The Intel Core i9-9980XE CPU Review: Refresh Until it Hertz
by Ian Cutress on November 13, 2018 9:00 AM ESTHEDT Performance: SYSMark 2018
You either love it or hate it. BAPCo’s SYSMark suite of tests is both an interesting combination of benchmarks but also clouded in a substantial amount of ire. The altruistic original goal was to develop an industry standard suite of real-world tests. AMD (and NVIDIA) left BAPCo several years ago citing that workloads were being chosen on purpose that favoured Intel processors, and didn’t include enough several emerging computing paradigms for the industry. Intel disagrees with that statement, and here we are today.
We run this disclaimer on our SYSMark testing primarily to emphasise that this benchmark suite, while some consider it more than relevant and encompassing a lot of modern professional software, others feel is engineered with specific goals in mind.
We haven’t run SYSMark on every processor, as it requires a fresh OS image compared to our automated suite, and requires refreshing that image every seven days. As a result we are trying to do sets of processors at a time where it makes sense and when time is available.
It’s clear from the Intel comparisons that the i9-9980XE takes a lead here, with a nice bump over the 7980XE of around 7% in the overall test. Given that the 7900X is the best of the Skylake-X processors, it will be interesting to see what the 9900X scores here.
143 Comments
View All Comments
coder543 - Tuesday, November 13, 2018 - link
I think you mean that Inte's 1 or 2 cores often beat AMD's 1 or 2 cores. In benchmarks that are highly multithreaded, AMD beats Intel. Intel currently has a frequency advantage, so they win in the lightly threaded tests.coder543 - Tuesday, November 13, 2018 - link
Intel's*AnandTech and Twitter both need an edit button.
HStewart - Tuesday, November 13, 2018 - link
But my question in real life do you need this many core - I still think it better to have single thread core speed. I say that even as developer that uses multiple threads.GreenReaper - Wednesday, November 14, 2018 - link
Depends on what your real life is, doesn't it? That's why reviews don't have one benchmark and the final paragraph of the conclusion here emphasizes that the right processor for you depends on your workload, even *without* considering the relative prices. For many workloads none of the CPUs tested here are appropriate; a 2400G would be the most efficient, cost-effective option.twtech - Thursday, November 15, 2018 - link
Yes. I could use about twice this many. I also need good single-core clockspeed as well. That's why I'm eagerly anticipating the 3175x launch.bji - Tuesday, November 13, 2018 - link
Windows is garbage. Put these chips onto a real operating system (Linux) and you will see the actual performance they are capable of without Windows holding them back. See Phoronix.twtech - Thursday, November 15, 2018 - link
It's nice if you have that option. Most of the time the software you run only works on certain operating systems and you don't have much choice.Lolimaster - Thursday, November 15, 2018 - link
In benchmarks where the 2950X at 1K will score similar too? What's the point?AMD top dogs are on another whole level, if you got the workload for them, take them, else 2950X.
Arbie - Tuesday, November 13, 2018 - link
Typo - you wrote "Intel will need to up its game here to remain competitive". Should have been "Intel will need to up its marketing here to remain competitive".eva02langley - Tuesday, November 13, 2018 - link
Unfortunately, they are selling everything. I would be happy if they were not selling anything.