The Ice Lake Benchmark Preview: Inside Intel's 10nm
by Dr. Ian Cutress on August 1, 2019 9:00 AM EST- Posted in
- CPUs
- Intel
- GPUs
- 10nm
- Core
- Ice Lake
- Cannon Lake
- Sunny Cove
- 10th Gen Core
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.
On AVX-512, the Ice Lake part destroys the competition.
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.
261 Comments
View All Comments
jcc5169 - Thursday, August 1, 2019 - link
It's like a long running Intel advertisement ......Ian Cutress - Thursday, August 1, 2019 - link
How so?jordanclock - Thursday, August 1, 2019 - link
Probably because you provided objective measurements instead of long winded conjecture.peevee - Friday, August 2, 2019 - link
"As for Ice Lake itself, our results lean towards Ice Lake outperforming Whiskey Lake, if only by a small margin."A laptop allowing more than 50W on a U chip and with cooler always running is certainly not representative of the real world performance a user would get from real systems. If Whiskey Lake results were obtained on a real system, then the results are not comparable enough that Ice Lake can actually be slower, especially in real world loads (note the performance on x264 and POV-Ray essentially has not changed).
But at least iGPU has improved nicely.
Samus - Saturday, August 3, 2019 - link
They have dedicated pages to 15w and 25w results. Not sure what you're on about about 'allowing more than 50W...' when they power limited the chip for specific-scenario benchmark results.Spunjji - Monday, August 5, 2019 - link
You may want to re-read that Power Results section - put simply, peevee's point still stands and your response is mistaken. At the 25W limit, the CPU spent more than 20 seconds operating between 50W and 35W before it finally dropped down. It's unlikely that the Coffee / Whiskey devices they tested allowed that sort of behaviour, as without a continuous 100% fan speed the thermals would have spiked and lowered the power limit far sooner. Even with the "power limits" you're talking about in place, the chip still boosts like that, so their benchmark results incorporate that behaviour. for any other company I might feel generously inclined, but Intel have already pulled a fast one multiple times by releasing CPUs that have nominally higher specifications but never reach their performance potential in shipping designs due to thermal constraints.Gondalf - Wednesday, January 8, 2020 - link
Spec suite (the reference) is a pretty loooong bench so the cpu stay at 15W all the time without your 50W. Your comment say a lot about you.yeeeeman - Wednesday, June 17, 2020 - link
Also, we not comparing power efficiency here but rather IPC of each core. And the bottom line is that Sunny Cove core has 5-10% better IPC than Zen 2.CSMR - Sunday, August 4, 2019 - link
There is no evidence that the iGPU has improved. We'd need a comparison against Iris Plus 655 to confirm that. We only know EUs have increased (64 vs 48) but EDRAM was removed.brakdoo - Thursday, August 1, 2019 - link
"First of all, I must say that Intel offering us to test a reference system in advance of a launch is a very good thing indeed"You are being played once again as a marketing tool for the Ice Lake launch. Don't tell me you don't see that.