Affiliate disclosure: as an Amazon Associate, we may earn commissions from qualifying purchases from Amazon.
Intel has just released their new CPUs, and they’re shaking things up with a fresh naming system: Intel Core Ultra, starting with the 200 series. Today, we’re diving into the Intel Core Ultra 9, specifically the 285K model. But what’s really intriguing about this generation is how radically different it is. Intel has switched over to TSMC for manufacturing and completely revamped the architecture, even dropping hyperthreading—a surprising move. Yet, despite these changes, Intel is promising similar performance to the previous generation, but with significantly less power consumption and lower heat output. It’s a bold shift, and one that could change the game.
Last-gen CPUs were notoriously tough to keep cool at full power, so it got me thinking—has Intel actually made cooling easier this time around? What kind of cooler do you need to really push these chips, and what’s the bare minimum to keep things under control? After putting them to the test, I can say that while Intel’s top-tier processors still run hot, they’re not nearly as difficult to manage as before. Let’s break down the results.
Let’s start by going over our setup. We’re testing with the Intel 285K, the flagship processor of this generation, featuring 8 Performance cores and 16 Efficiency cores. For the motherboard, we’re using the Gigabyte Z890 Elite, paired with 48 GB of DDR5 memory from Kingston, running at 8400 MT/s.
For cooling, we tested a range of options. At the cheaper end, we used the Pure Rock 2 FX air cooler from “be quiet!”, as well as the LE500 240mm water cooler from DeepCool. On the high end, we tested both the Dark Rock Elite air cooler and the massive 360mm LS720 water cooler, also from DeepCool. To push the limits, we cranked up the fan and pump speeds to 100% in every scenario, aiming to see the maximum cooling performance.
Let’s kick things off with a burn-in test using Prime95, where we push the CPU to its limits to see how each cooler performs. Right away, the temperature graph shows the lower-end air cooler and the 240mm water cooler both climbing above 100 degrees, while the higher-end coolers hover around 80 degrees. So, in a stock configuration, the high-end coolers manage to keep the CPU running at full throttle without overheating.
Just because the CPU runs hot doesn’t necessarily mean it’s thermal throttling, so we also looked at the frequency graph to check performance. This shows the average effective frequency across all cores, keeping in mind that the larger Performance cores and smaller Efficiency cores run at different speeds. The focus here is more on the relative differences between coolers than on absolute performance—we’ll dive into the test results in detail later.
In the frequency graph, the higher-end coolers are neck and neck, with the LS 720 possibly performing slightly better or at least more consistently. The smaller liquid cooler struggled, while the Pure Rock 2 air cooler landed somewhere in the middle. Notably, the difference between the best and worst coolers was around 700 to 900 MHz, which is a significant gap for an all-core workload, meaning we’re definitely losing performance with the lower-end options.
Now, looking at power consumption during the same test, the chip paired with the 240mm AIO dropped below 190W, while the other coolers allowed it to pull between 220 and 240W, which is significantly more.
When examining Vray benchmark results, this short test primarily evaluates a cooler's ability to rapidly dissipate heat from the CPU heat spreader within a span of about one minute. Rather than testing sustained cooling performance over extended periods, it focuses on how quickly each cooler can offload heat during intense bursts of computational workload.
In this comparison, the “be quiet!” Dark Rock Elite is just short of 3% behind the top-performing Deep Cool LS 720, while the “be quiet!” Pure Rock 2 FX lags by about 8%, and the Deep Cool LE 500 by almost 10%.
Next up, we tested Cinebench R23 for both multi-core and single-core performance. Both lower-end coolers had issues completing the single-core test, requiring several repeats, while the higher-end coolers had no problems. This could be due to chip stability rather than cooler performance, but the differences in cooling efficiency might also be a factor.
In multi-core, the LS 720 leads, with the Dark Rock Elite trailing by about 3%. The LE 500 and Pure Rock 2 FX are both around 5% behind. This suggests that while the lower-end coolers are capable, they can’t quite match the higher-end models in handling sustained multi-core loads, which demand better heat management over time.
For single-core, the LS 720 is again on top, with the Dark Rock Elite less than 1% behind, showing these two coolers are neck-and-neck when it comes to managing brief bursts of high performance. The Pure Rock 2 FX lags by 1%, and the LE 500 is 2% behind. The difference here is almost irrelevant as we are talking few percent here or there – I don’t think anyone would notice this in real world example.
Next up is the 7zip test, where we ran two separate tests—one for multi-core and one for single-core performance. Let’s start with the multi-core results.
For compressing, the Pure Rock 2 FX leads, with the LS 720 and Dark Rock Elite trailing by less than 2%. The LE 500 is about 3% behind the leader. These results suggest that all coolers perform similarly when compressing data, with only minor differences between them.
For decompressing, the Dark Rock Elite takes the top spot, with the LS 720 just 1% behind. The Pure Rock 2 FX lags by 8%, and the LE 500 by 9%. Here the higher-end coolers show a clear advantage in handling workload.
Now over to the 7zip single-thread test, for compressing, the LS 720 leads followed by the Pure Rock 2 FX, which is about 2% behind. The Dark Rock Elite and LE 500 trail by 6% and 10% respectively.
For decompressing, the LS 720 is again on top, but only by a small margin. The Pure Rock 2 FX is just 1% behind, with the LE 500 also very close, trailing by less than 1%. Interestingly, the Dark Rock Elite, which performed well in other tests, falls behind here, finishing last and trailing the leader by 4%.
The closely grouped results indicate that cooler efficiency has less impact on this single-threaded task, as all the coolers performed similarly, with their rankings shifting back and forth. We repeated the test multiple times, and each time the results were slightly different.
Which leads us to the conclusion. This is currently the highest end Intel CPU and to be honest considering how hard it has been to cool these for many years now – I am really impressed with the cooler performance here. We are using somewhat representative coolers which in many ways are able to handle the load without too much of a struggle. I would still recommend people who are getting higher-end chips to go with a dual tower air cooler or a 360mm liquid AIO to keep your temperatures and noise levels lower, but in a pinch the surprising little hero in our tests was the “be quiet!” Pure Rock 2 FX. It actually managed to keep most of the performance while being somewhat on the cheaper side.
What do you guys think of this new launch and what cooler are you planning to pair with the new Intel CPU’s ? Let us know in the comments below.