CPU Benchmark Performance: Power, Office, and Science

Our previous set of ‘office’ benchmarks have often been a mix of science and synthetics, so this time we wanted to keep our office section purely on real-world performance.

For the remainder of the testing in this review of the Core i7-12700K and i5-12600K, we are using DDR5 memory at the following settings:

  • DDR5-4800(B) CL40


(0-0) Peak Power

Comparing power draw to other competing CPUs, both the Core i7-12700K and Core i5-12600K are noticeably more power-efficient than previous generations including both Intel's 11th Gen and 10th Gen Core. Though at full-load with no overclocking, AMD's Ryzen 5000 and 3000 series processors remain much more power-efficient.


(1-1) Agisoft Photoscan 1.3, Complex Test

In our office benchmarks, Intel's 12th Gen Core reigns supreme in Agisoft Photoscan due to its higher core frequency and IPC performance.


(2-1) 3D Particle Movement v2.1 (non-AVX)

(2-2) 3D Particle Movement v2.1 (Peak AVX)

(2-3) yCruncher 0.78.9506 ST (250m Pi)

(2-4) yCruncher 0.78.9506 MT (2.5b Pi)

(2-4b) yCruncher 0.78.9506 MT (250m Pi)

(2-5) NAMD ApoA1 Simulation

(2-6) AI Benchmark 0.1.2 Total

(2-6a) AI Benchmark 0.1.2 Inference

(2-6b) AI Benchmark 0.1.2 Training

In the majority of our science-based benchmarks, both the Core i7 and Core i5 did well. The only benchmarks that didn't favor the 12th Gen Core series processors were in 3DPM 2.1, but more specifically in the AVX test. 

Intel Core i7-12700K and Core i5-12600K Review: Mid-Range Desktop CPU Benchmark Performance: Simulation And Rendering
Comments Locked


View All Comments

  • Otritus - Wednesday, March 30, 2022 - link

    Chiplets should not give much of a pricing advantage. Going from monolithic dies to chiplets adds manufacturing, validation, and design costs. You save money from the chiplets having better yields and being able to reuse chiplets (like AMD did with their IO die). Intel’s consumer processors aren’t that big and yields are good, so it’s probably a wash overall in pricing (excluding the money Intel saves from in-house manufacturing). Chiplets would be a competitive advantage for big processors (EPYC vs Xeon or Navi 31 vs AD102) because of yields.
  • vlad42 - Friday, April 1, 2022 - link

    Nonsense, chiplets absolutely provide a price advantage once you are using more than one CPU chiplet. There's a reason why AMD has done so well in the HEDT, workstation, and server markets. The chiplets allow them to provide more cores for less money than Intel (the yields on those high core count monolithic chips are abysmal by comparison).
  • Qasar - Friday, April 1, 2022 - link

    ive read somewhere thats part of the reason why intel cant offer more cores then they do, and why they also went with the P/E core setup. they just can't make the cpus that big with the big cores.
  • mode_13h - Saturday, April 2, 2022 - link

    Die size gets expensive as you scale it up, because not only do you get fewer chips/wafer, but yield becomes a major factor. This is even more true on a young process, such as "Intel 7", and why smaller chips tend to be the first to utilize them.

    In one way or another, cost is *always* a factor, whenever companies are deciding on the parameters of a new chip.
  • Qasar - Saturday, April 2, 2022 - link

    i think, the other reason was power consumption, if intel uses this much power with what 10 P cores ( and no e cores ) i dont want to know what 12, or even 16 cores would use.
  • mode_13h - Sunday, April 3, 2022 - link

    > if intel uses this much power with what 10 P cores ( and no e cores )
    > i dont want to know what 12, or even 16 cores would use.

    That's not necessarily a direct tradeoff. They could cap the power (i.e. clocks) and still deliver more performance than 8 P-cores.

    Of course, the other thing they could do is trade more die area & clock speed for a uArch with higher IPC, like Apple's. But die area costs $.

    The sad reality (for the planet) is that the winning strategy seems to be making less-sophisticated cores that simply clock really high. That gets you winning headline numbers on lightly-threaded benchmarks. Then, wedge in some throughput-optimized cores, so you can also do well on heavily-threaded workloads.
  • Qasar - Sunday, April 3, 2022 - link

    just saw a review of the 12900ks, to me, when a reviewer types " the Core i9-12900KS represents Intel throwing value and power consumption out the window in a no-holds-barred attempt to retain the performance crown, particularly in gaming. " screams desperation. looks like one of the ways intel is only able to compete, is by doing this.
  • Mike Bruzzone - Tuesday, April 5, 2022 - link

    @Otritus, I agree MCMs trade off packaging cost for monolithic die fabrication cost. Moving to Systems in Package in the near term will eclipse their dice fabrication costs; "manufacturing, validation, and design costs". mb
  • Khanan - Thursday, April 7, 2022 - link

    Last time I checked 5950X had still more performance despite being challenged by a 250W auto OC 12900K. I mean, imagine, you can easily activate auto OC on 5950X with activating PBO too. And yep it crushes the 12900K then. Reviews that don’t reflect this are as per usual kinda trashy.
  • mode_13h - Wednesday, March 30, 2022 - link

    > They have 90-95% of the performance for less than half, or in some cases, 1/3rd the power.

    I don't see how you can draw that conclusion, when the article only lists *peak* power. What's needed is joules per test, for the fixed-size tests.

Log in

Don't have an account? Sign up now