Apple: Performance claims about the iPad Air (2024) were accurate, despite GPU core count mistake
- D
- Dom
- raM
- 06 Jun 2024
Anonymous, 06 Jun 2024Both the CPU and GPU are important. The GPU is used to rend... moreI guess that’s true, but nanoreview scores still seem to be accurate based on how older processors have aged. For example the A12 and Snapdragon 865 both perform around the same in 2024, with neither being crazy fast, but both still feeling decently snappy. Nanoreview gave both of these processors a score of 57, which makes sense, but just by looking at the Antutu scores, it would seem like the Snapdragon 865 is a much better processor because it scored over 10,000 points higher even though both processors are about as good as each other.
- ?
- Anonymous
- 70d
- 06 Jun 2024
Dom, 06 Jun 2024Why shouldn’t cpu be considered more important than gpu in ... moreBoth the CPU and GPU are important. The GPU is used to render graphics. The CPU can be used to render graphics, but it's much slower than the GPU. The CPU is best used for non-graphics computations. Web browsers use the GPU to render graphics. You can look up WebGL and WebGPU.
- D
- Dom
- AAL
- 06 Jun 2024
Anonymous, 06 Jun 2024This website uses AnTuTu 10 and Geekbench 6, and it's ... moreWhy shouldn’t cpu be considered more important than gpu in ratings? I bet most people would rather have a phone that feels snappier opening apps and loading webpages than one that gets more frames playing games like Genshin Impact. And for those people, a faster cpu will also mean their phone will last longer. Plus, nanoreview scores also factor in battery life, which is important for longevity.
- ?
- Anonymous
- 70d
- 06 Jun 2024
Dom, 05 Jun 2024That’s why in my original comment I provided this link: ht... moreThis website uses AnTuTu 10 and Geekbench 6, and it's unclear how the ratings are determined. In any case, if the ratings are determined by combining these results, it's likely that SoCs that have higher CPU scores and lower GPU scores will be ranked higher than SoCs that have lower CPU scores and higher GPU scores. By the way, according to AuTuTu, the scores of Android and iOS versions are not comparable, which is silly. It's like saying that you can't compare the performance of the PS5 and the Xbox Series X because they use different OSes and APIs. Anyway, the Snapdragon 8 Gen 2 beats the A16 in AnTuTu 10 test which is a compound test that represents the overall performance.
- A
- Android-Never-Again
- K6X
- 06 Jun 2024
Anonymous, 05 Jun 2024Send it. But try not to ruin the canisters while you are ha... moreLmao says the person who's posted the most in this thread; a thread about an article on a brand they don't even like.
You Apple haters have Apple living rent-free in your head. It's so cute.
- D
- Dom
- raM
- 05 Jun 2024
Anonymous, 05 Jun 2024Geekbench multi-core/single-core tests are CPU tests. They ... moreThat’s why in my original comment I provided this link:
https://nanoreview.net/en/soc-list/rating
which combines data from multiple benchmarks and devices to provide an overall processor ranking
- ?
- Anonymous
- 70d
- 05 Jun 2024
Dom, 05 Jun 2024The highest scoring phone from this article with an A16 sco... moreGeekbench multi-core/single-core tests are CPU tests. They don't represent the overall performance. 3DMark tests are GPU tests. The A16 outperforms the Snapdragon 8 Gen 2 in CPU tests, but the Snapdragon 8 Gen 2 outperforms the A16 in GPU tests. Also, it's not only Samsung phones. Asus, Oppo, OnePlus, and Xiaomi phones with the Snapdragon 8 Gen 2 outperform Apple phones with the A16 in 3DMark tests. The Asus Zenfone 10 beats the iPhone 15 Plus by over 35% in Wild Life (1440p). The iPhone 15 Plus beats the Asus Zenfone 10 by about 16% in the Geekbench 6 multi-core test. Do the match.
- D
- Dom
- raM
- 05 Jun 2024
Anonymous, 05 Jun 2024https://www.gsmarena.com/samsung_galaxy_s24_ultra-review-26... moreThe highest scoring phone from this article with an A16 scored 6618 points in Geekbench, and the highest scoring phone with a Snapdragon 8 Gen 2 only scored 5698 points in Geekbench. 3Dmark only tests 1 part of a phone and isn’t a good representation of overall performance. Plus Samsung phone scores are unreliable as they have been known to cheat in benchmarks before, especially with the s22 series.
- ?
- Anonymous
- 70d
- 05 Jun 2024
Dom, 04 Jun 2024No, even the A16 beats the 8 gen 2
Source: https://nanor... morehttps://www.gsmarena.com/samsung_galaxy_s24_ultra-review-2670p4.php
Look at the 3DMark tests. The Snapdragon 8 Gen 2 outperforms the A16.
- ?
- Anonymous
- pe3
- 05 Jun 2024
Is it similar to how google uses 2 versions of their Tensor chips? One for the 8/8pro and lowerclocked one for the 8a but advertising them as the same thing?
Maybe the 8pro has also different clock speeds?
- ?
- Anonymous
- p$S
- 05 Jun 2024
Android-Never-Again, 05 Jun 2024Someone get this commenter some copium, STAT.Send it. But try not to ruin the canisters while you are handling it, fangurl tears are corrosive AF :)
- W
- Wat
- CbI
- 05 Jun 2024
Even if performance is the same, battery life will be lower.
- ?
- Anonymous
- vaS
- 05 Jun 2024
Anonymous, 04 Jun 2024If performance is truly less than what they claimed in the ... moreStill trying hard to defend a company that secretly throttled iphone 6/6s line up to cover up for the batches that shipped with defective batteries?
Ironically, it was also reddit that managed to exposed that BS.
So Apple said sorry lmao.
- A
- Android-Never-Again
- K6X
- 05 Jun 2024
Anonymous, 04 Jun 2024Pure unadulterated BS from you and Apple, in a weak attempt... moreSomeone get this commenter some copium, STAT.
- ?
- Anonymous
- tT7
- 04 Jun 2024
Probably as "accurate" as the unmarked graphs!
- D
- Dom
- raM
- 04 Jun 2024
Anonymous, 04 Jun 2024a17 got pwned by old snapdragon 8 gen 2 a17 cant rven match... moreNo, even the A16 beats the 8 gen 2
Source: https://nanoreview.net/en/soc-list/rating
- ?
- Anonymous
- NhR
- 04 Jun 2024
Why is Apple introducing a new iPad Air with a 2 years old chip though? iPad air is not as cheap as the iPad 10.9, which hardly justifies this..
That just makes the M1 iPad air a much better purchase, especially since if you care about their upcoming AI features, you won't get any of it anyways unless you have an M3/A17 Pro device.
- ?
- Anonymous
- p$S
- 04 Jun 2024
Anonymous, 04 Jun 2024Benchmarks don't compare core counts, they compare per... moreThe whole issue came to light when people started comparing performance numbers of the same chip, which was supposed to have the same cores, used in the same conditions.
Not all cores (even speaking specifically only about M-series chips) are the same, what are you even talking about? But the same cores, on the same chip, created using the same manufacturing process for the same generation is "generally" supposed to give the same performance, barring the margin of performance deviation values we call the "silicon lottery" that is inherent to semiconductors by nature and (ab)used by OEMs in the form of chip binning.
I am sensing a lot of confusion on this topic from your posts. More cores = more performance is pretty much the guideline (as long as power and cooling needs are met) for the very SAME chip. When things don't rationally add up, that's when you start looking deeper into things and asking questions on why this M2 chip is performing like this on this device and performing like that on another. That comparison effort is exactly how this matter blew up and people found out it was NOT the same chip. And forced Apple to start changing their website's techspecs at a later date and then give this verbal jugglery performance statement to that blog site.
And all of this "performance" talk is you skillfully/confusingly shifting goalposts about how Apple posted "10 cores" on their website, sold the device claiming it still had "10 cores" but only later the customers realized that it's not 10 cores but in fact a lower number of cores and Apple for whatever reason was untruthful in their advertising. The narrative Apple is banking on here about performance values are honestly irrelevant if it comes down to the letter of the law, since it does not change the fact that 10 cores were advertised and only 9 were delivered and the advertisement information was later amended to reflect this fact after customers were already sold the product using false marketing material.
That's the core topic which you have deviated from.
What you "think" or "believe", your feelings are subjective and non-relevant to the issue, but the fact of the lie is objectively real.
- ?
- Anonymous
- 0UU
- 04 Jun 2024
Anonymous, 04 Jun 2024The performance was always the correct number. The number o... morea17 got pwned by old snapdragon 8 gen 2 a17 cant rven match snapdragon 8 gen 3 nor 4
- ?
- Anonymous
- y26
- 04 Jun 2024
Anonymous, 04 Jun 2024"You can't simply compare one device to another b... moreBenchmarks don't compare core counts, they compare performance numbers.
You're acting like every core is created equal and performs equally in every device it's in. This is absolutely not how it works.
If Apple were to make a new product with 12 cores, but advertised performance was less than their 10 core product, that wouldn't be "false advertising". You're making the mistake of assuming "more cores = more performance" and that is not how the technology works.