AMD Ryzen 9 4900HS offers longer near-peak performance compared to the Intel Core i7-10875H under sustained load but suffers from an overall higher dip in scores

[Redaktion]

A preliminary sustained performance comparison between the AMD Ryzen 9 4900HS and the Intel Core i7-10875H in a looped Cinebench R15 Multi test shows that the Renoir chip can better hold on to its peak performance in the first few rounds compared to Comet Lake-H. What is interesting is that the 10th generation Intel chips have managed to nearly halve the second round-performance loss compared to the 9th generation chips.

https://www.notebookcheck.net/AMD-Ryzen-9-4900HS-offers-longer-near-peak-performance-compared-to-the-Intel-Core-i7-10875H-under-sustained-load-but-suffers-from-an-overall-higher-dip-in-scores.461668.0.html

[The Scott]

Where are the scores for AMD's 4900HS? You provide the Intel scores for various rounds of Cinebench R15 but not AMD. Why? This makes it difficult if not impossible to directly compare Ryzen 3 mobile to Comet Lake. A larger percentage dip doesn't matter if AMD is still outperforming Intel.

This is very shoddy benchmarking and journalism.

[Vaidyanathan]

Where are the scores for AMD's 4900HS? You provide the Intel scores for various rounds of Cinebench R15 but not AMD. Why? This makes it difficult if not impossible to directly compare Ryzen 3 mobile to Comet Lake. A larger percentage dip doesn't matter if AMD is still outperforming Intel.

This is very shoddy benchmarking and journalism.

Lol wut? Did you check the graph? The 4900HS tops it. Hover your mouse over the legend below and the respective curves will be highlighted. For even more details, the review notebook is hyperlinked as well.
Thanks.

[The Scott]

Sorry. I guess I overlooked the graph.

May I suggest including the numbers in the text paragraph next time?

I have a visual disability and my reading program handles text well but not graphs.

Again, I apologize.

Where are the scores for AMD's 4900HS? You provide the Intel scores for various rounds of Cinebench R15 but not AMD. Why? This makes it difficult if not impossible to directly compare Ryzen 3 mobile to Comet Lake. A larger percentage dip doesn't matter if AMD is still outperforming Intel.

This is very shoddy benchmarking and journalism.

Lol wut? Did you check the graph? The 4900HS tops it. Hover your mouse over the legend below and the respective curves will be highlighted. For even more details, the review notebook is hyperlinked as well.
Thanks.

[Vaidyanathan]

Sorry. I guess I overlooked the graph.

May I suggest including the numbers in the text paragraph next time?

I have a visual disability and my reading program handles text well but not graphs.

Again, I apologize.

Actually the primary numbers have been indicated along with percentage changes.
No worries, mate. Thanks for taking the time to read. Hope it was useful. :)

[Focusonskills]

Isn't this comparison unfair? A 17" laptop is compared with a 14" laptop.

[Padmakara]

Please include also the power consumption of the cpus.
Than it will be a great article.
Also down the graph, it should be first cpu, second as score, not mixed.
That's why probably the other guy didn't understand.
And the test should include r20 and not r15. Cinebench R20 uses a much larger and more complex test scene than R15, requiring about 8x the computational power needed to render it.  All the pros reviewers use it.
Than it will he a professional review.
Cheers!

[Padmakara]

Isn't this comparison unfair? A 17" laptop is compared with a 14" laptop.

Probably the sustained higher clock of weel cooled 17" will translate into 10-12% higher performance than a 14" same cpu

[Vaidyanathan]

Please include also the power consumption of the cpus.
Than it will be a great article.
Also down the graph, it should be first cpu, second as score, not mixed.
That's why probably the other guy didn't understand.
And the test should include r20 and not r15. Cinebench R20 uses a much larger and more complex test scene than R15, requiring about 8x the computational power needed to render it.  All the pros reviewers use it.
Than it will he a professional review.
Cheers!

Hi. Thanks for the feedback. Appreciated. Let me clarify on some of those.
1. Firstly, this is just a preliminary comparison. I am not concluding anything just yet. The idea I wanted to convey was, if you were to buy a 4900HS or 10875H laptop, what is the pattern of sustained performance you can expect to get.
2. The graph is the same layout we use for all reviews. It helps that the laptop name is mentioned first as this test is highly laptop design dependent. As I've alluded to in my article, there is a possibility of a better curve for the 4900HS if it were to have been in a more roomy chassis than the Zephyrus G14.
3. CB20 hasn't completely replaced CB15. Also, if you look at the previous CPU benchmark article, I have included both CB15 and CB20 scores. Using CB15 helps in this case as most of our previous comparisons have been using this test and if someone wants to add an older model, it becomes easy. Moreover, CB15 is perfectly fine for an 8 core chip ;)
4. Power consumption details will follow in individual reviews for the aforemoentioned notebooks. They are in progress.

Thank you :)

[Vaidyanathan]

Isn't this comparison unfair? A 17" laptop is compared with a 14" laptop.

Well, yes and no. Right now, we don't have any 17-inch 4900HS models yet. So, we are limited with what we can compare with. You can safely presume that a 17-inch laptop with a similar cooling as the Zephyrus G14 would probably offer even more rounds of near-peak perf and a reduced perf delta after that.
Also, while size plays an important role, it may not always translate as de facto higher performance though it is the case in theory. For example, we didn't find much diff in CB15 scores between a 15-inch Aero 15 and a 17-inch Aero 17 running the same CPU although I do anticipate some benefits for the latter in loop tests like this.

[Padmakara]

Please include also the power consumption of the cpus.
Than it will be a great article.
Also down the graph, it should be first cpu, second as score, not mixed.
That's why probably the other guy didn't understand.
And the test should include r20 and not r15. Cinebench R20 uses a much larger and more complex test scene than R15, requiring about 8x the computational power needed to render it.  All the pros reviewers use it.
Than it will he a professional review.
Cheers!

Hi. Thanks for the feedback. Appreciated. Let me clarify on some of those.
1. Firstly, this is just a preliminary comparison. I am not concluding anything just yet. The idea I wanted to convey was, if you were to buy a 4900HS or 10875H laptop, what is the pattern of sustained performance you can expect to get.
2. The graph is the same layout we use for all reviews. It helps that the laptop name is mentioned first as this test is highly laptop design dependent. As I've alluded to in my article, there is a possibility of a better curve for the 4900HS if it were to have been in a more roomy chassis than the Zephyrus G14.
3. CB20 hasn't completely replaced CB15. Also, if you look at the previous CPU benchmark article, I have included both CB15 and CB20 scores. Using CB15 helps in this case as most of our previous comparisons have been using this test and if someone wants to add an older model, it becomes easy. Moreover, CB15 is perfectly fine for an 8 core chip ;)
4. Power consumption details will follow in individual reviews for the aforemoentioned notebooks. They are in progress.

Thank you :)

Thanks. I thought you compare the cpus sustained performance and not the laptops performance, because in the title was R9 and  i7 and Not the laptop brands, that's why I was saying about the graph, also the guy before me misinterpreted your graph.
And there are some big differences between r15 and r20 also in the cpus performance. The  r20 is more complex and new than r15. Both are ok but R20 is more representative about the cpu performance more than >4 cores. If not the % performance difference between cpus would be the same in same tests. Which it isn't.
And of course a 17" has a much better cooling than a 14 but we're not talking about that here.
And when comparing the performance is good to compare also the power draw because is cpu comparison. So also the power draw should be regarded here. Not talking about laptops. If you compare cpu performance than also power W should be mentioned. Is very important for laptops buying.
All the best.
Namaskar

[Josh Miller]

Please include also the power consumption of the cpus.
Than it will be a great article.
Also down the graph, it should be first cpu, second as score, not mixed.
That's why probably the other guy didn't understand.
And the test should include r20 and not r15. Cinebench R20 uses a much larger and more complex test scene than R15, requiring about 8x the computational power needed to render it.  All the pros reviewers use it.
Than it will he a professional review.
Cheers!

Hi. Thanks for the feedback. Appreciated. Let me clarify on some of those.
1. Firstly, this is just a preliminary comparison. I am not concluding anything just yet. The idea I wanted to convey was, if you were to buy a 4900HS or 10875H laptop, what is the pattern of sustained performance you can expect to get.
2. The graph is the same layout we use for all reviews. It helps that the laptop name is mentioned first as this test is highly laptop design dependent. As I've alluded to in my article, there is a possibility of a better curve for the 4900HS if it were to have been in a more roomy chassis than the Zephyrus G14.
3. CB20 hasn't completely replaced CB15. Also, if you look at the previous CPU benchmark article, I have included both CB15 and CB20 scores. Using CB15 helps in this case as most of our previous comparisons have been using this test and if someone wants to add an older model, it becomes easy. Moreover, CB15 is perfectly fine for an 8 core chip ;)
4. Power consumption details will follow in individual reviews for the aforemoentioned notebooks. They are in progress.

Thank you :)

If you look at Ultrabookreview, their G14 with the 4800hs has significantly lower initial scores and a more gradual drop to roughly the same level your 4900hs sustains. Their two 4800h systems have slightly lower first run scores and minimal drop, with significantly higher sustained scores than the HS models. The two 4800h systems also have much higher cpu temperatures. That suggests the performance is tdp limited, not temperature limited.

[havefun]

So Asus Zephyrus G14 GA401IV  outperform all these notebooks, and is 500-1000 euro cheaper?

[A]

What about single core sustainability?

[Atul88]

So Amd still managed to beat intel even the dip in 2nd round was more than intel. Looks like your forgot to mention that. And what about power consumption throughout the test and where are the amd scores? You were kind enough to show intel scores. Right??