Today we can finally show you how Intel's new octa-core 9th-gen processors perform. There's a lot to go over, so we won't waste much time going over the specs considering the CPUs have been up for pre-order now for 10 days now and nothing from the spec sheet is a mystery.
On hand for testing we have the Core i9-9900K – actually we ended up with a few of these, with more tests and comparisons in the works for later on – and we also have the i7-9700K which is basically the same CPU, but crucially, with Hyper-threading disabled.
The Core i9-9900K is an 8-core processor with Hyper-threading enabled for 16 logical threads. It operates at a base frequency of 3.6 GHz but will boost as high as 4.7 GHz on all cores with a maximum single core frequency of 5 GHz. The L3 cache has been increased from the 8700K's 12MB up to 16 MB and quite shockingly despite packing 2 more cores and 4MB more cache, the TDP rating remains at 95 watts which was already a suspiciously low rating for the 8700K. We'll explore the impact of this in a bit.
The Core i7-9700K packs the same eight cores but can only process 8 simultaneous threads. It comes clocked at the same 3.6 GHz base frequency while the all-core and single-core clock speeds decreased by a marginal 100 MHz. The L3 cache capacity is dropped down to 12 MB, too.
For testing we're using the MSI Z390 Godlike, but also used the Asrock Z390 Taichi Ultimate to confirm the results. Both boards were tested using DDR4-3200 CL14 memory and this same memory was used on all platforms without any manually-tuned timings. The graphics card of choice is Gigabyte's RTX 2080 Ti Gaming OC. We have loads of results to go over so let's get started!
Benchmarks
First up we have the memory bandwidth results and unsurprisingly the new Coffee Lake refresh CPUs are on par with previous models such as the Core i7-8700K. So everything is as expected here, let's check out some Cinebench results.
As expected the 9900K and 9700K provide the highest out of the box single thread scores we've seen to date, easily breaking the 200 pts barrier thanks to a 5 and 4.9 GHz clock speed when using just a single core.
With all cores active the 9900K breaks the 2000 pt barrier making it 14% faster than the Ryzen 7 2700X. Meanwhile the 9700K was managed a score of just over 1500 pts that placed it just behind the old 1800X and just ahead of the 8700K. That also meant it was 26% slower than the 9900K.
Given what we saw in Cinebench it's no surprise that the 9900K outclassed the 2700X in Blender, reducing the workload completion time by rather large 23%. The 8-core Ryzen CPU was a fraction faster than the 9700K though.
Moving on to Corona and here we find a similar story, the 9900K reduced the render time by 20% from the 2700X, taking just 96 seconds. Though if your mostly rendering then the Threadripper 2950X makes more sense and I'll talk more about that a little later on.
The last rendering application we tested with is V-Ray and here the 9900K reduced the render time by 18%, taking just 62 seconds opposed to 76 seconds for the 2700X. The 9700K was a lot less impressive, taking a few seconds longer than the 8700K, making it slower than both the 1800X and 2700X.
The PCMark 10 synthetic gaming benchmark relies heavily on both clock speed and core count, that said it's interesting to see the 9900K only matching the 2700X here while the 9700K was able to edge ahead of the older 1800X.