The Intel 9th Gen Review: Core i9-9900K, Core i7-9700K and Core i5-9600K Tested
by Ian Cutress on October 19, 2018 9:00 AM EST- Posted in
- CPUs
- Intel
- Coffee Lake
- 14++
- Core 9th Gen
- Core-S
- i9-9900K
- i7-9700K
- i5-9600K
Test Bed and Setup
As per our processor testing policy, we take a premium category motherboard suitable for the socket, and equip the system with a suitable amount of memory running at the manufacturer's maximum supported frequency. This is also typically run at JEDEC subtimings where possible.
It is noted that some users are not keen on this policy, stating that sometimes the maximum supported frequency is quite low, or faster memory is available at a similar price, or that the JEDEC speeds can be prohibitive for performance. While these comments make sense, ultimately very few users apply memory profiles (either XMP or other) as they require interaction with the BIOS, and most users will fall back on JEDEC supported speeds - this includes home users as well as industry who might want to shave off a cent or two from the cost or stay within the margins set by the manufacturer. Where possible, we will extend out testing to include faster memory modules either at the same time as the review or a later date.
Test Setup | |||||
Intel 9th Gen | i9-9900K i7-9700K i5-9600K |
ASRock Z370 Gaming i7** |
P1.70 | TRUE Copper |
Crucial Ballistix 4x8GB DDR4-2666 |
Intel 8th Gen | i7-8086K i7-8700K i5-8600K |
ASRock Z370 Gaming i7 |
P1.70 | TRUE Copper |
Crucial Ballistix 4x8GB DDR4-2666 |
Intel 7th Gen | i7-7700K i5-7600K |
GIGABYTE X170 ECC Extreme |
F21e | Silverstone* AR10-115XS |
G.Skill RipjawsV 2x16GB DDR4-2400 |
Intel 6th Gen | i7-6700K i5-6600K |
GIGABYTE X170 ECC Extreme |
F21e | Silverstone* AR10-115XS |
G.Skill RipjawsV 2x16GB DDR4-22133 |
Intel HEDT | i9-7900X i7-7820X i7-7800X |
ASRock X299 OC Formula |
P1.40 | TRUE Copper |
Crucial Ballistix 4x8GB DDR4-2666 |
AMD 2000 | R7 2700X R5 2600X R5 2500X |
ASRock X370 Gaming K4 |
P4.80 | Wraith Max* | G.Skill SniperX 2x8 GB DDR4-2933 |
AMD 1000 | R7 1800X | ASRock X370 Gaming K4 |
P4.80 | Wraith Max* | G.Skill SniperX 2x8 GB DDR4-2666 |
AMD TR4 | TR 1920X | ASUS ROG X399 Zenith |
0078 | Enermax Liqtech TR4 |
G.Skill FlareX 4x8GB DDR4-2666 |
GPU | Sapphire RX 460 2GB (CPU Tests) MSI GTX 1080 Gaming 8G (Gaming Tests) |
||||
PSU | Corsair AX860i Corsair AX1200i |
||||
SSD | Crucial MX200 1TB | ||||
OS | Windows 10 x64 RS3 1709 Spectre and Meltdown Patched |
||||
*VRM Supplimented with SST-FHP141-VF 173 CFM fans ** After Initial testing with the ASRock Z370 motherboard, we noted it had a voltage issue with the Core 9th Gen processors. As a result, we moved to the MSI MPG Z390 Gaming Edge AC for our power measurements. Benchmarking seems unaffected. |
We must thank the following companies for kindly providing hardware for our multiple test beds. Some of this hardware is not in this test bed specifically, but is used in other testing.
274 Comments
View All Comments
vext - Friday, October 19, 2018 - link
Very good article, but here are my beefs.Why is there no mention of temperatures?
According to Techspot the 9900k runs ridiculously hot under heavy loads. At stock clocks under a heavy Blender load it reaches 85C with a Corsair H100i Pro, or Noctua NH-D15. Pushed to 5Ghz, it hits 100C. At 5.1 Ghz it FAILS. I suggest that Anandtech has failed by not discussing this.
Techspot says:
"There’s simply no way you’re going to avoid thermal throttling without spending around $100 on the cooler, at least without your PC sounding like a jet about to take off. Throw in the Corsair H100i Pro and the 9900K now costs $700 and you still can’t overclock, at least not without running at dangerously high temperatures."
Why the focus on single threaded benchmarks? For the most part they are irrelevant. Yet they are posted in their own graph, at the front of each testing section, as though they were the most important data point. Just include them as a separate bar with the multi-thread benchmarks. Good Grief!
Why post MSRP prices in every single benchmark? You can't even buy them for MSRP. There should be a single chart at the front of the article with a rough retail estimate for each processor, and links to the retailers. If the MSRP is necessary, then just add a column to the chart. Sheesh.
Why no in depth cost/benefit comparison? A Ryzen 2600 with included cooler at $160 costs only one quarter of a 9900k with an aio cooler at $700. The $540 difference would buy a new RTX 2070 video card. Or three more Ryzen 2600's. For crying out loud.
I like the 9900k, it's a good processor. It's intended for hobbyists that can play with custom loop cooling. But it's not realistic for most people.
mapesdhs - Sunday, October 21, 2018 - link
All good questions... the silence is deafening. Thankfully, there's plenty of commentary on the value equation to be found. A small channel atm, but I like this guy's vids:https://www.youtube.com/watch?v=EWO5A9VMcyY
abufrejoval - Friday, October 19, 2018 - link
I needed something a little bigger for my lab two or three years ago and came across an E5-2696v3 on eBay from China, a Haswell generation 18-core at $700.That chips didn't officially exist, but after digging a little deeper I found it's basically an E5-2699v3 which clocks a little higher (3.8 instead of 3.6GHz) with 1-2 cores active. So it's basically a better chip for a fraction of the going price of the lesser one (E5-2699v3 is still listed at €4649 by my favorite e-tailer). And yes, it's a perfect chip, Prime95'd it for hours, POVrayd and Blendered for days until I was absolutely sure it was a prime quality chip.
Officially it has 145Watts TDP, but I've only ever seen it go to 110Watts on HWiNFO with Prime95 in its meanest settings: It must be a perfect bin. With the particle pusher it's never more than 93Watts while no part of the CPU exceeds 54°C with a Noctua 140mm fan practically inaudible at 1000rpm cooling it: That because the 18 cores and 36 threads never run faster than 2.8GHz fully loaded. They also don't drop below it (except for idle, 1.855 Watts minimum btw.), so you can pretty much forget about the 2.3GHz 'nominal' speed.
It gets 2968.245803 on that benchmark, slightly above the i9-9900k, somewhat below the ThreadRipper. That's 22nm Haswell against 14++/12nm current and 18 vs 8/12 cores.
This is rather typical for highly-threaded workloads: It's either cores or clocks and when the power ceiling is fixed you get higher throughput and energy efficiency when you can throw cores instead of clocks at the problem.
I think it's a data point worth highlighting in this crazy clock race somewhat reminiscent of Pentium 4 days, heat vs. efficiency, a four year old chip beating the newcomer in performance and almost 3:1 in efficiency at far too similar prices.
Yet, this specific chip will clock pretty high for a server chip, easily doing 3.6 GHz with eight cores seeing action from your game engine, while the remaining ten are often ignored: Perhaps that's a Ryzen effect, it used to be 4:14 earlier.
I've done BCLK overclock of 1.08 to have it reach the magic 4GHz at maximum turbo, but it's not noticeable in real-life neck-to-neck to an E3-1276v3 which also turbos to 4GHz on three cores out of four available, 3.9 at 4/4 with HT.
abufrejoval - Friday, October 19, 2018 - link
2968.245803 on the particle pusher benchmark... need editicoreaudience - Friday, October 19, 2018 - link
Move away from rar/lzma : the new darling of data compression is called Zstandard :https://www.zstd.net
It comes with a nice integrated benchmark, which can easily ramp up with multithreading :
zstd -b -1 -T8 fileToTest # benchmark level one on fileToTest using 8 threads
Windows user can even download a pre-compiled binary directly in the release notice :
https://github.com/facebook/zstd/releases/latest
It would be great to see some numbers using this compressor on latest Intel cores !
Kaihekoa - Friday, October 19, 2018 - link
Looks like all your gaming benchmarks are GPU bound and there pointless. Why not use a 2080 Ti to eliminate/reduce GPU bottleneck?Kaihekoa - Friday, October 19, 2018 - link
therefore*palladium - Friday, October 19, 2018 - link
Can you please run some SPEC2006 benchmarks and see if Apple's SOC really has caught on to Intel's performance (per core), as mentioned in Andrei in his iPhone XS review? ThanksVirpZ - Friday, October 19, 2018 - link
Apart from blender, your review is full Intel biased software for rendering.Hifihedgehog - Friday, October 19, 2018 - link
Hey Ian. I see your updated full load power consumptions results. Question: Why is it that the six-core i7-8086K is drawing so little power in comparison to everything else including the quad-cores? Is this due to its better binning or is this simply an error that crept in?