Alder Lake Benchmark Results Surface, but Beware the Fine Print

There are some new leaked benchmark results for Alder Lake and the Core i9-12900K. Even better, they come from Puget Systems, an OEM reseller and benchmark developer that serves the high-performance video editing and content-creation crowd. We’ve used Puget Systems benchmarks in the past — they do good work — and it’s great to see some results in a benchmark explicitly intended to mimic real-world workloads as opposed to a more synthetic score.

The system in question was benchmarked using Adobe After Effects. Adobe After Effects is not tuned for high core count systems and does not particularly benefit from additional cores above the eight-core mark. Earlier this year, Puget compared the Ryzen 5000 and Intel 11th Gen families. Those results show AMD and Intel both gaining about 7 percent more performance when shifting from six cores to eight. There was no discernible benefit to more than eight cores.

Adobe After Effects does respond well to IPC gains, however. The Core i7-10700K (8C/16T) scored a 1243, while the Core i7-11700K (8C/16T) scored 1407. That’s a 13 percent performance improvement when the 11th Gen CPUs maximum clocks are very slightly lower than the 10700K (5.0GHz versus 5.1GHz). Given that Alder Lake is thought to maintain approximately the same frequencies as Rocket Lake and After Effects lack of response to higher core counts, we can assume that any performance boost in this application comes courtesy of better IPC or an improved ability to maintain maximum clock.

Here’s a reference slide from a different Puget Systems After Effects article, comparing performance in a number of recent AMD and Intel solutions. This graph gives some additional context about how Intel and AMD platforms perform in this test overall:

Now that we’ve discussed the particulars of the benchmark in more detail, let’s talk about these leaked results. WCCFTech spotted these figures. While they’ve since been removed, we can evaluate what the score means for future systems:

A GPU score of 156.7 is only barely faster than the Core i9-11900K’s score of 153.1. The 10900K scored 126.5 in this benchmark, while the 5800X scored 141.7. It looks as though Intel might continue to lead this subtest, but not by dramatically more than it currently does. Overall, Intel leads AMD on this test by 8 percent and its own Rocket Lake by 2 percent.

A RAM preview score of 138 is slightly faster than the 128 reported by the 11900K or the 5800X’s 125.9. The 5950X actually outscores the 5800X in this test, at 130.8. That’s a 5.5 percent lead for Alder Lake in this specific subtest.

A render score of 137 is 12 percent higher than the Core i9-11900K’s 122.1 and the Ryzen 7 5800X’s 118.5. The Ryzen 9 5950X closes the gap here with a 126.2, but that’s not quite enough to tie things up. Intel, again, leads by about 8 percent. I’m using comparison figures published by Puget itself, which means our numbers are slightly different than those reported by WCCFTech.

Finally, there’s the tracking score of 196.7. First, the good news. For Intel users on older platforms, this is a significant bump of about 11 percent. It’s not quite enough to catch the 5800X, however, which posts a 214.1. The overall reported score of 1573 is faster than the recently reported 1528 for AMD (5800X) or the Core i9-11900K (1426). This works out to roughly a 1.1x gain for Intel platforms.

Ryzen’s current dominant position in A:AE is a very new phenomenon. Back in 2020, before the launch of the 5000 series, Intel hardware was neck and neck with its AMD counterpart. Many Adobe applications have generally favored Intel for years, and while that’s changed recently, it’s still been the dominant status quo. When you look at these Alder Lake benchmarks versus Ryzen, keep in mind that the test does not scale well above eight cores. What we’re seeing here is a narrow victory for Intel, but in a test that caters to lower CPU core counts at higher clock speeds. It’s a perfectly valid point of comparison, but it affords AMD’s high-core-count CPUs no room to stretch their metaphorical legs.

In aggregate, these test results show modest generational improvement for Intel. AMD will likely gain something off its V-Cache improvements dropping later this year, but we don’t know how much.

Continue reading

Linus Tovalds Blames Intel for Killing ECC RAM in Consumer Systems

Intel stripped ECC RAM support off its consumer products over a decade ago, and Linus Torvalds is still unhappy about it.

Reusing Your Current Intel Heatsink for Alder Lake May Be a Bad Idea

There may be good reason to upgrade your CPU cooler when Alder Lake ships, even if conversion kits are available. We'll have to see how well they work.

Alder Lake Leak: Intel Core i9-12900K Offers 5.3GHz Boost Clock, Lower PL2

Specifications for Intel's upcoming Alder Lake have leaked, and they imply a CPU that's at least somewhat more power-efficient than its predecessors without giving up any top-end clock.

What Kind of Performance Should We Expect from Intel’s Alder Lake?

What kind of performance should we expect from Alder Lake? Recent rumors have been all over the map.