Intel’s Core i7 and Core i9 H-Series chips are no longer a slam-dunk choice for CPUs to power high-end gaming and multimedia-content-creation laptops. There’s a new kid on the block: AMD Ryzen 7 and Ryzen 9 H-Series processors. And they just got even more powerful.
Our early performance tests suggest that the new fifth-generation Ryzen chips (code-named “Cezanne” during their development), most of which use AMD’s latest “Zen 3” processor architecture, outperform their predecessors in niche areas important to laptop power users. This is especially true when it comes to the performance of each single processor core operating individually. And the fourth-generation Ryzen H-series were already giving comparable Core i7 and Core i9 chips a run for their money in overall performance. Add the improvements, plus the commanding starting point of “Zen 2,” and you get some formidable bits of silicon that should excite laptop shoppers of all kinds.
The early tests are based on just a single Ryzen 9-powered laptop, and it’s too early to tell how any of AMD’s new silicon will compare with Intel’s yet-to-be-released 11th Generation Core i9 flagship chips. Intel and AMD together expect hundreds of new H-series laptops based on the latest chip generations to hit the market this year, so the situation could change as we test more hardware.
But for now, the picture is simple. If the high-performance laptop you’re eyeing offers a choice of the latest AMD or Intel H-series processors, you may well want to select the former; AMD is no longer to be dismissed as an also-ran in high-powered mobile CPUs. To better understand why, let’s look at the new innovations available in AMD’s fifth-generation Ryzen 5000 mobile processors, and then get into how they stack up on some benchmark tests.
The New Lineup: The Ryzen 5000 Mobile Family
The most important thing to understand about AMD Ryzen and Intel Core CPUs for laptops is the dizzying array of chip models in a given generation. And since new generations tend to appear at least once a year, this can make for some confusing spec comparisons. To simplify things for the latest AMD Ryzen generation, here’s a handy chart showing all the models that are available for laptop makers to put in their new designs:
Why are there 13 chips? A few key differences among them make some of the models suited for beefy gaming rigs and others suited for thin-and-light designs. First, look at the letters appended to each of the model names. Anything with an “H” consumes more power—seen in the thermal design power (TDP) column—and requires more robust cooling hardware than an equivalent chip with a “U” at the end of the model name.
For example, the Ryzen 7 5800U needs a maximum of 15 watts, while the otherwise-similar Ryzen 7 5800H needs a maximum of 45 watts. The latter is intended for workstations, gaming rigs, and other powerful laptops, while the former is for ultraportable notebooks that are mainly destined for web browsing, spreadsheet editing, and the like.
To ensure that H-series chips are further suited to the capabilities of various laptops, AMD adds more variation than it does with the U-series models. The HX-series chips can avail themselves of more than 45 watts, while the H-series ones max out at 45 watts. And the HS-series processors are intended for the burgeoning category of slim, sleek gaming laptops that use 35 watts.
Zen 3 Architecture Improvements
While the general power requirements that differentiate these laptop processors haven’t changed much from one Ryzen generation to the next, the switch to the Zen 3 architecture does confer some noteworthy benefits.
First, you’ll note from the chart that every fifth-generation chip listed supports multi-threading. This means that each CPU core can process two instruction threads at once, which can mean a dramatic speedup on modern apps that scale well with as many cores and threads as they can access. Some Ryzen mobile chips in previous generations do not support multi-threading.
Second, Zen 3 cores are simply more advanced than their predecessors. They use a 7-nanometer (7nm) production process AMD refers to as “Cezanne,” while Intel’s latest 11th Generation chips continue to use a 10nm “Tiger Lake” process. Smaller chip-fabrication processes usually mean added performance and greater operating efficiency. Previous AMD architectures lagged their Intel counterparts slightly on the performance of each individual core, which is important for running some types of apps efficiently, especially graphics-intensive games and some legacy apps that don’t do much or anything in the way of multi-threading. AMD claims that single-core performance is up to 23% better on Zen 3 cores compared with Zen 2 cores operating with the same TDP.
Also important to evaluating a processor’s efficiency is how well all of its various components work together to meet a PC’s processing demands. Memory latency and cache are critical here, and AMD says it has improved both. Each Zen 3 model has an increased 20MB of total cache. Of that, 16MB is L3 cache accessible to each of the processor cores. It’s a more efficient and plentiful arrangement than the 8MB of L3 cache in Zen 2, only half of which is directly accessible to each core.
AMD says it has also improved the memory controller in Zen 3 and made adjustments to the processor power management. A new “sleep state” can reduce the available memory when it’s not needed, and instantly power it back up when it is. And the processor itself ditches traditional power states in favor of a new infinitely adjustable regime that lets it select any voltage and frequency it feels is most appropriate for the task at hand. Graphics capabilities get a boost in Zen 3, too. The Ryzen chips still use integrated Radeon graphics processors, but now they come with higher clock frequencies, which can boost performance by about 15% when a laptop is running on the integrated Radeon graphics chip (instead of a discrete GPU).
Ryzen 9 5000 Series: The First Benchmark Results
To independently check these and other performance claims that AMD makes about the new Ryzen models, we ran a few benchmark tests on the brand-new Asus ROG Flow X13 laptop, one of the first to come with the Ryzen 5000-series CPU models. (Hit the link for a full review of the Flow X13.)
Our test laptop has the flagship of the 35-watt Ryzen 5000 chips, the Ryzen 9 5980HS, which operates at clock speeds up to 4.8GHz. The ROG Flow X13 is an archetypal slim gaming laptop, prioritizing space savings over all-out processing muscle. To that end, it has a relatively pedestrian graphics chip, the Nvidia GeForce GTX 1650. You can see how its specs compare with those of a few key competing laptops in the chart below…
The competitors include the Asus ROG Zephyrus G14, which represents the best of what AMD’s fourth-generation Zen 2 chips have to offer. Three Intel-based machines round out the list. Two are gaming laptops with Core i7 chips. The Gigabyte Aero 15 has Intel’s previous 10th Generation H-series Core i7, while the Asus TUF Dash F15 (debuting today alongside the Flow X13 and AMD’s Ryzen 5000 mobile line) has the latest 11th Generation Core i7 model, the Core i7-11375H.
Finally, the Dell Precision 3551 workstation represents Intel’s Core i9 H-series flagship, from the 10th Generation. Remember, 11th Generation Core i9 laptops aren’t available yet.
Cinebench R15, POV-Ray 3.7 and 7-Zip
Straightforward CPU tests like Maxon’s Cinebench R15 tool scale well with increasing numbers of cores and threads, giving us a bird’s eye look at CPU performance. With all cores churning, there are few surprises. The eight-core/16-thread Ryzen 9 5980HS takes the lead, followed closely by its predecessor. The Intel-based machines are all distinctly behind on this test.
Limiting Cinebench to running on a single core reveals some surprises, however. The Ryzen 9 5980HS does indeed demonstrate significantly improved single-core performance over its predecessor, but the difference between it and the Core i7-11375H is far smaller. (The Core i7-11375H and its 5GHz Turbo Boost Max 3.0 boost clock also outpaced the previous-generation Core i7 H-Series shown here, the Core i7-10875H, on single-core performance.) And remember that when the 11th Generation Core i9s arrive, they will almost certainly have better single-core performance than the 11th Generation Core i7 represented here.
Cinebench uses CPU cores to render a 3D image on the laptop’s screen, while POV-Ray uses them to render ray-traced graphics. The output is different, but the relative test results are often parallel when all CPU cores are used. Run on all cores, POV-Ray shows no improvement between the two Ryzen generations, but all of the Intel machines are significantly slower. The Ryzen 9 5980HS beats the field on the single-core test.
Simpler tasks like compressing folders into a zipped file also depend on a CPU’s capabilities. The in-app benchmark of the 7-Zip compression utility helps us approximate this task. There isn’t a huge leap forward between Ryzen generations, but the Core i7-11375H’s low score is especially notable. Likely it’s down to the Tiger Lake-H chip’s four cores and eight threads on a test that uses all the cores it can get; the Ryzen 9 chips have eight and 16, respectively. No match.
Handbrake 1.1.1 and Adobe Photoshop CC
Our final two tests approximate tasks that pretty much every power user engaged in content creation will need to perform on a laptop at some point: video transcoding and image editing. We test video encoding by transforming a 4K clip to a 1080p one using the Handbrake app, which is highly CPU-dependent. While there’s no meaningful improvement between the Ryzen generations, they far outperformed their Intel competitors. The TUF Dash again trailed the field; no surprise, as Handbrake eats all the cores and threads it can for every meal.
While it’s not strictly a pure CPU test, we’ve also included the results of our Photoshop image editing benchmark below. In this trial, we time how long it takes to apply a series of filters and effects to a standard test JPG. It is somewhat dependent on a laptop’s graphics processor and is sensitive to RAM amounts, so it’s not 100% relevant to a CPU comparison. Still, it shows a dramatic improvement in the Ryzen 9 5980HS that mirrors the Cinebench single-core results.
Will Core i9 Take on Ryzen 9?
The test results above aren’t always consistent, but they do suggest that the latest Ryzen 9 is more capable than its predecessor in most of the ways that AMD says it is. But we should note some caveats.
In addition to missing Intel’s latest-generation Core i9, which will be the direct competitor to the Ryzen 9 5000-series models when it’s introduced later this year, there’s no look at gaming performance. As nice as it is to have gaming-friendly features like improved cache and single-core performance, laptop gamers should care more about the GPU and thermal capabilities of their machines than how well the CPU performs in a raw-CPU-performance sense. There’s also too much variability in gaming laptops’ GPUs, RAM amounts, and thermals to draw conclusions on the Ryzen 9 here without a much bigger pool of tested Ryzen 5000 laptops from which to draw data. As a result, we’re not using gaming benchmarks or actual games in this sneak peak of CPU performance. If you’d like to see how each of the gaming laptops mentioned above matches up against its key competitors on real-world titles and graphics benchmarks, head on over to the full reviews, summarized here one more time:
We’re looking forward to revisiting how H-series AMD and Intel chips compare once the latest generations are available in more laptops later this year. For now, this early look at Zen 3 in the Ryzen 5000 mobile line suggests that AMD is just as ascendant as ever when it comes to giving high-performance Intel silicon a run for its money.
