MSI MEG Z890 Ace: 28 phases for Arrow Lake CPUs and…

Extremely robust power delivery, 10-gigabit ethernet, a huge number of USB ports and more – that’s what comes with one of the most feature-packed LGA 1851 platform motherboards, the MSI MEG Z890 Ace. This is mirrored, of course, by the remarkably high price tag. Such a wealth of features comes at a price, and it’s all about whether you see a potential in it that you can appreciate. Be that as it may, MSI has packed it really properly.

Conclusion

Without a doubt one of the most extreme motherboards to pass the HWCooling testlab. For some, the MSI MEG Z890 Ace may be too expensive (and underutilized), for others it is exactly what they were looking for. For example, because of the very strong power delivery.

Don’t be fooled by the relatively high VRM temperatures with no power limits. This is because, in factory settings, the motherboard pushes for as much stability as possible even at the cost of higher power consumption. After manual tuning, the situation can be reversed. In a way that you will already benefit from the high efficiency that can be achieved on this motherboard. The Z890 Ace motherboard is made for more advanced user tuning.

In performance tests, the Z890 Ace scores top-notch results. Not just in terms of CPU performance, but also things related to M.2 SSD slot throughput. The Ethernet is also extremely fast. We’ve only seen transfer speeds over 1 GB/s (from a 10 Gb connection) on the X670E Ace besides this motherboard.

USB port speeds are as expected, within the norm, and overall we can say that there are no performance anomalies on the Z890 Ace motherboard. Not even when it comes to gaming performance. These things are usually not impacted by the motherboard. That is, not positively, but in this case not negatively either.

From the memory tests in Aida64 we can highlight the above-standard read and write speeds within the tested platform (LGA 1851 with the CU9 285K), but in turn it is also worth pointing out the relatively higher L3 cache latency. In any case, the differences are always small, and practical tests have shown that there is no point in getting excited over these measurements. The impact on the results of “real world” tests is negligible to nonexistent.

If there’s anything to highlight about the Z890 Ace, it’s the extremely rich internal and external connectivity. As many as five M.2 SSDs can be installed, with two of them supporting PCI Express 5.0. There’s only one, the first slot, for a graphics card with PCIe ×16 5.0 support. Or rather, the second PCIe ×16 slot also provides PCIe 5.0, but that one is “only” an 8-lane one.

Then you also have two headers for four USB 3.2 gen. 1 connectors (i.e. you can fully plug in the front panel even on PC cases with premium connectivity), which isn’t the case with cheaper boards (where there’s only one 19-pin USB header). There are also a lot of external USB ports on the rear panel. Overall, including the internal USB ports, we end up with a total of 24.

And finally the 10 Gb network adapter, proper audio and in the end the Z890 Ace comes in at a rather high price of around 650 EUR. Bearing that in mind, yes, this is not a motherboard for everyone. As with other expensive models, it has to be said that you have to be able to benefit from the added value, which is certainly there with the Z890 Ace. While it may not make sense to one, it may to another. No matter how high the price is, we can’t deny the “Top-Notch” award to such a well-equipped motherboard. It’s one of the most feature-rich options for powerful Core Ultra 200S processors for gaming PCs, as well as workstations or a combination of both types of setups.

English translation and edit by Jozef Dudáš

Wordpress Table Plugin

* Applies to a configurations with BIOS v1.A50 and Intel Core Ultra 9 285K processor.

Some of the tested boards are also available in the Datacomp e-store

Special thanks also to Blackmagic Design (for the license to DaVinci Resolve Studio) and Topaz Labs (for the licenses to DeNoise AI, Gigapixel AI and Sharpen AI)