System Results (15W)

When testing a laptop system, there are various angles to consider on how to test: either user experience benchmarks, that are mostly single threaded and give a good boost to how systems implement a deal of turbo, or sustained benchmarks that test how the system performs when you push it. Intel has gone out of its way to emphasise the former for the next generation of mobile CPUs: they would prefer that reviewers stick to very user experience-like tests, rather than say, rendering programs. The problem there is that outside a number of canned benchmarks, it can be difficult. Users, and especially creators, that typically spend a lot on a premium device, might actually be doing sustained benchmarks.

Given the time that we had to test, we were actually limited in what we could arrange.

Application Loading (GIMP 2.10.4)

3DPM v2.1 (non-AVX)

3DPM v2.1 (AVX2 / AVX-512)

On AVX-512, the Ice Lake part destroys the competition.

Blender 2.79b (cpu-bmw27)

POV-Ray 3.7.1

CineBench R20 ST

CineBench R20 MT

7-zip 1805 Combined

WinRAR 5.60b3

AES Throughput (minus AES instructions)

These last two tests are typically our more memory sensitive tests, and the LPDDR4X-3733 really does win out over the LPDDR3-2133 in the other systems.

Power Results (15W and 25W) Synthetic and Legacy Results (15W)
Comments Locked

261 Comments

View All Comments

  • Fulljack - Thursday, August 1, 2019 - link

    lpddr4 would make it higher priced and stuck in premium device, as it's a soldered ram.
  • RSAUser - Friday, August 2, 2019 - link

    I think the commenter means DDR4L, low powered SO-DIMM and not the usual phone/tablet soldered RAM.
  • Rudde - Thursday, August 1, 2019 - link

    There are rumors (check AT forums) about a Gen 12 (ice lake is 11) discrete gpu with 512 EUs (arctic sound). I believe 64 EUs is confirmed max on Ice lake, but the generation after that (Tiger lake?) might have more.
  • Mr Perfect - Thursday, August 1, 2019 - link

    I notice that these parts all have Hyper threading. Has Intel addressed their security concerns with HT in Ice Lake, or do they see the benefits of HT to be greater then the risks in the mobile space?
  • djayjp - Thursday, August 1, 2019 - link

    And what about Zombieload...? Or was that just quietly swept under the rug?
  • justaviking - Thursday, August 1, 2019 - link

    QUOTE: "After attending the event, to which fewer than 10 press were invited, I now understand why. Some of the press invited didn’t have OS images, didn’t bring benchmarks with them, and were quite happy to go along with the flow. Intel provided benchmarks like Geekbench and 3DMark, which those press with their audiences were happy to run. I came prepared with both a new 1903 OS image and my benchmark suite, ready to rock and roll."

    And that is why we love AnandTech. :)
  • justaviking - Thursday, August 1, 2019 - link

    And QUOTE: "It was clear that some of the press in attendance only needed a day (or half a day), but for what we do at AT, then two days would be better."

    Thank you for the article, and for doing the best you could in the time available.
  • ManDaddio - Thursday, August 1, 2019 - link

    Well I'm glad that we got a little bit more looking to this part of Intel's 10nm. I'm looking forward to the next iteration to be honest.

    Actually I was more interested in the features of these new Intel laptops.

    The benchmarks as you stated are just surface stuff until you can actually truly spend a while testing these things when they come out.

    All the rave right now in some circles is AMD but I always liked Intel laptops for their features as well as performance.
    And I have owned a couple AMD laptops. They were good and did the job but my Intel laptops were always much better. That's my experience.

    I understand the challenges of putting an article like this out to the public who complains a lot or just wants to troll. Thanks for sharing.
  • Bulat Ziganshin - Thursday, August 1, 2019 - link

    >AMD for Zen 2 decided to halve the L1-D, double increase the L1-D associativity

    it was L1-I cache
  • Drumsticks - Thursday, August 1, 2019 - link

    In terms of pure performance, Icelake is definitely let down by it's available frequency. It's in bad need of a 10nm++ process or a backport to 14nm.

    It was a good idea to release it for laptops though, I think. Laptops definitely are about more than just the CPU, and the overall experience the Icelake featureset enables might actually be a good selling point.

    I'd love to see an Icelake based microarchitecture released on Intel 14+++ though. Even though 14+++ has been around forever, I think it's still a good node. It's still completely untouched by any other node in frequency, power consumption isn't as important to desktop users, and the node must be so optimized by now that costs are good even with bigger dies.

Log in

Don't have an account? Sign up now