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

Power

(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.

Office

(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.

Science

(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
POST A COMMENT

196 Comments

View All Comments

  • mode_13h - Thursday, March 31, 2022 - link

    Somewhat off-topic: the top of my page is serving me an from ASML, indicating they're seeking optical engineers to work on EUV.

    Crazy. You'd think it would be a dream job for anyone with the proper skillset, but I guess their growth is outpacing the available talent pool.

    Well, if anyone reading this thread knows an optical engineer, you might tell them to go have a look at ASML. I didn't check to see whether relocation would be necessary, but perhaps that's the catch.
    Reply
  • mode_13h - Thursday, March 31, 2022 - link

    "... serving me an ad from ASML, ...", I meant to say. Reply
  • milli - Thursday, March 31, 2022 - link

    Peak power consumption is handy data but not very useful by itself.
    How about average power consumption during applications / games / ...
    You also used to represent power consumption to complete a task, now that was useful data.
    Reply
  • eloyard - Thursday, March 31, 2022 - link

    Peak power is essential for understanding how robust power delivery system and heat sink need to be.

    "Average is the only important thing" is typical Intel's bs pushing cost of more expensive components different than processor (mobo, psu, cooler) to the consumer.
    Reply
  • mode_13h - Friday, April 1, 2022 - link

    > Peak power is essential for understanding how robust power delivery system and heat sink need to be.

    Yes and no. It really depends on the minimum duration over which a "peak" is defined. If we're talking about a couple milliseconds, then absolutely not. If we're talking about a couple seconds, then yes.

    > "Average is the only important thing" is typical Intel's bs ...

    I understand your point, but it's still relevant for people trying to determine which CPU will deliver the most performance per unit of energy. With increasing energy prices and the warmer months approaching those of us in the northern hemisphere, that's not merely a theoretical concern.
    Reply
  • milli - Friday, April 1, 2022 - link

    Your post oozes fanboism.
    Firstly, I didn't dismiss peak power consumption (as you assume) but by itself it's useless. The graph doesn't even state how long it's holding this peak power consumption.
    Secondly, where do I say "Average is the only important thing". Right, nowhere.
    But if a CPU peaks 200W for 1s and then goes to 65W, I would like to know that. That's important data. It's data that Anandtech used to provide but stopped doing.
    Reply
  • Mike Bruzzone - Thursday, March 31, 2022 - link

    A look at Alder Lake Production data;

    i9s = 55.01% of all Alder Lake with 12900K @ 43.56% and KF @ 2.07% and 900 @ 8.92%

    i7s = 34% of Alder Lake with 12700K @ 29.42% and KF @ 3.82%

    i5s = 9.3% with 600K @ 3.13%, 600KF @ 1.5% and 400 @ 2.63%

    i3s = 1.13% with no SKU greater than 1% of full line volume.

    Pentium = 0.08%
    Celeron = 0.45%

    Alder Lake full line can be characterized, essentially, all i9 12900K and i7 12700K and remainder fall out from sort. Similar Coffee Refresh where most of full line volume are disabled i9s. Cores disablement is discouraged on industrial management best practice subject manufacturing cost optimization not incorporating the added production step to purposely disable that at bottom bin can drive cost > price, despite most bottom bin given away anyway cost : price averaged across the total sales package.

    Analyst suspects OEM volume procurement price for Alder mirroring exactly the category percent supply is range $254 to $239 each per i9/i7 component calculated $1K Average Weighed Price of i9/i7 category / 2 and that all i5/i3/P&C are thrown in for free as sales package 'close' incentive.

    Net cost of Alder at $1K AWP < Cost of Sales, < Cost of R&D, < Tax Accrual = $214.
    Average Marginal Cost of production = $63.34 per component
    Average Marginal Cost of fabrication = $25.88 per component

    Mike Bruzzone, Camp Marketing
    Reply
  • Mike Bruzzone - Thursday, March 31, 2022 - link

    Adding on Alder, there are supposedly two mask sets. The first a 215 mm2 full P+E and the other a cost optimized for area 116 mm2 P only. Not much has been said can anyone confirm these two mask sets and die area?

    mb
    Reply
  • dwade123 - Thursday, March 31, 2022 - link

    Already sold my 5800x for a 12900k. I will never buy AMD again and forced to beta test their crap for them for free. Reply
  • mode_13h - Friday, April 1, 2022 - link

    > I will never buy AMD again

    AMD is still growing and maturing. With that, their quality control should hopefully improve.

    About 11 years ago, Intel suffered a similarly embarrassing problem, with the SATA controller in their Sandybridge motherboard chipsets. Had you forsworn never to buy Intel again, you now be up a creek without a paddle.

    https://www.anandtech.com/show/4142/intel-discover...
    Reply

Log in

Don't have an account? Sign up now