Small business workstations: The 65 W TDP keeps power costs manageable, and the 32 GB GPU VRAM enables CAD, GIS, and machine learning inference workloads.

The build is not ideal for budget-conscious users or those needing 24/7 server-grade reliability, given the non-ECC memory and locked multiplier.

---

Upgrade Path and Platform

The platform is built on Intel Socket 1851, which supports DDR5 memory in a dual-channel configuration. The CPU provides 20 PCIe Gen 5 lanes, and the GPU uses PCIe 5.0 x16, leaving 4 lanes for additional devices — enough for a Gen 5 NVMe SSD. The 65 W TDP CPU is modest, but the GPU's 575 W TDP and 950 W suggested PSU mean the system requires a robust power supply. The single 16-pin power connector on the GPU simplifies cabling.

A sensible next upgrade would be adding more DDR5 memory, as the 102.4 GB/s bandwidth may become a bottleneck in memory-heavy workloads. The 24 MB L3 cache is fixed, so the only cache upgrade would be a CPU swap to a higher-tier Arrow Lake processor with more L3, but the socket 1851 platform's longevity depends on Intel's future support. The locked multiplier prevents CPU overclocking, but the GPU can be overclocked via software tools. The 950 W PSU leaves headroom for adding more storage drives or a second GPU, though PCIe lane constraints limit multi-GPU setups.

The 3 nm process and 17,800 million transistors on the CPU suggest a modern, efficient design, but the 575 W GPU TDP dominates the system's power profile. Users upgrading from a lower-tier GPU will need to verify their PSU meets the 950 W recommendation.

---

Balance and Bottleneck

This pairing is well-balanced overall, with the CPU at the 90th percentile and GPU at the 92nd percentile. However, the bottleneck direction changes by workload:

  • CPU-bound scenarios: In esports games at 1080p or lightly-threaded applications, the 14-thread ceiling may limit frame rates. The GPU's massive compute power (104.8 TFLOPS) will sit idle while the CPU's single-thread performance (4,607 Cinebench R23) becomes the limiting factor. The 0.2% performance delta to the Core Ultra 5 245HX indicates the 235A is not the fastest Arrow Lake part, so CPU-bound tasks will show some headroom loss.
  • GPU-bound scenarios: In 4K gaming or heavy rendering, the GPU is the primary driver. The 5090 D's 32 GB VRAM and 1.79 TB/s bandwidth ensure it is not memory-limited, but the CPU's 14 threads may struggle to feed the GPU in draw-call-heavy scenes, potentially reducing GPU utilization.
  • Mixed workloads: For video editing or 3D rendering, both components are active. The CPU's multi-core score of 32,633 in R23 handles scene setup and physics, while the GPU accelerates final rendering. The 91st combined percentile suggests these workloads will perform strongly.

The data implies the CPU is the more likely bottleneck in gaming, while the GPU is the bottleneck in compute-only tasks. The 1.1% delta to the RX 6650M XT in GPU benchmarks shows the 5090 D is not dramatically faster than that rival, so pairing it with a 90th percentile CPU is appropriate — a higher-tier CPU would offer diminishing returns.

---

Build Overview

This build pairs the Intel Core Ultra 5 235A (14-core, 14-thread Arrow Lake-S desktop CPU) with the NVIDIA GeForce RTX 5090 D (Blackwell 2.0 architecture GPU with 32 GB GDDR7). It is a desktop-class system, as indicated by the buildClass field. The combined percentile of 91 places this build in the top 9% of all systems in the benchmark database.

The CPU's 90th percentile rank and the GPU's 92nd percentile rank are closely matched, creating a cohesive platform. The CPU's average benchmark score of 48,201 is within 0.6% of its nearest rivals, and the GPU's 77,712 average is within 2.4% of its closest competitors. This means the build performs at a consistent high tier across both CPU and GPU workloads, with neither component dramatically outclassing the other.

The 65 W TDP CPU and 575 W TDP GPU create a power asymmetry, but the 950 W suggested PSU accommodates both. The system is production-active for both components, with the CPU released on 2025-07-28 and the GPU on 2025-01-29. The build is a modern, high-end desktop workstation/gaming hybrid.

---

GPU Analysis

The NVIDIA GeForce RTX 5090 D is built on the Blackwell 2.0 architecture with the GB202 chip, manufactured on a 5 nm TSMC process with 92,200 million transistors on a 750 mm² die. It has 32 GB of GDDR7 memory on a 512-bit bus, delivering 1.79 TB/s of bandwidth. The base clock is 2017 MHz with a boost of 2407 MHz, and memory runs at 1750 MHz (28 Gbps effective).

Compute resources are substantial: 21,760 shading units, 680 TMUs, 176 ROPs, 170 RT cores, and 680 tensor cores. The pixel rate is 423.6 GPixel/s and texture rate is 1,636.8 GTexel/s. FP32 throughput is 104.8 TFLOPS, with FP16 at 104.8 TFLOPS (1:1 ratio). This makes the 5090 D a rendering powerhouse — the FP32 figure alone suggests it can handle complex 3D scenes, scientific simulations, and AI inference with ease.

In benchmarks, the 3DMark Steel Nomad DX12 score of 14,326 and Geekbench OpenCL of 310,674 indicate strong rasterization and compute. The Vulkan score of 376,915 is notably higher than OpenCL, suggesting excellent driver optimization for Vulkan-based games and applications. Passmark G3D at 44,065 and GPU compute at 28,396 confirm solid all-around performance, though the nearest rivals (RX 6650M XT at +1.1% and RX 6850M XT at -1.6%) show that aggregate scores cluster tightly.

The 170 RT cores enable hardware-accelerated ray tracing, and the 680 tensor cores accelerate DLSS and AI workloads. The 32 GB VRAM is double the typical 16 GB found in many high-end cards, making this GPU particularly suited for 8K texture packs, large machine learning models, and multi-GPU rendering scenes. The 512-bit memory bus and 1.79 TB/s bandwidth ensure that memory-bound workloads like 4K video editing or scientific visualization do not stall.

The 575 W TDP is substantial, requiring the 950 W suggested PSU. The dual-slot design and 304 mm length mean case compatibility must be verified. Display outputs include 1x HDMI 2.1b and 3x DisplayPort 2.1b, supporting high-refresh monitors. API support includes DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, ensuring future software compatibility.

---

Usage Scenarios

High-refresh gaming: The GPU's 104.8 TFLOPS and 32 GB VRAM push frame rates well beyond 144 FPS at 1080p and 1440p in most titles, with the CPU's 5.00 GHz boost clock maintaining high minimum frames. At 4K, expect 60-100 FPS in AAA games, with ray tracing enabled thanks to the 170 RT cores. Esports titles at 1080p could exceed 200 FPS, though the CPU's 14-thread limit may cap performance in the most demanding scenes.

Streaming: The 14-core CPU handles encoding workloads, with the Passmark extended instructions score of 31,625 supporting AV1 and HEVC encoding. The GPU's tensor cores can also offload encoding via NVENC, freeing CPU resources for gameplay. The 32 GB VRAM prevents memory pressure during simultaneous game capture and encoding.

Video editing: The CPU's Cinebench R23 multi-core of 32,633 handles timeline decoding and effects, while the GPU's 104.8 TFLOPS accelerates exports and GPU-accelerated effects. The 1.79 TB/s bandwidth ensures smooth 4K multi-stream playback. The 92nd GPU percentile indicates this is a top-tier editing workstation.

3D rendering: The GPU's 170 RT cores and 32 GB VRAM enable GPU-accelerated ray tracing renders in software like Blender or Octane. The FP32 throughput of 104.8 TFLOPS dramatically reduces render times compared to lower-tier GPUs. The CPU's multi-core score supports scene simulation and particle effects.

Software development: The 14 threads and 24 MB L3 cache speed up compilation, with Passmark integer math of 88,626 indicating strong arithmetic throughput. The 5.00 GHz boost clock accelerates single-threaded build steps. The 90th CPU percentile means this handles large codebases efficiently, though developers needing massive parallel builds may want more threads.

Student and office work: The CPU's single-thread score of 4,607 in R23 makes everyday applications like office suites and web browsing feel instant. The integrated Arc Xe-LPG Graphics (24 EU) can handle dual displays without the discrete GPU, saving power. The 65 W TDP keeps the system quiet and energy-efficient for dorm or office environments, while the GPU provides headroom for occasional rendering or gaming tasks.