On average, Apple’s ultra-slim flagship appears to be faster, but that may quickly change.
Apple has just announced the iPhone Air, equipping it with a version of the A19 Pro, which is its best-performing chipset. That puts the ultra slim device right against the Galaxy S25 Edge, and now we have the first opportunity to see which one is faster.
Unfortunately, the answer may turn out to be a little more complicated than fans of either brand would like. On the bright side, you can say either is faster, as long as you’re okay with a bit of cherry picking.
We don’t have a clear winner between the iPhone Air and the Galaxy S25 Edge
The first benchmarks for the iPhone Air have started showing up on Geekbench, revealing that Apple has kept its promise to provide top-tier performance for the ultra-slim device. The single-core score of the iPhone Air ranges from 3,206 to over 3,700 points, and the multi-core score is between 8,546 and 9,737 points in different runs.
Comparing that to the results of the Galaxy S25 Edge makes for an interesting journey. Samsung’s ultra-slim phone has been available for months, so there is more data on Geekbench’s database. However, the single-core scores from the last few days range from 1,579 to 3,205, and the multi-core scores go as low as 6,564 and as high as 10,102.
Keep in mind that many people own a Galaxy S25 Edge and are running tests in all sorts of conditions. Once the iPhone Air reaches its first owners, it is likely to see a similarly wide range of scores. Even with that in mind, the S25 Edge outperforms the iPhone Air in its best multi-core tests and lags in the single-core tests. That doesn’t tell the full story, though.
Which phone do you want to appear faster?
If we compare the average scores of all available iPhone Air tests and the averages of all Galaxy S25 Edge since September 9, Apple is clearly doing better. In fact, the difference between the average scores looks jarring.Before Samsung fans get mad, I need to clarify that with a bit of cherry picking, we can make either of the devices look superior. And that’s the main issue with such comparisons – they’re rarely representative of how the devices perform in real life.Looking at the scores, the only thing I can safely say is that both phones are fast. I don’t think any of them would stumble with any task. What’s left to see is how the iPhone Air holds up under sustained load, which is probably more important than the peak performance.
Rediscover some of the most unique and memorable phones of the last two decades! «Iconic Phones» is a beautifully illustrated book that we’ve been working on for over a year — and it’s coming out in just a couple short month!
«Iconic Phones: Revolution at Your Fingertips» is a must-have coffee table book for every phone lover out there. Covering the stories of more than 20 fan-favorite phones, it takes you on a memorable journey through the technological revolution that shaped our lives. Sign up now to secure an early discount price!

ChatGPT maker wants to be able to detect early signs of AI bad behavior
OpenAI’s hardware chief warns future AI models need real-time hardware kill switches
Richard Ho highlights networking, memory, and power challenges in scaling infrastructure
He calls for benchmarks, observability, and cross-industry partnerships to address reliability and trust
A senior OpenAI executive has warned that future AI infrastructure will require hardware-level safety features, including kill switches.
Richard Ho, head of hardware at the company, made the remarks during his keynote at the AI Infra Summit in Santa Clara.
“It has to be built into the hardware,” Ho said. “Today a lot of safety work is in the software. It assumes that your hardware is secure. It assumes that your hardware will do the right thing. It assumes that you can pull the plug on the hardware. I am not saying that we can’t pull the plug on that hardware, but I am telling you that these things are devious, the models are really devious, and so as a hardware guy, I want to make sure of that.”Silicon-level safety measures
Ho argued that the growth of generative AI is forcing a rethink of system architecture and described how future agents will be long-lived, interacting in the background even when a user is not actively engaged.
This shift requires memory-rich, low-latency infrastructure to manage continuous sessions and communication across multiple agents.
Networking, Ho said, is becoming a bottleneck. “We’re going to have to have real-time tools in these – meaning that these agents communicate with each other. Some of them might be looking at a tool, some might be doing a website search. Others are thinking, and others need to talk to each other.”
Ho outlined several hardware challenges that must be addressed, including limits on high-bandwidth memory, the need for 2.5D and 3D chip integration, advances in optics, and extreme power requirements that could reach 1 megawatt per rack.
The safety measures OpenAI put forward include real-time kill switches built into AI clusters, telemetry to detect signs of abnormal behavior, and secure execution paths in CPUs and accelerators.
Ho wrapped things up by saying, “We don’t have good benchmarks for agent-aware architectures and hardware, and I think it is important to know about latency walls and latency tails, what is the efficiency and power and things like that. We need to have good observability as a hardware feature, not just as a debug tool, but built in and constantly monitoring our hardware.»
«Networking is a real important thing, and as we head towards optical, it is unclear that the reliability of the network is there today. We need to get there with enough testing of these optical testbeds and these other communication testbeds that show that we actually have the reliability.”