Intel’s new Core Ultra 9 285K may struggle against its predecessor

News Room

Intel Arrow Lake processors are here, and the Core Ultra 9 285K stands at the top of the lineup as Intel’s current best CPU. However, the chip faces a lot of competition not just from AMD, but also from Intel’s own last-gen Raptor Lake refresh. Even Intel itself can’t deny that performance-wise, the Core Ultra 9 285K and the Core i9-14900K are not that far off.

Although the performance uplift might be subtle (or even non-existent) at times, Intel’s Core Ultra 9 285K does bring some interesting changes. Here’s a rundown of how the Core Ultra 9 285K and the Core i9-14900K stack up against each other.

Pricing and availability

The Core Ultra 9 285K and the Core i9-14900K are almost exactly one year apart.

The Core i9-14900K, part of the Raptor Lake Refresh lineup and a 14th-gen processor, was officially released on October 17, 2023. Meanwhile, the Core Ultra 9 285K was announced alongside the rest of the Arrow Lake-S lineup on October 10, 2024, and will hit the shelves on October 24.

Both processors were launched with the exact same recommended list price (MSRP) of $590. However, one year later, you’ll find the Core i9-14900K went through several price cuts and is now a great deal cheaper. You can pick it up for as little as . When the Core Ultra 9 285K is up for sale, we expect it to be at around the MSRP for a while, but it might help push down the price of the Core i9-14900K even further.

Specs

Core Ultra 9 285K Core i9-14900K
Socket LGA1851 LGA1700
Cores/Threads 24 (8+16) / 24 24 (8+16) / 32
L3/L2 Cache 36MB / 40MB 36MB / 32MB
Max turbo frequency 5.7GHz 6GHz
Base/Turbo power 125W / 250W 125W / 253W
NPU 13 TOPS N/A
Current price $590

When you look at the specs, these two processors seem like they’re almost the same thing — and if one had to pick a winner, the Core i9-14900K might look like the better deal. It has more threads, but that’s because Intel finally killed Hyper-Threading. It also sports a higher clock speed.

The Core Ultra 9 285K retains the same core configuration as its predecessor, complete with eight P cores and 16 E cores. The lack of threads is a big one, as after over a decade, Intel is no longer using multi-threading in its desktop CPUs. This sets this generation apart not from just previous Intel products, but also from AMD, which uses SMT (simultaneous multithreading) in its latest Zen 5 processors. In fact, the flagship Ryzen 9 9950X only sports 16 cores, but it also comes with 32 threads.

The lower maximum clock speed is not a surprise, given that Intel seems to be targeting efficiency with this generation as opposed to pushing for huge performance gains. On the other hand, the base P core clocks are much higher on the Core Ultra 9 285K than on the Core i9-14900K. The newer CPU sits at 3.7GHz, and the older chip sports 3.2GHz. The same can be said about E cores: The Core Ultra 9 285K offers a base frequency of 3.2GHz and a maximum clock of 4.6GHz, and its predecessor maxes out at 2.4GHz and 4.4GHz, respectively.

Although Intel spoke a lot about the efficiency of Arrow Lake-S — which we’ll talk about below — the CPU still has practically the same TDP as the previous-gen part.

One more major spec change is that Intel is ditching the LGA1700 socket and moving on to LGA1851. This means new motherboards, but also a huge difference in memory support, because the Core Ultra 9 285K abandons DDR4 RAM and only supports up to 192GB of DDR5-6400 memory. Meanwhile, the Raptor Lake chip can run either DDR4 or DDR5.

Lastly, the Arrow Lake processor has some bragging rights for featuring the first neural processing unit (NPU) found in a desktop CPU. Designed for AI workloads, the NPU in the Core Ultra 9 285K only sports 13 tera operations per second (TOPS). This is much less than what Lunar Lake CPUs had to sport in order to qualify for the Copilot+ program. However, on a desktop computer, AI workloads are usually handled by the GPU, which makes the NPU a nice perk but not a dealbreaker one way or another.

Efficiency

Power draw for Intel's Arrow Lake CPUs.

A big selling point for the Core Ultra 9 285K, as well as the rest of the Core Ultra 200-S lineup, lies in its efficiency. Although we’ve yet to test it ourselves, Intel talked a lot about how much cooler and more efficient the new CPUs are, providing some benchmarks to prove it. The company’s main claim is that Arrow Lake-S, with the Core Ultra 9 285K leading the charge, can offer the same performance at half the power draw. Let’s take a closer look.

Efficiency figures for the Core Ultra 9 285K.

First, Intel talked about the way the Core Ultra 9 285K stacks up against the Core i9-14900K in productivity tests, including Procyon Office Productivity and the Cinebench 2024 single-core test. Based on these benchmarks, Intel claims package power reductions ranging from 42% to 58%.

Efficiency chart comparing the Core Ultra 9 285K to the Ryzen 9 9950X and the Core i9-14900K.

Intel also claims to offer the same performance as not just the Core i9-14900K, but also the Ryzen 9 9950X, while only consuming half the power — 125 watts instead of 250 watts. This is in a Cinebech multi-core test.

A comparison of gaming performance-per-watt for the Core Ultra 9 285K and the Core i9-14900K.

Lastly, Intel dips into performance per watt in gaming scenarios by comparing the power draw of the entire system. That means not just the CPU but also all the rest of it, such as the Z780 motherboard used for testing the Core i9-14900K and the 800-series board Intel paired with the Core Ultra 9 285K.

In many titles, the Core Ultra 9 285K managed to offer roughly the same frame rates as the Core i9-14900K. This includes games like Metro Exodus and Black Myth: Wukong. In Total War: Pharaoh, the Arrow Lake CPU maintained the same frames per second (fps) while reducing system power by 58 watts. The biggest gains are seen in Warhammer 40,000: Space Marine 2, where the Core Ultra 9 285K pushes ahead by 4% while cutting system power by a whopping 165 watts.

Arrow Lake-S temperature drops in various games.

In addition to lower system power, Intel also claims that the new CPU runs up to 15 degrees Celsius cooler than the Core i9-14900K.

These figures are promising for the Arrow Lake chip, but it’s important not to take them at face value — we’ll have to test it ourselves and see just how much the new chip can keep up with its predecessor at a lower power draw.

Performance

Intel Arrow Lake performance versus Ryzen 9000.

Aside from efficiency, Intel also provided a slew of benchmarks in gaming and productivity scenarios. The results are a bit all over the place, and unfortunately, Intel chose the rival Ryzen 9 9950X to compare to in most of these tests. However, we’ve compared the Ryzen 9 9950X to the Core i9-14900K in our in-depth benchmarks, so that should give you a good idea of how the two CPUs stack up. For reference, we’ve found that the 9950X is about 10% faster in productivity, but it lagged behind in our gaming tests.

As for the Core Ultra 9 285K, Intel claims a lead of up to 28% in Total War: Warhammer III, but the harsh reality is that most titles are even — that game is the one single major victory for Intel. The CPU trails behind AMD in Cyberpunk 2077 and Red Dead Redemption 2, and that’s with Intel APO enabled, so that might tip the scales in Intel’s favor, too.

Charts comparing the Core Ultra 9 285K to the Ryzen 9 7950X3D.

Intel also pits the Core Ultra 9 285K against the last-gen Ryzen 9 7950X3D. AMD’s 3D V-Cache is difficult to win against, hence why we’re seeing a major 21% drop in Cyberpunk 2077 — but that drop would still very much be there if Intel compared the Core i9-14900K to the 7950X3D. Intel also claims one victory over the 7950X3D, with Civilization VI scoring an extra 15% fps.

Given that the Ryzen 9 9950X is generally slower in gaming than the Core i9-14900K, we expect the Core Ultra 9 285K and its predecessor to end up mostly on par. We’ll have to check this thoroughly once the CPU is available for testing.

Content creation tells a different story. Intel’s Arrow Lake CPU wipes the floor with the Ryzen 9 7950X3D in some tests, with a 21% win in Cinebench. But if this was a comparison against the 9950X, we’d probably see a much different result.

Multi-threaded performance for Intel's Arrow Lake CPUs.

Back to the Core i9-14900K, Intel touts an up to 15% gain in multi-threaded performance when compared to the Raptor Lake Refresh chip. Intel also compared the Core Ultra 9 285K against the 9950X again here, with a win of up to 19% in 3DMark and 18% in Cinebench’s multi-core test.

Additionally, Intel showed off some improved render times — an important thing for content creators. Unfortunately, as those comparisons were once again made against the Ryzen 9 9950X, it’s hard to imagine how the CPU will stack up against the Core i9-14900K here. Intel’s biggest wins were huge, and the losses were quite small, but again, we’ll need to draw our own conclusions once we’ve tested the chip.

A hard sell

A render of an Intel Core Ultra 200-S chip.

Intel’s Arrow Lake-S flagship is both lucky and unfortunate. It’s lucky because it arrives met with pretty average competition, as AMD’s Zen 5 CPUs also didn’t provide much of a generational uplift compared to their predecessors. However, it’s unfortunate because it’s certainly not the huge improvement people may have hoped for.

Instead of pushing for higher performance, Intel chose to target efficiency. This is a good approach, but with an enthusiast CPU, it may not play as big a part as it would with a lower-end chip. Most people don’t buy a $600 processor to save on their electric bills or in order to buy a budget cooler; they want performance gains, which the Core Ultra 9 285K might not deliver much of.

However, the main reason why the Core Ultra 9 285K might be a tough sell is that the Core i9-14900K is now much cheaper than it was when it was first released. Although both chips have the same MSRP, the Core i9-14900K is $150 cheaper.

If our benchmarks align with what Intel’s shown us so far, the Core Ultra 9 285K is unlikely to be worth a $150 premium. Once we’ve tested it ourselves, we’ll be able to tell you more about its performance.






Read the full article here

Share This Article
Leave a comment

Leave a Reply

Your email address will not be published. Required fields are marked *