Intel Core i9-9900K 9th Generation CPU Review

The new 9th generation Intel i9-9900K CPU is upon us! AMD has been pushing into Intel's desktop market and Intel knows it. Today Intel is pulling the curtain back on "not paid for" reviews and we are happy to be serving you one of those up here today. Is the i9-9900K better than the Ryzen 7 2700X, and is it worth the staggering price premium?


Thread Scaling

All our data below is with all CPUs compared running at stock clock settings. We did not put the 8600K in here because it basically runs at locked clocks. It is interesting to see how both Intel's Boost Clocks and AMD's Precision Boost clocks react under load.

For these workloads we used Prime95, so it is in fact a very heavy workload using SmallFFTs. We allow the CPU to heat-load then refresh and start recording our data using HWInfo64. We use this process for both recording work-loaded clocks and CPU package power.

Article Image

We were wondering if the 9900K was in fact going to be able to pull off its stated 5GHz single threaded clock with Prime95 and it did in fact stay true to its word. Also the stated full load clock for the 9900K is 4.7GHz and it did in fact manage this as well. The reason we are seeing this slightly elevated is because the MSI MEG Z390 ACE ($290) has a base clock of just over 100MHz.

We see our 2700X clock bottom out to 4GHz under a full core load once heat-soaked. And as expected we see our 2950X fall off even further to 3.5Ghz once it is fully loaded across all 32 threads.

Power Scaling

Article Image

As we stated previously, the 2950X is not a truly a comparable CPU in the desktop realm, and that is why even though we included it in our charts, we did not speak to its performance, but I wanted you to be aware of it. Using the 2950X with Precision Boost 2, we see that our CPU package power is capped at the CPU's 180 watt TDP. We do not see our TDP hit on the 2950X till we have a 12 thread load.

A lot of people like to speak to Performance Per Watt or "PPW." There are a lot of ways to apply this metric, so we are not going to nail down any math to assign a value to it. You could use a multitude of workloads to prove PPW data wrong or right. What we do know is that at a full 16 thread load is that the 9900K shows a ~31% power usage increase compared to the 2700X, which greatly outpaces its performance increases shown in our benchmarks (16% average). Now you may or may not care about PPW, or the power usage at all, as long as we can dissipate the heat produced. What we do know is that when it comes to overclocking our CPUs, power usage is a big damn deal and we are going to cover that on the next page as I know many of us have very high expectations for overclocking the 9900K.