Apple's M1 processor often surpasses the graphics performance of desktop GPUs, including the Nvidia GeForce GTX 1050 Ti and AMD Radeon RX 560, according to a new benchmark submission spotted by Tom's Hardware.
According to Apple, the M1's octa-core GPU can simultaneously handle close to 25,000 threads and deliver up to 2.6 TFLOPS of throughput. This is the same TFLOPS achieved by the Radeon RX 560, and just below the GeForce GTX 1650's 2.9 TFLOPS.
The GFXBench 5.0 benchmarks were tested under Apple's Metal API, and show that the M1 often performs better than the Nvidia GeForce GTX 1050 Ti and AMD Radeon RX 560 by a fair margin. There is not yet a Metal benchmark for the GeForce GTX 1650 for the sake of comparison.
In the Aztec Ruins Normal Tier test, the Radeon RX 560 achieves 146.2 FPS, the GeForce GTX 1050 Ti achieves 159 FPS, and the M1 achieves 203.6 FPS. Similar results are seen across the board, with the M1 almost consistently surpassing the performance of the two desktop GPUs.
It is worth noting that GFXBench 5.0 benchmarks are primarily used to test mobile devices, and the GeForce GTX 1050 Ti and Radeon RX 560 are older GPUs. Nonetheless, the M1 contains integrated graphics and the fact that it can surpass 75W desktop graphics cards is impressive. More detailed GPU comparisons will likely be coming soon as the first Macs equipped with the M1 arrive to customers this week.
That means you dont know the diff between an gpu that is build on the chip vs that nvdidia dedicated gpu.... Tell us what integrated gpu build into the chip can come close to M1
I'm reminded of this classic quote, when Apple was building the iPhone “We’ve learned and struggled for a few years here figuring out how to make a decent phone. PC guys are not going to just figure this out. They’re not going to just walk in.” -- Palm CEO Ed Colligan
I get the feeling Nvidia aren't expecting Apple to 'just walk in' and make competitive graphics processors, and I get the feeling they might be surprised.
The M1 is beating the snot out of the Intel CPUs and now some GPUs. Now don't get me wrong. I recognize that these values may not be sustained and only win in the short term. I don't care because, before now, the Intel equivalents (and I mean equivalent in terms of low power chips) were simply pitiful and didn't even pretend to compete.
This is just the beginning. The M1 is the low end chip. Remember that!
Alright, so the graphics is now plenty fast on an ordinary MacBook Air. The only thing left to see is sustained performance of M1 with passive cooling and whether there's thermal throttling involved and when it kicks in specifically and I might be able to switch from a Pro to an Air in due time.
For integrated GPU's on low power high efficiency platforms (MacBook etc.), this is great to see. Nice. That said a nVidia 1050 is 2 generations old and then a lowish mid range GPU - just for reference. For integrated GPU's that's still great.
The real test will be the GPU's on the higher end iMac Apple Silicon when it comes out. How far up the GPU performance chain will they go is the question (or will they offer a discreet GPU for that platform)?
Wednesday April 24, 2024 2:05 pm PDT by Joe Rossignol
Apple is set to unveil iOS 18 during its WWDC keynote on June 10, so the software update is a little over six weeks away from being announced. Below, we recap rumored features and changes planned for the iPhone with iOS 18. iOS 18 will reportedly be the "biggest" update in the iPhone's history, with new ChatGPT-inspired generative AI features, a more customizable Home Screen, and much more....
Wednesday April 24, 2024 3:39 pm PDT by Juli Clover
Apple today released several open source large language models (LLMs) that are designed to run on-device rather than through cloud servers. Called OpenELM (Open-source Efficient Language Models), the LLMs are available on the Hugging Face Hub, a community for sharing AI code. As outlined in a white paper [PDF], there are eight total OpenELM models, four of which were pre-trained using the...
Apple has announced it will be holding a special event on Tuesday, May 7 at 7 a.m. Pacific Time (10 a.m. Eastern Time), with a live stream to be available on Apple.com and on YouTube as usual. The event invitation has a tagline of "Let Loose" and shows an artistic render of an Apple Pencil, suggesting that iPads will be a focus of the event. Subscribe to the MacRumors YouTube channel for more ...
Saturday April 27, 2024 12:41 am PDT by Tim Hardwick
There are widespread reports of Apple users being locked out of their Apple ID overnight for no apparent reason, requiring a password reset before they can log in again. Users say the sudden inexplicable Apple ID sign-out is occurring across multiple devices. When they attempt to sign in again they are locked out of their account and asked to reset their password in order to regain access. ...
Best Buy is discounting a collection of M3 MacBook Pro computers today, this time focusing on the 14-inch version of the laptop. Every deal in this sale requires you to have a My Best Buy Plus or Total membership, although non-members can still get solid second-best prices on these MacBook Pro models. Note: MacRumors is an affiliate partner with Best Buy. When you click a link and make a...
Apple used to regularly increase the base memory of its Macs up until 2011, the same year Tim Cook was appointed CEO, charts posted on Mastodon by David Schaub show. Earlier this year, Schaub generated two charts: One showing the base memory capacities of Apple's all-in-one Macs from 1984 onwards, and a second depicting Apple's consumer laptop base RAM from 1999 onwards. Both charts were...
Top Rated Comments
Tell us what integrated gpu build into the chip can come close to M1
I get the feeling Nvidia aren't expecting Apple to 'just walk in' and make competitive graphics processors, and I get the feeling they might be surprised.
This is just the beginning. The M1 is the low end chip. Remember that!
Make me proud, Apple.
The real test will be the GPU's on the higher end iMac Apple Silicon when it comes out. How far up the GPU performance chain will they go is the question (or will they offer a discreet GPU for that platform)?