SYSTEM ANALYZER

Rate My PC: AMD Ryzen 9 7945HX3D + Intel Arc B390

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

86 / 100
HIGH-END

Power Build

Top 14% of systems. Excellent for 1440p Ultra or 4K High gaming.

1440p Ultra4K High

System Balance Analysis

CPU vs GPU performance ratio
Well Balanced
CPU
95%
VS
GPU
77%
PROCESSOR

AMD Ryzen 9 7945HX3D

59,641 Benchmark Score
Top 5% Market Ranking
View Full Specs →
GRAPHICS CARD

Intel Arc B390

1,482 Benchmark Score
Top 23% 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

Strong Performance

Excellent for 1440p gaming. Most games will run at high/ultra settings smoothly.

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 AMD Ryzen 9 7945HX3D paired with the Intel Arc B390 is, on paper, one of the most lopsided laptop combinations in the benchmark database. The CPU sits in the 92nd percentile of all processors measured, while the integrated GPU lands in the 9th percentile — a gap so wide that the combined percentile of 51 tells you almost nothing about how this machine actually behaves. One important caveat up front: the FACT PACK contains no measuredFps data for this exact pairing, so every frame-rate statement below is an estimate extrapolated from the benchmark scores rather than a recorded result.

GPU Analysis

The Intel Arc B390 is an integrated graphics solution built on Intel's Xe3-LPG architecture, fabricated on Intel's own 3 nm process as part of the Panther Lake family. Its core configuration is modest: 1536 shading units, 48 texture mapping units, 24 render output units, and 12 ray tracing cores. Peak compute lands at 7.680 TFLOPS FP32, with FP16 throughput doubling to 15.36 TFLOPS at a 2:1 ratio. Pixel fill rate is rated at 60.00 GPixel/s and texture fill rate at 120.0 GTexel/s.

Because this is an IGP, memory is system shared — there is no dedicated VRAM pool, no fixed bus width, and bandwidth is entirely dependent on the host system's RAM. That has real consequences for rendering performance. The CPU's dual-channel DDR5 interface provides 83.2 GB/s of theoretical bandwidth, and that single figure has to feed both the processor and the GPU. In any workload where the Arc B390 needs to pull large texture or geometry data — modern games at higher settings, viewport rendering in 3D applications — that shared bandwidth becomes the defining constraint. A discrete card with its own memory subsystem would not face this limitation.

Clock behavior is wide but unremarkable: a 300 MHz base and a 2500 MHz boost, with the boost figure only reachable within the GPU's 80 W thermal envelope. That power figure matters in a laptop chassis, because sustained GPU load will compete with the 16-core CPU for the same cooling capacity and the same power budget.

The single recorded benchmark score is 1482 in 3DMark Steel Nomad (DX12). Context makes that number meaningful: the B390's nearest rivals are the NVIDIA GeForce GT 520MX at 1463 (the B390 leads by 1.3%), the NVIDIA GeForce 800M at 1460 (1.5% ahead), the GT 625 OEM at 1446 (2.5% ahead), and the GT 710 at 1443 (2.7% ahead). These are entry-level parts from much older generations. The percentile placement — 9th against all GPUs in the database — confirms the B390 belongs to the bottom decile of graphics performance. For rendering work, that means the 12 RT cores exist on the spec sheet but cannot be leveraged into meaningful ray tracing throughput; the card's DirectX 12 Ultimate (12_2), Vulkan 1.4, and OpenGL 4.6 API support is complete, but the hardware behind it is thin. Any rendering estimate derived from the Steel Nomad score points toward low-resolution, low-settings gaming and basic GPU compute only.

Benchmark Performance

The CPU side is where this build earns its identity. The Ryzen 9 7945HX3D posts a Cinebench R23 multi-core score of 32782 and a single-core score of 1935. In Cinebench R15 it scores 5600 multi-core and 311 single-core. Geekbench returns 16758 multi-core and 2769 single-core. PassMark multithread reaches 57476, with a single-thread score of 4050. The 3DMark CPU profile scales sensibly across thread counts: 1071 single-thread, 2095 at two threads, 3937 at four, 7330 at eight, 12459 at sixteen threads, and 12936 at maximum threads.

That scaling curve is worth reading carefully. Going from one thread to eight yields nearly a sevenfold gain, but going from sixteen threads to the full thirty-two adds only 477 points — under 4% more. The 3D V-Cache slice is the distinguishing hardware feature here (a 64 MB slice stacked onto the base 128 MB of shared L3), and it is the reason this chip carries the X3D badge. In thread-limited workloads that fit in the enlarged cache, the chip behaves like a much narrower but faster part; in heavily threaded throughput workloads, the scaling story is standard.

The average benchmark score of 59641 places this CPU at the 92nd percentile of all CPUs. Its nearest rivals are all within roughly a percentage point: the Intel Core i9-14900F averages 60008 (this CPU is 0.6% behind), the AMD Ryzen 9 7945HX averages 60099 (0.8% behind), the AMD Ryzen 7 8745HX averages 60104 (0.8% behind), and the AMD Ryzen 5 PRO 9645 averages 58916 (this CPU is 1.2% ahead). In practical terms, the 7945HX3D trades blows with desktop-class and mobile-flagship silicon at the very top of the chart.

The GPU's average score of 1482 against a database-wide 9th percentile completes the picture. The combined percentile of 51 is an arithmetic artifact of averaging a 92nd-percentile CPU with a 9th-percentile GPU — it does not describe any real workload. No workload exists where this machine performs like a 51st-percentile system. CPU-bound tasks run like a top-tier workstation; GPU-bound tasks run like an entry-level laptop from a much weaker era.

FAQ

Q: Is this combination good for gaming?

A: Not for modern gaming at high settings. The Arc B390's Steel Nomad score of 1482 and 9th-percentile GPU placement put it roughly on par with a GT 710-class part — about 2.7% ahead of that card. Esports titles at reduced settings are the realistic ceiling. Note again that no measured FPS data exists for this pairing, so all frame-rate expectations are estimates from these scores.

Q: How strong is the CPU compared to the market?

A: Very strong. It sits in the 92nd percentile of all CPUs, within 1% of the Intel Core i9-14900F (0.6% behind) and effectively tied with the Ryzen 9 7945HX (0.8% behind) and Ryzen 7 8745HX (0.8% behind).

Q: What does the 3DMark CPU thread scaling show?

A: Scores run 1071 (single), 2095 (two threads), 3937 (four), 7330 (eight), 12459 (sixteen), and 12936 (max threads). Scaling is near-linear through eight threads and flattens sharply beyond sixteen.

Q: Does the GPU have dedicated VRAM?

A: No. Memory size, type, bus width, and bandwidth are all system shared. It draws from the same dual-channel DDR5 pool that feeds the CPU, limited to 83.2 GB/s of total bandwidth.

Q: Can this system do ray tracing?

A: It has 12 RT cores and DirectX 12 Ultimate support, but at 7.680 TFLOPS FP32 and a 9th-percentile overall placement, real ray tracing throughput will be minimal. The API support is present; the performance is not.

Q: Is the CPU good for video encoding and compression work?

A: Yes. PassMark data compression scores 699988, encryption 42367, integer math 201217, and floating point math 121619 — numbers consistent with its 92nd-percentile standing.

Q: Is this a desktop or laptop build?

A: It is classified as a laptop build, which fits both parts: the 7945HX3D is a mobile-segment Dragon Range processor on Socket FL1, and the B390 is an IGP.

Balance and Bottleneck

The bottleneck analysis here is unusually clean-cut: the GPU bottlenecks everything visual, and the CPU bottlenecks nothing. When a GPU sits in the 9th percentile and the CPU sits in the 92nd, no realistic workload exists where the graphics side keeps pace with the processor.

In gaming terms, the CPU's 3DMark max-threads score of 12936 and its physics score of 4321 indicate enormous frame-delivery headroom — headroom that the B390's 1482 Steel Nomad score cannot use. Estimated frame rates in any modern title will be GPU-limited at every resolution, and increasing resolution or quality settings will barely move the estimate because the GPU is already saturated. The one scenario where this inverts is extremely old or extremely lightweight games running at low resolution, where even the B390 can push high frame estimates and the CPU's 4050 PassMark single-thread score takes over as the limiter — but at that point the hardware is being wasted.

For productivity, the picture reverses. Cinebench R23 multi-core at 32782, Geekbench multi-core at 16758, and PassMark multithread at 57476 describe a machine that chews through compilation, encoding, and multi-threaded rendering without waiting on anything. The GPU contributes essentially nothing to these workloads, and the shared-memory design means GPU activity can actually steal bandwidth from CPU tasks — a rare case where the weak GPU is not merely absent but mildly counterproductive under mixed load.

The memory subsystem deserves mention as a secondary constraint. Dual-channel DDR5 at 83.2 GB/s is adequate for the CPU alone at this class, but when the IGP borrows from the same pool, effective bandwidth for both drops. Any workload that loads CPU and GPU simultaneously — streaming with hardware encoding, GPU-accelerated video effects — will feel this pinch first.

Usage Scenarios

High-refresh gaming: Not viable. A GPU in the 9th percentile with rivals like the GT 710 cannot produce high-refresh estimates in modern titles at any settings. Competitive esports at low settings and modest resolutions are the only realistic gaming target, and even those estimates come without measured FPS confirmation for this pairing.

Streaming: Partially viable. The 57476 PassMark multithread score and 32 threads handle encoding, compositing, and chat overlays with ease, but game capture will strain the shared memory pool, and the GPU cannot render a demanding game while assisting the stream. Streaming lightweight titles works; streaming AAA games does not.

Video editing: Strong on the CPU side. The 32782 Cinebench R23 multi-core score makes timeline scrubbing, proxies, and software encoding fast. GPU-accelerated effects and exports will disappoint, given the 7.680 TFLOPS FP32 figure and system-shared memory. A codec-driven, CPU-based editing workflow is the right fit.

3D rendering: CPU rendering is excellent — 12936 in 3DMark max threads and a 92nd percentile standing mean CPU-based render engines finish scenes quickly. GPU path tracing is effectively off the table with 12 RT cores at this performance tier. Viewport work will be the daily friction point.

Software development: Ideal. The 16758 Geekbench multi-core and 2769 single-core scores mean both large parallel builds and snappy single-threaded tooling. The 446 find-prime-numbers score and 83919 string-sorting score confirm strong general-purpose throughput. The GPU is irrelevant here.

Student and office work: Massively overprovisioned on compute. Documents, browsing, and video calls need a fraction of this CPU's capability. The B390 handles display output and video playback adequately, and the integrated Radeon 610M on the CPU side exists as a fallback graphics option.

Who Should Build It

Realistically, nobody chooses this exact pairing — it is best understood as a database comparison case, since these parts come from different manufacturers and different platform generations. But the underlying performance profile maps to clear users. Developers and engineers who need a 92nd-percentile, 32-thread mobile CPU for compilation, simulation, or data work are the primary audience; the 16 cores, 128 MB L3 plus the 64 MB V-Cache slice, and ECC memory support make this a portable workstation-grade processor.

Data professionals benefit from the PassMark suite: 699988 in compression, 42367 in encryption, and 201217 in integer math describe a machine for on-the-go dataset work. Content creators whose workflows are CPU-render-centric can use this confidently; creators dependent on GPU acceleration cannot, given the 9th-percentile graphics placement.

Gamers should look elsewhere entirely — or accept that this configuration limits them to esports-class titles at entry settings. Students and office users get far more CPU than they need and exactly the modest graphics they need, which is functional but imbalanced. Small businesses needing ECC-capable mobile workstations for field compute would find the CPU specification compelling, with the GPU as an acceptable accessory.

CPU Analysis

The Ryzen 9 7945HX3D is a 16-core, 32-thread Zen 4 processor from the Dragon Range family, built on TSMC's 5 nm node. It runs a 2.30 GHz base clock with a 5.40 GHz boost, inside a 55 W TDP envelope, on Socket FL1. The die configuration is 2x 71 mm² with 17,840 million transistors. Cache is the headline: 64 KB L1 per core, 1 MB L2 per core, 128 MB shared L3, and — the defining feature — one 64 MB 3D V-Cache slice stacked on top.

That cache hierarchy is what separates this chip from its nearest rivals, and yet the rival data shows the advantage is small in aggregate: it trails the Ryzen 9 7945HX by only 0.8% on average score. The V-Cache benefit shows up in the single-thread numbers instead — 1071 in 3DMark single-thread and 4050 in PassMark single-thread — where cache-resident workloads run faster than raw clock speed alone would suggest. For real workloads, this means the chip excels at simulation loops, game logic, and any latency-sensitive task where working sets fit in the enlarged cache, while matching (not beating) its siblings in pure throughput.

The platform features are generous for a mobile part: 28 PCIe Gen 5 lanes from the CPU, DDR5 support with ECC, and an unlocked multiplier for users who want to tune within thermal limits. The integrated Radeon 610M handles basic display duties. PassMark physics at 4321 and extended instructions at 52310 round out a profile that is genuinely workstation-class in a mobile power envelope.

Build Overview

This is a laptop-class combination in the strict sense: the 7945HX3D is a mobile-segment processor, and the Arc B390 is an integrated GPU with an IGP bus interface and no power connectors. The pairing is unusual — an AMD CPU matched to an Intel iGPU is not a configuration that occurs on a single platform in practice — so this page should be read as a comparative analysis of two parts' benchmark records rather than a shippable system.

By tier, the split is extreme. The CPU at the 92nd percentile is top-decile hardware. The GPU at the 9th percentile is bottom-decile hardware. The combined percentile of 51 is the arithmetic mean of those extremes and should not be quoted as a performance indicator; no single workload blends them evenly. In practice, this configuration behaves as either a high-end mobile workstation (for CPU work) or an entry-level laptop (for graphics work), with nothing in between.

Upgrade Path and Platform

The CPU's platform anchors are Socket FL1, DDR5 memory on a dual-channel bus at 83.2 GB/s, ECC support, and 28 PCIe Gen 5 lanes from the CPU. The GPU, being integrated, offers no upgrade path of its own — its 80 W envelope, IGP slot width, and absent power connectors are fixed attributes of the silicon. There is no suggested PSU figure because there is no discrete power delivery to size.

That asymmetry defines the sensible upgrade direction: graphics. The only meaningful performance ceiling in this configuration is the GPU, sitting in the 9th percentile while the CPU idles at the 92nd. A discrete graphics solution — one that brings its own memory and does not share the 83.2 GB/s DDR5 pool — would transform the machine's gaming and rendering estimates. The CPU, conversely, needs nothing; at 0.6% behind a Core i9-14900F, there is no headroom worth chasing within this class.

On the CPU side, the 55 W TDP and unlocked multiplier leave modest tuning room, though in a laptop chassis the shared cooling between a 55 W CPU and an 80 W iGP limits how much of that headroom is usable under combined load. Memory is the other practical lever: since the iGPU's bandwidth is system dependent, faster DDR5 directly improves estimated graphics performance, making memory the cheapest effective upgrade for this configuration's weakest link.