Intel Core Ultra 9 285K vs Intel Core i9-14900K vs AMD Ryzen 9 9950X Comparison

On the 10th October 2024, Intel announced the launch of their next-gen desktop processors – Core Ultra (Series 2). It received somewhat of a mixed reaction and the reception was only lukewarm. This was mainly due to the fact that Intel themselves stated in their presentation slides, that the gaming performance of their flagship Core Ultra 9 285K was in fact, worse than its predecessor the Core i9-14900K! Oh dear … what a disappointing launch for Intel.

More worrying, is that the new Core Ultra 9 285K now only has 24 threads, compared to 32 threads on the older 14900K, and offers a lower max boost speed of only 5.7GHz (6.0GHz on the 14900K). It also has a new pin configuration and that corresponds to a new LGA1851 socket … in other words, if you want to run one of these new Core Ultra CPUs from Intel, you’ll need a new motherboard that has the new Z890 chipset. Check out the comparison table between the Core Ultra 9 285K vs the Core i9-14900K taken from Intel.

So what real advantages does the new Core Ultra (Series 2) have over previous Core i-Series of processors? Well first off, the new chip is now made by TSMC (not Intel) using the 3nm fabrication process, plus it’s now more energy efficient and consumes less power. I guess lower power consumption means less heat, so now the CPU should run a little cooler.

You can’t make a new chip nowadays without integrating some sort of AI. The new Core Ultra CPUs now includes Intel’s AI Boost NPU (neural processing unit) with a performance of 13 TOPS  (Trillions or Tera Operations per Second). I’m not exactly sure how these NPUs will help with everyday tasks… but there you go, you now got some NPUs in your CPUs 😀

For our analysis and verdict of the AMD Ryzen 9 9950X go here.

Anyway, let’s get back to our comparison between the Intel Core Ultra 9 285K vs Intel Core i9-14900K vs AMD Ryzen 9 9950X. We’re going to do a quick comparison of the performance of these processors. So let’s begin …

Quick Comparison

	Core Ultra 9 285K	Core i9-14900K	Ryzen 9 9950X
Generation (Code Name)	15th Gen (Arrow Lake-S)	14th Gen (Raptor Lake-S)	Zen 5 Granite Ridge
Cores	24 (8 Performance-cores/16 Efficient-cores)	24 (8 Performance-cores/16 Efficient-cores)	16
Threads	24	32	32
Default Base clock rate	3.7 GHz	3.2 GHz	4.3 GHz
Max Turbo Boost	5.6 – 5.7 GHz	5.8 – 6.0 GHz	5.7 GHz
Fabrication Process	3nm (TSMC)	10 nm	4nm
Cache L3	36 MB	36 MB	64MB
Integrated Video	Intel® UHD Graphics	Intel® UHD Graphics 770	Radeon Graphics
TDP	125W Turbo: 250W	125W Turbo: 253W	170W Boost: 200W
Socket Type	LGA1851	LGA1700	LGA1718 (AM5)
PCIe/DDR support	PCIe 5.0 / DDR5	PCIe 5.0 / DDR5	PCIe 5.0 / DDR5
Current Price from Amazon (Oct 2024)		USD $577	USD $625

Network Diagram

Test Rig

We prepared 3 test rigs with the configuration as per below.

Installation and BIOS

All BIOS settings were at default, with XMP profiles enabled (only for Intel based systems). Graphics resolution set to 1920×1080 (HD 1080p). High or Ultra settings enabled.

CPU-Z Info

CPU-Z indicates that the Core Ultra 9 285K is manufactured using a “7nm” process, but according to Intel, the CPU Lithography uses TSMC’s N3B (3nm) node. I swear, before I wrote this comparison, I saw the specification on Intel’s page showing 7nm!

Load Temps

As originally stated by Intel, the Core Ultra 9 285K did run at lower temperature than the Core i9-14900K, even though it has the same TDP of 125W.

Benchmarks (Part 1)

I’ll let the numbers do all the talking … 1st column shows the Intel Core Ultra 9 285K on a Z890 motherboard, while the 2nd column shows the Intel Core i9-14900K on a Z790 motherboard, while the 3rd column shows the Ryzen 9 9950X on X870E motherboard.

AIDA64 Memory Test

This benchmark measures the bandwidth and latency of the CPU caches and the system memory. By double-clicking any rectangle, column or row in the window, we can launch benchmarks or benchmark types individually. For example, if we double-click “Memory”, only system memory read, write, copy and latency benchmarks will be run, that is, only the operations in the given row are executed. Similarly, if we double-click “Read”, only read benchmarks will be run on all memory types, that is, only the operations in the given column are executed. If we double-click any rectangle, only the selected single benchmark will be run.

PassMark 11

Fast, easy to use, PC speed testing and benchmarking. PassMark PerformanceTest allows you to objectively benchmark a PC using a variety of different speed tests and compare the results to other computers.

HyperPi  

A simple-to-configure and portable software program that helps you calculate the pi value for a given number of digits and perform overclocking operations on multi-core machines.

PCMark 10

PCMark 10 is the latest in our series of industry standard PC benchmarks. Updated for Windows 10 with new and improved workloads, PCMark 10 is also faster and easier to use. PCMark 10 features a comprehensive set of tests that cover the wide variety of tasks performed in the modern workplace. With express, extended, and custom run options to suit your needs, PCMark 10 is the complete PC benchmark for the modern office and an ideal choice for organizations that buy PCs in high volumes.

Steel Nomad (High-end Gaming Benchmark)

3DMark Steel Nomad is a cross-platform, non-raytraced benchmark for high-end gaming PCs. It uses the DirectX 12 API by default on Windows. In Explorer mode, you can freely explore the scene and change rendering settings.

Your device must have at least 6GB of video memory to run this test. Systems with Integrated GPUs need 16 GB total system RAM.

3DMark Speed Way

3DMark Speed Way is a graphics card benchmark for testing DirectX 12 Ultimate performance. To run this test, you must have a graphics card that supports DirectX 12 Ultimate and has 6GB or more of video memory.

3DMark – Port Royal (DXR – Ray Tracing)

Port Royal is a graphics card benchmark for testing real-time ray tracing performance. To run this test, you must have a graphics card and drivers that support Microsoft DirectX Raytracing.

Time Spy

3DMark Time Spy is a new DirectX 12 benchmark test for Windows 10 gaming PCs. Time Spy is one of the first DirectX 12 apps to be built “the right way” from the ground up to fully realize the performance gains that the new API offers. With its pure DirectX 12 engine, which supports new API features like asynchronous compute, explicit multi-adapter, and multi-threading, Time Spy is the ideal test for benchmarking the latest graphics cards.

UL Procyon Video Editing Benchmark

The UL Procyon Video Editing Benchmark uses Adobe Premiere Pro in a typical video editing workflow. Using relevant apps ensures that the benchmark score reflects the real-world performance of the whole system.

The benchmark starts by importing two video project files. The project timelines include various edits, adjustments and effects. The second project uses several GPU-accelerated effects.

Each video project is exported in Full HD encoded with H.264 and again in 4K UHD encoded with HEVC (H.265). The benchmark score is based on the time taken to export all four videos.

UL Procyon AI Computer Vision Benchmark

Windows Machine Learning (Windows ML) is an API developed by Microsoft enabling high-performance AI inferences on Windows devices. Windows ML lets app developers write standard code that guarantees highly optimized AI inference performance across different hardware such as CPUs, GPUs and AI Accelerators by handling hardware abstraction.

Microsoft Windows ML hardware acceleration is built on top of DirectML, a low-level DirectX 12 library suitable for high-performance, low-latency applications such as frameworks, games, or machine learning inferencing workloads.

 

Cinebench 2024

Cinebench 2024 utilizes the power of Redshift, Cinema 4D’s default rendering engine, to evaluate your computer’s CPU and GPU capabilities. Cinebench 2024 is designed to accommodate a broad range of hardware configurations – while it seamlessly supports x86/64 architecture (Intel/AMD) on Windows and macOS, it also extends its reach to Apple Silicon on macOS and Arm64 CPUs on Windows, ensuring compatibility with the latest advancements in hardware technology. Additionally, Cinebench 2024 streamlines the benchmarking process by utilizing a consistent scene file for both CPU and GPU testing.

Geekbench 6 CPU 

Geekbench 6 is a cross-platform benchmark that measures your system’s performance with the press of a button. In Geekbench 6, new tests measure performance in popular application areas, including Artificial Intelligence, Augmented Reality, and Machine Learning, so you’ll know how close your device is to the cutting edge.

Verdict and Conclusion

It seems that with each passing year, Intel is becoming less and less innovative. They used to be the market leaders in the CPU space, producing chips that not only outperform its competition by a significant margin, but also offer new innovative features and technologies. However recently, with so much internal troubles (organizationally) in 2024, and the lack of new and fresh ideas, sadly those glorious golden days are a distant past …

To give you an idea of how dire the situation is with Intel, they’re literally 5 years behind! … Picture this, they’ve only just moved away from the 10nm fabrication process, while its competitor AMD started producing 7nm chips way back in 2019. In fact, AMD have advanced even further and achieved 4nm technology on their flagship desktop processor the Ryzen 9 9950X. And what’s even more embrarassing, is that the new Core Ultra processors are not produced by Intel themselves, but outsources to TSMC. Oh, how the mighty have fallen!

There are some advantages with the new Core Ultra processors (if you can call it that). These new chips are now more energy efficient and consumes less power. More performance per watt is Intel’s new marketing strategy, that’s my guess. Any way, the truth is the lower the power consumption, the less heat it will produce … so hopefully the CPU will run a little cooler.

Now let’s talk about performance between the 3 systems. Individually, each system performed really well in all of our benchmarks. In fact, I wouldn’t mind owning any one of these systems. If you’re AMD fan, then of course you would go for the Ryzen 9 9950X. But when it comes to picking one of the Intel systems, the choice becomes is a little harder.

Based on our tests and benchmarks, I can tell you that there’s literally no real seeable differences between the Core i9-14900K and the newer Core Ultra 9 285K. Seriously. In gaming benchmarks (3D Mark Suite) the Core Ultra 9 285K perform does perform slightly lower than the Core i9-14900K. However, in single and multicore tests (Cinebench 2024), there is a marginal improvement. It does run a little cooler and consumes less power, but apart from that … it hasn’t really improved or increased my productivity to be honest.

If you already own an Intel 13th or 4th Gen processor, which by the way is more than capable for today’s applications, then there’s no real need to upgrade to the new Core Ultra (Series 2). But for those who’s on the older 10th or 11th Gen Intel processor and planning to upgrade … now maybe a good time to get the Core Ultra. You’ll need a new Z890 motherboard anyway, so you might as well go all in 😀

There is one thing which I found quite odd. Looking at the CPUZ screenshot, Intel ‘s new Core Ultra (Series 2) and the older 14th Gen processors does not have AVX-512, but the Ryzen 9000 series does have it. According to Intel, they said it wasn’t enabled due to the inclusion of two different architectures (E-cores + P-cores). However, some motherboard manufacturers allow users to enable AVX-512 after disabling the E cores in the BIOS.

When it comes to thermals … the Core Ultra 9 285K did really well. With the same TDP of 125W as the Core i9-14900K, the load temperature reach a high of only 82 degrees Celsius. So it’s true what Intel said … the Core Ultra 9 285K processor does run a little cooler for the same performance and TDP.

Which ever processor you decide to go with, we highly recommend you install a 360 AIO cooler.

So the question you’ve all been waiting to ask … should I get one? Well, if your PC uses the older 10th or 11th gen processor, then sure go ahead and upgrade to the new Core Ultra (Series 2) processor, along with a new Z890 chipset motherboard. But it will be a major upgrade, costing you well over $800 just for the chip and board.

For those who’s already on the 12th, 13th and 14th Gen processors … I suggest you wait, as I don’t see any real benefit the new chip will bring, other than it’s new, plus you need to upgrade your motherboard, which is a real pain in the ass!

With all things taken into consideration … we give the Core Ultra 9 285K a rating of 6.5 to 7/10. Here’s a quick breakdown below.

Quick Breakdown

So here’s my final thoughts … Whether you’re a gamer, streamer or content creator, and you need a NEW desktop PC, then all the above processors will offer great performance. Just choose one based on your budget and needs.

Related articles

• AMD Ryzen 9 9950X vs Intel Core i9-14900K Comparison
• Intel Core i9-14900K vs Core i9-13900K vs Core i9-12900KF Processor Comparison
• Intel Core i7-14700K vs Core i9-12900KF vs Core i9-13900K Processor Comparison
• AMD Ryzen 9 7900X vs Intel Core i9-12900KF vs Intel Core i9-13900K Comparison
• Intel Core i9-10900K vs 11900K vs 12900KF Processor Comparison

• ASRock X870E Taichi Motherboard Review
• Gigabyte X870 AORUS ELITE WiFi7 Motherboard Review
• How to Install Your DDR5 Ram Correctly (in the correct slots) in Order to Run XMP Mode
• How to Install Intel LGA1700 Mounting Kit from Cooler Master