First Benchmarks Of Intel’s New CPUs Appear

First Benchmarks Of Intel’s New CPUs Appear
Image: Youtube (Karl MrTechQc)

Heads up: there’s some new silicon on the way very soon. From Intel, to be exact. And if you’re wondering roughly how far ahead it’ll be of the last generation, the first benchmarks have appeared online.

The figures come courtesy of tech blogger Karl Morin, who happened to come across a HP Omen PC with the unreleased i7-8700K CPU at a tech event in Germany. For reference, the 8700K is part of Intel’s upcoming Coffee Lake line, and sports 6 core/12 threads with a base clock speed of 3.7GHz.

There wasn’t a nearby monitor around to test, so Morin grabbed one and fired up the multi-threaded test of Cinebench R15.

1230 isn’t bad. It’s higher than the Ryzen 5 1600 CPU according to the screenshot above, although my tests with the Ryzen 1600X at launch found that it was basically on par with that.

That’s a good thing! The key will be seeing how much higher users can push the CPU. Ryzen CPU’s don’t have a great deal of headroom, and games often get more mileage out of higher clock speeds than an extra two or four cores.

For good measure, Morin also fired up Cinebench’s single core test:

And some tests in CPU-Z:

Not too shabby. CPUs aren’t expected out until at least the second week of October, but we’ll keep you posted as to precisely what’s launching, when it’s available, and for how much.


  • Higher than a Ryzen 1600, not 1600X.

    That puts it at a 500mhz disadvantage.

    If you take a 1600X, which has a base clock of 3.6Ghz, and scaling those numbers as a percentage, the 1600X is a decent bit faster.

    • Made an update so it says Ryzen 1600. Also, you’d have to factor in the IPC disadvantage that Ryzen had against the last Intel generation (5% AFAIK) so even when overclocked to the same base, it should be pretty much an even keel (if not a nose ahead for the 8700K). But we’ll get more real world stuff soon.

        • AMD’s SMT implementation isn’t superior, they’re both essentially the same this generation. The reason Ryzen edges out competitors like the 7820X on single-to-multicore performance ratio is because they have a ton of extra L3 cache slapped on – the 1800X has 45% more than the 7820X and 16% more than even the 10C/20T 7900X.

          That aside, the 1600X benches 8% faster than the 1600 on both single core and multithreaded tests, which based on the results in the photos above would still put it below the 8700K on both metrics. Scaling for the extra 100MHz (2.7% difference) would put them about equal on multithreading performance, but still 25% slower on single core.

          • It’s not just a victim cache, it’s the first level of shared cache at both the CCX and die level, which is essential for inter-core communication since L1 and L2 are private caches. It’s not that expensive to access, with proper scheduling it’s only half L2’s read/copy speeds and fairly close on write speed, but still 10-15 times faster than RAM.

  • ummmmmm….
    Intel’s not expected to unveil their 8th generation of CPUs until 1:00 AM August 22.

    Aren’t we in September? Or is that August 22 in 2018?

Log in to comment on this story!