A18 vs. A18 Pro: What's the Difference?

[chucker23n1]

Yes. The Snapdragon gen 4 single core geekbench score is 5% lower than the A18 pro, but multicore is 20% higher.

I'm seeing conflicting numbers, so it may be too early to tell.

Here's an iPhone 16 with 3114 / 6666 scores.

Here's an iPhone 16 Pro with 3409 / 8492 scores.

The first doesn't give an estimated CPU clock; the second says 4.04 GHz. That's plausible. It would give us a bump of 17.7% / 18% over the iPhone 15 Pro, or if we normalize that to the same clock, a bump of 10.2% / 10.5%.

Especially the single-core IPC change is close to what I see comparing M3 and M4, which is 10.5% / 4.0%. (The low multi-core improvement is explained by having only three p-cores, not four.)

The relatively low 16 number suggests that the A18 non-Pro is more different than I would've expected. However, the multi-core score is strange. The iPhone 15 on the A16 already scored 6325. The core count is the same (2p4e on A16, A18, and A18 Pro), according to Apple, so why would the multi-core improvement be so low?

But similarly, the Qualcomm data feels incomplete.

Here's a score of 2884 / 8840.

But here's a score of 3236 / 10049.

 

[ToomeyND]

I think they are wise to show performance differences against a range of iPhones, I remember seeing in the presentations bar graphs showing performance increase against iPhone 12, 13, 14, and 15. The thing is, people are holding onto phones for longer, and it’s convenient for viewers to be able to compare their phone.

whether performance increases are that relevant is another question. It’s been a long time since I’ve thought of my iPhone as ’too slow’ at anything.

I don't mind showing more info, as you noted. But to highlight those things is a bit over the top. My actual issue is when they selectively compare once aspect to the A17, but another aspect to the A16, a camera quality over the iPhone 15, but a screen characteristic to the iPhone 14, etc. It's like they are recycling slides from previous years because they didn't improve them over last year but still want to fill time.

 

[eldho]

bragging about speed increases compared to the iPhone 12 are insane at this point (and I have an iPhone 12).

Maybe you are a target audience for the advertising that Apple is doing with this information to lure you to upgrading. This makes it perfectly rational for Apple to point out how much progress there has been over a longer period of time rather than just since the last iteration.

 

[AndiG]

I would change the title to:
"

What's the Difference? And does anyone care?"​

 

[chucker23n1]

I would change the title to:
"

What's the Difference? And does anyone care?"​

I do.

 

[CarAnalogy]

A17 Pro is faster than A18.

Instead of buying iPhone 16, you should go for iPhone 15 Pro and get 120Hz display and 1 more GPU core.

Except that the A17 Pro was on a bad process node and runs way hotter than it should. Reading between the lines, the A18 Pro is far more efficient and should run far cooler, and the 15 Pro has had heat issues since day one.

Apple very rarely openly admits a design flaw, but it seems clear to me the A18 Pro is what they were trying to make, and the A17 Pro is kind of a dud. Reminds me of the Pentium 4. Not necessarily badly performing for its time, but running waaaaay too hot compared to the Core that came out the next year.

 

[HylianKnight]

15%, which is claimed by Apple, not proven yet (at this point we don't know what exactly anything they claim means until tested, like their fake 2x telephoto optical zoom on base model that looked liked crap or claiming to have 10x optical zoom on Pro's spec sheet), is basically nothing.
they also claimed A17 Pro is 3x more efficient than the competition (Nokia 1100?) but 15 Pro turned out to be a molten core with less battery life than base model 15.
Snapdragon on the other hand saw 35-40% CPU performance increase without having access to 3nm silicon last time, so I'm really excited to see if CPU performance is going be another L after display, battery and camera.

Which processor? Snapdragon 7 Gen 3 to 7+ saw about a 15% improvement in CPU, and about 45% in GPU. Also, the spec sheet says 5x telephoto lens. Please, check your sources and share?

 

[kiranmk2]

I wonder if either chip will be able to maintain a solid 30 fps and smooth frame time in the AAA games launched over the last year. All the reviews I've seen show Resident Evil Village and 4 REmake struggling on A7Pro iPhone and M1 devices. typically needing an M2 to get smooth 30 fps.

 

[gphome]

Realistically, there will be very little difference between the A18 and A18 Pro, and in 99.9% of use for average users zero perceivable difference.

For standard iPhone users, Apple Intelligence has been a godsend, its forced Apple to put their latest chip in the standard iPhone and bump the RAM upto 8GB, its an excellent deal now.

I think that's why they've kept the standard iphones on ancient 60ghz screens, worried they'll eat into the Pro profit margins too much

 

[Ramchi]

Storage QLC or TLC? That is not mentioned in some of the articles.

 

[HylianKnight]

Maybe you are a target audience for the advertising that Apple is doing with this information to lure you to upgrading. This makes it perfectly rational for Apple to point out how much progress there has been over a longer period of time rather than just since the last iteration.

Apple likely shows these figures because they have a good idea how many people are on older model phone. So, while tech nerds like me who tend to get the latest and greatest may be a bit annoyed with Apple comparing their latest flagship to a model other than the previous year, I can understand why they do it.

 

[Ramchi]

Realistically, there will be very little difference between the A18 and A18 Pro, and in 99.9% of use for average users zero perceivable difference.

For standard iPhone users, Apple Intelligence has been a godsend, its forced Apple to put their latest chip in the standard iPhone and bump the RAM upto 8GB, its an excellent deal now.

I think that's why they've kept the standard iphones on ancient 60ghz screens, worried they'll eat into the Pro profit margins too much

Yes, by pushing the customers envelope by making them to buy higher editions and of course more profits. If you keep comparing each segment you can see there will be compelling difference that customers will opt for. If you consider iPad Pro the RAM is linked to the storage option and hence the price will be substantially higher than the base versions.

 

[crsh1976]

Why does this article seem way more confusing than it needs to be?

I'd wager the article is really just trying to make sense of Apple's marketing speak - the A18 and A18 Pro are nearly identical, some minute (and overstated, imo) bits are there to make one of them "pro" somehow.

 

[9081094]

Why does this article seem way more confusing than it needs to be?

Can’t you rewrite it with AI?

 

[HylianKnight]

I wonder if either chip will be able to maintain a solid 30 fps and smooth frame time in the AAA games launched over the last year. All the reviews I've seen show Resident Evil Village and 4 REmake struggling on A7Pro iPhone and M1 devices. typically needing an M2 to get smooth 30 fps.

I’d be surprised. Thermal limitations imposed on the form factor of these devices will prevent significant sustained speeds.

 

[9081094]

I’m pretty sure it is 8GB.

Yep… if it was 12GB they would scream it out loud and probably would’ve taken 10 minutes keynote time 😎

 

[HylianKnight]

I don't mind showing more info, as you noted. But to highlight those things is a bit over the top. My actual issue is when they selectively compare once aspect to the A17, but another aspect to the A16, a camera quality over the iPhone 15, but a screen characteristic to the iPhone 14, etc. It's like they are recycling slides from previous years because they didn't improve them over last year but still want to fill time.

Apple wants people on older devices to upgrade, and they have a much better idea how many of those devices are active out there. So, from a business standpoint, it actually makes a lot of sense to focus on those numbers. And, for those like me who always want to know how performance compares to last year, you typically don’t have to wait long to find out. Not sure it’s something worth getting upset over.

 

[HylianKnight]

Yep… if it was 12GB they would scream it out loud and probably would’ve taken 10 minutes keynote time 😎

Yeah. Rumor mill says they are moving to 12GB for the 17 pros.

 

[9081094]

Yes. The Snapdragon gen 4 single core geekbench score is 5% lower than the A18 pro, but multicore is 20% higher.

😱 and away goes the last area where Apple had a lead over the competition. I think the same will happen with their M-series next year. Snapdragon X-elites first generation was already on par and much faster on many fronts with the M3. Once the M4 comes out Apple probably brags about the fastest laptop chip in the industry. Then Qualcomm comes out with their second generation X-elite to leapfrog Apple on all fronts.

I’ve said it before… wait and watch 😎

 

[ACHD]

I'm seeing conflicting numbers, so it may be too early to tell.

Here's an iPhone 16 with 3114 / 6666 scores.

Here's an iPhone 16 Pro with 3409 / 8492 scores.

The first doesn't give an estimated CPU clock; the second says 4.04 GHz. That's plausible. It would give us a bump of 17.7% / 18% over the iPhone 15 Pro, or if we normalize that to the same clock, a bump of 10.2% / 10.5%.

Especially the single-core IPC change is close to what I see comparing M3 and M4, which is 10.5% / 4.0%. (The low multi-core improvement is explained by having only three p-cores, not four.)

The relatively low 16 number suggests that the A18 non-Pro is more different than I would've expected. However, the multi-core score is strange. The iPhone 15 on the A16 already scored 6325. The core count is the same (2p4e on A16, A18, and A18 Pro), according to Apple, so why would the multi-core improvement be so low?

But similarly, the Qualcomm data feels incomplete.

Here's a score of 2884 / 8840.

But here's a score of 3236 / 10049.

More than likely the efficiency clock speed is lower. You can see the single core speeds drop so we know the core clocks are lower on single so why not all around..

The lower Multi Core can be explained by having a lesser ability to dissipate heat compare to the 16 pros and lower clock speeds all around.


The biggest thing apple focus'ed on wasn't performance. It was the the fact it could perform the same as previous gen but with 20-30% more efficiency therefore doing same performance for less.

Its obvious this is a more finalized 17 since they tightened up performance a bit and its more efficient.

 

[HylianKnight]

😱 and away goes the last area where Apple had a lead over the competition. I think the same will happen with their M-series next year. Snapdragon X-elites first generation was already on par and much faster on many fronts with the M3. Once the M4 comes out Apple probably brags about the fastest laptop chip in the industry. Then Qualcomm comes out with their second generation X-elite to leapfrog Apple on all fronts.

I’ve said it before… wait and watch 😎

And this kind of competition can only ever benefit us consumers. Why people get huffy about it (not you), I’ll never understand.

 

[gphome]

And this kind of competition can only ever benefit us consumers. Why people get huffy about it (not you), I’ll never understand.

Agreed, Microsoft getting its act together on ARM processors in partnership with Qualcomm is only good news for Apple users, will keep them on their toes and then some! Intel too!

 

[bradman83]

I find it interesting that Apple would say that the A18 compares well to current desktop chips. With the latest boosts in speed it would comfortably outperform older desktop chips like the M1, which is already a really good level for a mobile chip to be at.

But that means that if they followed the same process as previous years, and developed new performance and efficiency cores first for iPhone, we can expect these large performance gains to come to M-series chips in the M5, which would help keep the M-series ahead of the pack.

It's much more likely that the A18 shares the same core microarchitecture as the recently released M4 given that both are on the 3NE process node. Case and point: Apple's performance comparison chart says the A18 is 50% faster than the A13 chip; the M4 is advertised as being 50% faster than the M2 chip, which is based on the A13's "Avalanche" and "Blizzard" microarchitecture.

(Side note: Apple hasn't made their core codenames discoverable since the A16 and M2, and I kind of miss that! I can only assume they ran out of meteorological phenomena to name their cores after since they switched to mountain names for the A16.)

 

[Mac Fly (film)]

Can’t you rewrite it with AI?

I can rewrite it without AI. Pro iPhones and slightly bigger, slightly faster, with some more battery life and a titanium chassis. ☑️

 

[bkendig]

Why does this article seem way more confusing than it needs to be?

Because most of the article compares the A18 against the A16 Bionic, and the A18 Pro against the A17 Pro, instead of giving any specific details on how the performance of the A18 and the A18 Pro compare.

For an article titled "A18 vs. A18 Pro: What's the Difference?", that's kind of disappointing.