CPU Benchmark Performance: Simulation And Rendering

Simulation and Science have a lot of overlap in the benchmarking world, however for this distinction we’re separating into two segments mostly based on the utility of the resulting data. The benchmarks that fall under Science have a distinct use for the data they output – in our Simulation section, these act more like synthetics but at some level are still trying to simulate a given environment.

We are using DDR5 memory at the following settings:

  • DDR5-4800(B) CL40

Simulation

(3-2a) Dwarf Fortress 0.44.12 World Gen 65x65, 250 Yr

(3-2b) Dwarf Fortress 0.44.12 World Gen 129x129, 550 Yr

(3-2c) Dwarf Fortress 0.44.12 World Gen 257x257, 550 Yr

(3-3) Dolphin 5.0 Render Test

(3-4a) Factorio v1.1.26 Test, 10K Trains

(3-4b) Factorio v1.1.26 Test, 10K Belts

(3-4c) Factorio v1.1.26 Test, 20K Hybrid

When it comes to simulation, the higher core clock speeds of the Core i9-12900KS didn't make any effect in our Factorio benchmark, which AMD's Ryzen 7 5800X3D with 3D V-Cache gives a more effective boost to performance. In our Dolphin test, the Core i9-12900KS was quicker by around 7-8%, and in our Dwarf Fortress benchmark, the Core i9-12900KS was better than the K, but not as good as the Ryzen 7 5800X3D.

Rendering

(4-1) Blender 2.83 Custom Render Test

(4-2) Corona 1.3 Benchmark

(4-4) POV-Ray 3.7.1

(4-5) V-Ray Renderer

(4-6a) CineBench R20 Single Thread

(4-6b) CineBench R20 Multi-Thread

(4-7a) CineBench R23 Single Thread

(4-7b) CineBench R23 Multi-Thread

Focusing on rendering performance, the results between the Core i9-12900KS and Core i9-12900K are consistently close. In single-threaded benchmarks such as CineBench R23, we saw impressive performance in comparison to other chips on test, with a 7% jump in R23 single thread performance over the Core i9-12900K.

The multi-threaded performance though between the two Intel 12th Gen Core i9 chips wasn't much difference, but there's a clear advantage in single-thread performance when rendering. It's worth pointing out that rendering is typically a multi-threaded workload.

CPU Benchmark Performance: Power, Office, And Science CPU Benchmark Performance: Encoding and Compression
Comments Locked

56 Comments

View All Comments

  • DazFG - Friday, July 29, 2022 - link

    performance per watt for these CPUs is abismall!
  • bug77 - Friday, July 29, 2022 - link

    Which is weird, because high-end CPUs (and GPUs, for that matter) are usually hallmarks of efficiency. Right?

    But yeah, Alder Lake can take so much punishment, your PSU might cave in before the CPU does...
  • techjunkie123 - Friday, July 29, 2022 - link

    Sadly AMD is also headed in this direction with their new CPUs (Zen 4)...... 5.5 GHz and 70% higher TDP for what, 40% or so better MT performance?

    Apples high end CPU and GPU performance scales quite well with power. They just throw a lot more silicon at the problem, but it is effective.
  • tamalero - Sunday, July 31, 2022 - link

    Eyup, this is another Netburst..
  • bennyg - Monday, August 8, 2022 - link

    When you get up to 5ghz, yeah, but OC has always been that way, turbo boost is just factory auto OC
    Down at 80W in mobile the perf per watt looks far better for the 6p8e parts
    It's all Ohm's fault for that law of his
  • blanarahul - Friday, July 29, 2022 - link

    Huh. Makes you wonder how a stock 12900KS would perform under Dry Ice since I doubt any phase change cooler can keep up with the 400 watt heat output.
  • ddhelmet - Friday, July 29, 2022 - link

    Factoria updates per seconds are too low.
  • healbunny - Friday, July 29, 2022 - link

    stock voltage is way too high, please try undervolt. my 12900ks with undervolt able to get P core all core 5.3Ghz, Ecore 4.2Ghz (vcore 1.2v). 2 core boost is 5.6Ghz at 1.3v vcore. All core full load in cinebench r23 consume 240w, cpu package temperature around 80C. Single core consume around 48W, 53C. (using Msi K360 aio). Score for R23, multi 30006 pts, single 2143pts.
  • Oxford Guy - Saturday, July 30, 2022 - link

    You are clearly unfamiliar with their justification for only using JEDEC RAM settings. Undervolting is more complex than picking a vetted XMP profile.
  • Carls Car - Saturday, July 30, 2022 - link

    healbunny didn't mention XPM at all.

Log in

Don't have an account? Sign up now