SYSTEM ANALYZER

Rate My PC: Intel Core i7-13700T + NVIDIA GeForce RTX 5090

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

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
91%
VS
GPU
98%
PROCESSOR

Intel Core i7-13700T

35,403 Benchmark Score
Top 9% Market Ranking
View Full Specs →
GRAPHICS CARD

NVIDIA GeForce RTX 5090

79,842 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

The Intel Core i7-13700T paired with the NVIDIA GeForce RTX 5090 is an unusual desktop combination: a 35 W thermal-design-power CPU from Intel's 13th Gen Raptor Lake family sitting alongside the most powerful GPU in NVIDIA's GeForce 50-series Blackwell lineup. The data shows a combined percentile of 89 across all tracked builds, meaning this pairing outperforms roughly nine in ten systems in the database — but the split is lopsided. The GPU sits in the 92nd percentile of all GPUs while the CPU sits in the 85th percentile of all CPUs, and the gap in raw compute capability is far wider than those percentiles suggest. One important caveat applies throughout this page: no measured FPS rows exist for this exact combination in the database, so every frame-rate discussion below is estimated from the benchmark scores rather than directly measured.

Upgrade Path and Platform

The i7-13700T uses Intel Socket 1700, which gives the platform a clear and practical upgrade path within Intel's desktop lines that share that socket. The chip is a 16-core, 24-thread part built on Intel's 10 nm process with a die size of 257 mm², and it supports both DDR4 and DDR5 memory on a dual-channel bus — a useful flexibility, since a builder can carry over older DDR4 kits or move to DDR5 without changing platforms. ECC memory is not supported, which matters for anyone considering this as a workstation base.

PCIe connectivity is Gen 5 with 20 lanes from the CPU, which is exactly what the RTX 5090 wants. The GPU uses a PCIe 5.0 x16 bus interface, so it runs at full speed with headroom left over for a fast Gen 5 NVMe drive on the remaining lanes. That is a genuine strength of this pairing: the platform does not bottleneck the card's interface.

Power is the interesting part. The CPU's TDP is just 35 W, while the RTX 5090 carries a 575 W TDP and NVIDIA's suggested PSU rating is 950 W. In practice the GPU dominates the power budget almost entirely — the processor is a rounding error next to the card. A builder speccing this system should follow the 950 W suggested PSU figure, because the CPU contributes almost nothing to total system draw. That asymmetry also has a cooling implication: the CPU can run under a capable air cooler without strain given its 35 W rating, while the GPU is a dual-slot card measuring 304 mm in length, 137 mm in height, and 40 mm in width, drawing power through a single 16-pin connector. Case clearance for the card's length is the real constraint, not CPU thermals.

A sensible next upgrade is straightforward: replace the 13700T with a higher-TDP Socket 1700 processor. The chip has an unlocked multiplier, though the thermal envelope limits how far that can be pushed, and the benchmark data shows the i7-12700K scoring within 0.3 percent of it on average — evidence that a standard desktop chip in this family delivers similar aggregate performance with far more thermal headroom. On the GPU side, the fact pack lists the GeForce 60-series as the successor line, so a future card upgrade within NVIDIA's roadmap is the natural long-term move.

GPU Analysis

The RTX 5090 is the flagship of the GeForce 50-series, built on the GB202 chip with TSMC's 5 nm process. The silicon is enormous by any standard: 92,200 million transistors on a 750 mm² die, giving a transistor density of 122.9 million transistors per square millimetre. The architecture is Blackwell 2.0.

The memory subsystem is the headline. The card carries 32 GB of GDDR7 on a 512-bit bus, running at 1750 MHz with 28 Gbps effective speed, producing 1.79 TB/s of memory bandwidth. That combination of capacity and bandwidth is what makes the card viable for professional rendering and large-scene work, not just gaming — 32 GB of VRAM means huge textures, high-resolution timeline scrubbing, and large 3D scenes fit without spilling to system RAM.

Compute resources are equally extreme: 21,760 shading units, 680 texture mapping units, 176 render output units, 170 RT cores, and 680 tensor cores. The pixel rate is 423.6 GPixel/s and the texture rate is 1,636.8 GTexel/s. Raw FP32 throughput is 104.8 TFLOPS, with FP16 running at a 1:1 ratio for the same 104.8 TFLOPS figure. The 170 dedicated RT cores matter for ray-traced rendering workloads, and the 680 tensor cores drive AI-accelerated features in both render engines and upscaling technologies.

Clocks are a 2017 MHz base and a 2407 MHz boost. Display output consists of one HDMI 2.1b and three DisplayPort 2.1b connectors, and the card supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

What do the benchmark scores mean for rendering? The Geekbench OpenCL score of 334,370 and Vulkan score of 376,728 indicate very strong GPU compute throughput, and the 3DMark Steel Nomad DX12 score of 18,355 confirms top-tier rasterization and DX12 performance. The PassMark G3D score of 39,650 and GPU compute score of 26,756 round out a picture of a card in the 92nd percentile of all GPUs, with an average benchmark score of 79,842. The database's nearest-rivals list puts it within 0.3 percent of the NVIDIA Tesla P100 PCIe 16 GB, 0.6 percent ahead of the Tesla P100 PCIe 12 GB, 1.1 percent ahead of the AMD Radeon RX 6850M XT, and 1.4 percent behind the AMD Radeon Pro Vega 64X on aggregate average score — a reminder that aggregate averaging can group it with very different classes of hardware, but the percentile and individual scores tell the real story.

Benchmark Performance

Starting with the CPU: Cinebench R23 multi-core comes in at 23,482 with a single-core score of 3,282. Cinebench R20 shows 9,764 multi-core and 1,378 single-core, and Cinebench R15 shows 2,343 multi-core and 330 single-core. Geekbench lands at 11,573 multi-core and 2,490 single-core, and PassMark single-thread scores 3,821 with a multi-thread result of 28,211. The specialized PassMark subtests cover the practical bases: 105,397 in integer math, 73,928 in floating point math, 327,700 in data compression, 19,230 in encryption, 19,306 in extended instructions, 34,939 in random string sorting, and a physics score of 1,769. The find-prime-numbers result is 127. The CPU's average benchmark score across the database is 35,403, placing it in the 85th percentile of all CPUs.

The rival data contextualizes that. The Intel Core 5 213PE averages 35,428 — just 0.1 percent ahead. The i7-12700KF averages 35,365, the i5-13600T averages 35,305, and the i7-12700K averages 35,287, all within half a percent of the 13700T. In other words, this low-TDP chip performs in the same aggregate bracket as full-desktop 12th and 13th Gen parts, which is impressive for a 35 W processor — but it also means it is not a step above that class.

On the GPU side, the RTX 5090's numbers are detailed in the previous section: an average benchmark score of 79,842, the 92nd percentile, and individual results spanning 3DMark, Geekbench compute, and the PassMark DirectX suites (395 in DX9, 341 in DX11, 226 in DX10, 185 in DX12), plus 1,413 in G2D.

The combined picture is a build in the 89th percentile overall. The GPU is clearly the stronger half, and the practical implication is that GPU-bound workloads — rendering, GPU compute, high-resolution gaming — will flourish, while CPU-bound stages of a pipeline will run at a good-but-not-flagship level.

Who Should Build It

This pairing fits a specific kind of user. The obvious candidate is a content creator or 3D artist who needs the 32 GB of GDDR7 and 104.8 TFLOPS of FP32 for rendering and simulation but does not need a flagship CPU to feed it. The i7-13700T's 16 cores and 24 threads with a Cinebench R23 multi-core score of 23,482 are enough for timeline editing, scene assembly, and light simulation, and the 35 W TDP makes for a quiet, compact machine.

Gamers at 4K are the second group. Because there are no measured FPS rows for this combination, frame rates must be estimated from the benchmark scores — and with the GPU in the 92nd percentile, 4K gaming is where the card's 423.6 GPixel/s pixel rate and 1.79 TB/s of bandwidth are best utilized, and where the CPU is least likely to be the limiter. High-refresh-rate 1080p or 1440p gaming is a weaker fit: at those resolutions the CPU typically becomes the constraint, and an 85th-percentile processor with a 35 W power limit leaves performance on the table.

Developers and students benefit differently. The 24 threads handle parallel compilation well, and the 680 tensor cores plus 32 GB of VRAM make this a serious platform for local AI and machine learning work that would overwhelm lesser cards. Small business workstations doing video, visualization, or GPU compute get a capable, quiet base — with the caveat that ECC memory is not supported, which rules out workloads that demand memory error correction.

CPU Analysis

The i7-13700T is a 16-core, 24-thread Raptor Lake-S processor on Intel's 10 nm process, released on January 3, 2023. Its base clock is 1400 MHz — a very low number reflecting the power-constrained design — but it boosts to 4.90 GHz, which explains the strong single-core results: a PassMark single-thread score of 3,821 and Cinebench R23 single-core of 3,282 put responsiveness and lightly-threaded tasks squarely in mainstream-flagship territory.

The cache layout supports that: 80 KB of L1 per core, 2 MB of L2 per core, and 30 MB of shared L3. Sixteen cores with 24 threads means a hybrid arrangement of performance and efficiency cores, and the multi-core scores reflect it — 23,482 in Cinebench R23 and 28,211 in PassMark multithread are strong numbers for a 35 W part. The multiplier is unlocked, giving tuners room to experiment, though the thermal envelope constrains how much is practical.

For real workloads, the PassMark subtests map directly onto everyday tasks. The 327,700 data compression score indicates fast archive handling; the 19,230 encryption score covers disk and file encryption; the 34,939 random string sorting result reflects text and database-adjacent performance; and the 73,928 floating point score with 19,306 in extended instructions covers scientific and media workloads. The 1,769 physics score is relevant for game physics and simulation. In aggregate, this is a CPU that behaves like a standard i7-class desktop processor of its generation despite the low power rating — the rival data confirms it trades blows within half a percent of the i7-12700K and i7-12700KF on average.

FAQ

Q: Does this pairing have measured frame-rate data?

A: No. The database contains no measured FPS rows for this exact combination, so all gaming performance discussion on this page is estimated from the CPU and GPU benchmark scores, not direct measurement.

Q: What PSU should this build use?

A: Follow NVIDIA's suggested PSU rating of 950 W. The GPU's TDP is 575 W delivered through a single 16-pin connector, while the CPU adds only a 35 W TDP, so the GPU dictates the power budget.

Q: Can the CPU be upgraded later without changing motherboards?

A: Yes. The i7-13700T uses Intel Socket 1700, so other processors on that socket are drop-in upgrade candidates. Notably, the i7-12700K scores within 0.3 percent of this chip on average, with much higher thermal headroom.

Q: Is 32 GB of VRAM actually useful?

A: For this card's target workloads, yes. The 32 GB of GDDR7 on a 512-bit bus with 1.79 TB/s of bandwidth supports large 3D scenes, high-resolution video timelines, and local AI models that smaller cards cannot hold in memory.

Q: Does the CPU support ECC memory?

A: No. The fact pack lists ECC memory support as false, so this platform is not suited to workloads requiring memory error correction.

Q: How does the RTX 5090 compare to its nearest rivals in the database?

A: On average benchmark score it sits within 0.3 percent of the Tesla P100 PCIe 16 GB, 0.6 percent ahead of the Tesla P100 PCIe 12 GB, 1.1 percent ahead of the Radeon RX 6850M XT, and 1.4 percent behind the Radeon Pro Vega 64X — while sitting in the 92nd percentile of all GPUs overall.

Q: Is the CPU overclockable?

A: The multiplier is unlocked, so tuning is possible. The 35 W TDP is the practical limiter on how much headroom is extractable.

Balance and Bottleneck

This is an unbalanced build, and deliberately so. The GPU sits in the 92nd percentile; the CPU sits in the 85th percentile. But percentiles understate the gap, because the 5090 is the top of its product stack while the 13700T is a power-limited variant in the middle-upper range of its own. The average benchmark scores make the asymmetry concrete: 79,842 for the GPU versus 35,403 for the CPU on their respective scales.

Where does that leave the bottleneck? In GPU-bound workloads — 4K gaming, ray-traced rendering, GPU compute via OpenCL or Vulkan — the 5090's 104.8 TFLOPS of FP32 and 1.79 TB/s of bandwidth are the engine, and the CPU simply dispatches work. The Cinebench and PassMark CPU scores are more than adequate there. In CPU-bound workloads, the picture reverses: high-refresh gaming at lower resolutions, heavy compilation, or multi-threaded simulation will be limited by a 35 W processor whose aggregate performance matches standard 12th Gen desktop chips but cannot exceed them — the rival deltas of under half a percent against the i7-12700K and i7-12700KF cap expectations.

Because no measured FPS rows exist for this combination, FPS scaling cannot be cited directly from the database; the estimate, based purely on the benchmark scores, is that GPU-bound scenarios will scale with the 5090's 92nd-percentile performance while CPU-bound scenarios will plateau at 85th-percentile levels. The 20 Gen 5 CPU PCIe lanes and PCIe 5.0 x16 GPU interface remove the bus from the equation entirely — the imbalance is purely about processing power, not connectivity.

Build Overview

This is a desktop build, classified as such in the database, combining a 13th Gen Intel Core i7 processor in its low-power "T" configuration with NVIDIA's flagship GeForce 50-series GPU. The pairing ranks in the 89th percentile of all tracked combinations overall — an elite aggregate figure driven almost entirely by the graphics card. The i7-13700T brings 16 cores, 24 threads, dual-channel DDR4/DDR5 support, and integrated UHD Graphics 770 (useful as a backup display output, though redundant here). The RTX 5090 brings Blackwell 2.0 silicon, 32 GB of GDDR7, and a 575 W TDP that dwarfs the CPU's 35 W. It is best understood as a rendering and GPU-compute workstation in a gaming card's clothing, with the CPU chosen for efficiency rather than maximum throughput.

Usage Scenarios

High-refresh gaming. Estimated from benchmark scores only, since no measured FPS data exists for this pair: at 1080p and 1440p with high refresh targets, the CPU becomes the limiting factor, and its 35 W TDP means it cannot exploit the 5090's full potential in CPU-bound titles. This is not the ideal build for competitive high-refresh play.

4K gaming. This is where the pairing makes sense. The 92nd-percentile GPU with 1.79 TB/s of bandwidth and a 423.6 GPixel/s pixel rate is built for high-resolution output, and at 4K the load shifts onto the graphics card, where the 5090's Steel Nomad score of 18,355 indicates top-tier rasterization performance.

Streaming. The 24 threads and 28,211 PassMark multithread score can handle encoding alongside gameplay, and the 327,700 data compression score helps with recording and file management. The imbalance works in the streamer's favour at 4K output resolutions, less so at 1080p60 where CPU headroom matters more.

Video editing. Strong fit. The 32 GB of GDDR7 accommodates high-resolution timelines, the GPU compute scores (334,370 OpenCL, 376,728 Vulkan) accelerate effects and export stages, and the CPU's 23,482 Cinebench R23 multi-core score handles timeline scrubbing and proxy generation comfortably.

3D rendering. The best-case scenario for this build. The 104.8 TFLOPS of FP32, 170 RT cores for ray-traced rendering, and massive VRAM mean GPU render engines will run at full speed, and the CPU's role in scene preparation is adequately covered by its multi-core results.

Software development. The 24 threads and strong single-core results — 3,821 PassMark single-thread, 2,490 Geekbench single-core — handle compilation and IDE workloads well, and the 680 tensor cores plus 32 GB of VRAM make local AI-assisted development tooling and model experimentation practical on one machine.

Student and office work. Overkill for documents and browsing, but the low 35 W CPU TDP makes for a quiet everyday machine, and the same hardware doubles as a capable research and content-creation platform — the PassMark string sorting (34,939) and integer math (105,397) scores indicate snappy general responsiveness.