Image: Intel

A Chinese YouTuber has shared some of the first benchmarks for Intel’s upcoming 11th Gen Core “Rocket Lake-S” flagship, the i9-11900K. Unfortunately, the metrics do not do the processor any real favors, as they allude to the chip performing worse than its Comet Lake-S predecessor in various tests.

In the 1080p gaming benchmarks, Intel’s i9-11900K slips behind the i9-10900K in numerous titles such as Shadow of the Tomb Raider, Cyberpunk 2077, Assassin’s Creed Valhalla, and PUBG when it comes to average frame rates. The Rocket Lake-S part also loses to its Comet-Lake-S sibling in every 3DMark test (e.g., Fire Strike) that the reviewer presented.

You can check out the benchmarks compiled by harukaze5719 below. The reviewer used a Z490 motherboard to test the CPUs, which were both clocked at 5.2 GHz (BIOS locked to DDR4 3600 MHz). For Intel’s sake, let’s hope that newer BIOSes and motherboards will put the Rocket Lake-S chip in a better light.

Join the Conversation

9 Comments

  1. In a world where nobody is surprised that Intel’s vaunted best in performance is less than that projected.. nobody is surprised.
  2. Of course no one looked at the 1% lows, 0.1% lows, and minimum framerates.

    Those all show a decent bump, approaching 50% in some cases. That’s a trade I’ll make vs. 5% average, every day of the week.

    [obligatory engineering sample + prerelease BIOS + some random benchmarker disclaimer statement]

  3. Of course no one looked at the 1% lows, 0.1% lows, and minimum framerates.

    Those all show a decent bump, approaching 50% in some cases. That’s a trade I’ll make vs. 5% average, every day of the week.

    [obligatory engineering sample + prerelease BIOS + some random benchmarker disclaimer statement]

    Minimums really are where the story lies for playability. Look at CP2077 on the graph there. Avg 10900 77FPS, 11900 64FPS. Both look playable. The mins tell a different story.

    10900 dips to 18FPS, with the 1% at 22FPS. So roughly every 1.4 seconds (100 frames rendered) it drops to about 22FPS, which would probably be unplayable.

    11900 has a lower average, but the minimum is still 30FPS, 1% at 32FPS. Still probably noticeable, but more playable.

    Minimums determine how smooth the game appears.

  4. Given the high prices on these 11th gen processors that leaked out (supposedly), It’s going to be a toss up on how well they sell overall.
  5. Of course no one looked at the 1% lows, 0.1% lows, and minimum framerates.

    Those all show a decent bump, approaching 50% in some cases. That’s a trade I’ll make vs. 5% average, every day of the week.

    [obligatory engineering sample + prerelease BIOS + some random benchmarker disclaimer statement]

    Is it enough of a bump to stake an entire new generation on it?

  6. Is it enough of a bump to stake an entire new generation on it?

    It is for Intel, apparently. No idea where Rocket Lake is going to land overall, but since it’s still 14nm, not going backward is perhaps the least they can do.

    As we’ve seen, people are going to buy it because it’s available.

  7. Of course no one looked at the 1% lows, 0.1% lows, and minimum framerates.

    Those all show a decent bump, approaching 50% in some cases. That’s a trade I’ll make vs. 5% average, every day of the week.

    [obligatory engineering sample + prerelease BIOS + some random benchmarker disclaimer statement]

    There is some truth to this, but given that it’s launching soon I wouldn’t expect something drastically different than what we are seeing now. Your point about the benchmarker in question is valid though. I have no idea who they are. We don’t know the methods used here and the game sampling is rather small.

  8. Minimums really are where the story lies for playability. Look at CP2077 on the graph there. Avg 10900 77FPS, 11900 64FPS. Both look playable. The mins tell a different story.

    10900 dips to 18FPS, with the 1% at 22FPS. So roughly every 1.4 seconds (100 frames rendered) it drops to about 22FPS, which would probably be unplayable.

    11900 has a lower average, but the minimum is still 30FPS, 1% at 32FPS. Still probably noticeable, but more playable.

    Minimums determine how smooth the game appears.

    Was going to say. I was looking for where the poor benchmark numbers were and saw the exact opposite. Looks like a decent upgrade to the 10900K. I would imagine the lower averages and 3DMark scores are due to the 11900K have two fewer cores compared to the 10900K.

  9. Was going to say. I was looking for where the poor benchmark numbers were and saw the exact opposite. Looks like a decent upgrade to the 10900K. I would imagine the lower averages and 3DMark scores are due to the 11900K have two fewer cores compared to the 10900K.

    Most likely. People need to remember that 3DMark 10 is not a game. It doesn’t use a game engine. It’s a test that generates arbitrary numbers. The scores are often influenced by specific hardware or configuration changes that would not impact performance in actual games. For example: Most games do not leverage high core count / thread count CPU’s. However, core count can have an impact on 3DMark scores. CPU clocks can also impact 3DMark scores in a way that doesn’t necessarily make sense.

Leave a comment