We ran every test you could think of on the M1 Ultra

We've now tested every version of Apple's M1 processor, from the first M1 chip in the 13-inch Macbook Pro all the way up to the M1 Ultra in the new Mac Studio, and the chip's ability to scale performance is pretty incredible. The M1 Ultra fuses two M1 Max chips together to get you a processor with 20 CPU cores and 64 GPU cores, along with up to 128GB of RAM, and it's one of the fastest processors we've ever tested.

We asked what tests you'd like to see run on the M1 Ultra and assembled quite a list, including Adobe Lightroom and Premiere Pro, Davinci Resolve and Fusion, 3D modeling in Blender, machine learning tests like TensorFlow and Pytorch, and even some gaming.

Amazingly, the M1 Ultra really does seem to be around twice as fast as the M1 Max in most applications. Whatever overhead is required to shuffle data around such a large chip, it rarely impacts CPU performance. The GPU scaling doesn't fare quite so well. The M1 Ultra was typically 40-80% faster than the M1 Max, still speedy, but not quite as impressive in its scaling as the CPU.

The M1 Ultra does best when its hardware accelerators can kick in. These are the parts of the chip built to speed up specific tasks, namely video rendering and AI processing. In a test processing ten 8K video clips at once, the M1 Ultra did the job in just 29 seconds when its accelerators were able to help out. This was about twice as fast as the PC we were testing, despite it having a 16-core AMD 5950X processor and Nvidia RTX 3080 Ti graphics card.

Apple's M1 ecosystem does at times feel unfinished. There were more strange bugs than I'd expected, and software compatibility can still be a problem, but it's improving rapidly. The M1 Ultra wasn't always faster than my PC, but in some tests it was two or three times quicker, and in only a few cases was it significantly slower. If you use Apple and need a high-end desktop, there's really no decision, and for everyone else looking for a workstation, the M1 Ultra Mac Studio is a contender.

Check out the video above for the full details of our testing.

Apple’s M1 Pro and Max chips are living up to the hype

It's now been three months since Apple brought its own silicon to its pro-grade laptops. As we said in our review, the M1 Pro and Max are fantastically capable chips that offer more than enough performance for the vast majority of users.

But for our explainer show Upscaled, we wanted to go a little deeper, with a suite of benchmarks that really push the Pro and Max to their limits, as well as some requests that our readers really wanted to see. Ever wondered how long it takes to open a 100MB Excel file?

Watch the video above for the full story, and we'll see you for the next episode soon.

Apple’s M1 Pro and Max chips are living up to the hype

It's now been three months since Apple brought its own silicon to its pro-grade laptops. As we said in our review, the M1 Pro and Max are fantastically capable chips that offer more than enough performance for the vast majority of users.

But for our explainer show Upscaled, we wanted to go a little deeper, with a suite of benchmarks that really push the Pro and Max to their limits, as well as some requests that our readers really wanted to see. Ever wondered how long it takes to open a 100MB Excel file?

Watch the video above for the full story, and we'll see you for the next episode soon.

Is Apple’s M1 Max really the fastest laptop chip ever?

After months of rumors, Apple has revealed not one but two new processors, along with a pair of new Macbook Pro laptops. The original M1 processor paired four high-performance and four low-power cores with a 7- or 8-core GPU, and with that set up it delivered impressive results. The M1-powered 13-inch MacBook Pro bested a 16-inch MacBook Pro with an 8-core Intel i9 CPU and dedicated GPU when we tested it last year, and did so while staying impressively cool and quiet.  

The new M1 Pro and M1 Max bump the core count to eight high-performance and two low-power cores, and add 16, 24, or 32 GPU cores. With twice the high-performance CPUs, and up to four times the GPU cores as the original M1, these chips should be incredibly fast, but expect it to come at the cost of more heat and noise.

Despite a shortage of compatible games, the original M1 proved to be a remarkably adept processor for gaming, and these new chips should be even better, outpacing most Intel CPUs and maybe even besting a mobile NVIDIA RTX 3080 GPU's performance. The new 14-inch and 16-inch MacBooks also come with a 120Hz display with HDR, and improved speakers. Add that all together, and suddenly the new MacBook Pro looks like a pretty capable gaming machine, compatibility issues aside. If only Apple still supported Bootcamp.

For all the details on these new chips, check out the video above.

Is Apple’s M1 Max really the fastest laptop chip ever?

After months of rumors, Apple has revealed not one but two new processors, along with a pair of new Macbook Pro laptops. The original M1 processor paired four high-performance and four low-power cores with a 7- or 8-core GPU, and with that set up it delivered impressive results. The M1-powered 13-inch MacBook Pro bested a 16-inch MacBook Pro with an 8-core Intel i9 CPU and dedicated GPU when we tested it last year, and did so while staying impressively cool and quiet.  

The new M1 Pro and M1 Max bump the core count to eight high-performance and two low-power cores, and add 16, 24, or 32 GPU cores. With twice the high-performance CPUs, and up to four times the GPU cores as the original M1, these chips should be incredibly fast, but expect it to come at the cost of more heat and noise.

Despite a shortage of compatible games, the original M1 proved to be a remarkably adept processor for gaming, and these new chips should be even better, outpacing most Intel CPUs and maybe even besting a mobile NVIDIA RTX 3080 GPU's performance. The new 14-inch and 16-inch MacBooks also come with a 120Hz display with HDR, and improved speakers. Add that all together, and suddenly the new MacBook Pro looks like a pretty capable gaming machine, compatibility issues aside. If only Apple still supported Bootcamp.

For all the details on these new chips, check out the video above.

How are camera sensors still improving?

Cameras keep getting better, but it sometimes feels like the pace of improvement has slowed. You can look at photos from the original Canon 5D, released in 2005, and they look pretty good! In some tests of performance, sensors seem to have barely changed in the past decade. However, there have been big advancements in noise reduction, dynamic range and how fast sensors can be read, which have not only improved image quality, but enabled photographers to work in more challenging environments and shoot in new ways. While the original 5D could capture three 13-megapixel images per second at a native ISO 1600, Sony's latest a1 can shoot 30 50-megapixel images at a native ISO 32,000 in the same timeframe.

In the latest episode of our Upscaled series, we explain how sensors work, and how camera makers have overcome the physical challenges of trying capture a lot of light in a very short time. For the full story, check the video below.

Intel has a plan to go beyond 3nm chips

Earlier this year, Intel announced they were planning to retake the CPU manufacturing lead and "unquestioned leadership" in the PC world. These were impressive goals, but what was missing was any sense of how they'd actually achieve them. Now, we finally know Intel's plan.

Intel's CEO Pat Gelsinger and SVP of Technology Development Dr. Ann Kelleher, laid out the company's plan for the future. For starters, Intel is renaming its manufacturing nodes. What used to be 10nm "Enhanced Superfin" is now just "7." This may feel a little duplicitous — "just wave a wand a you've got better technology!" — but to be fair to intel, the nanometer measurements of process nodes don't really correspond to anything physical any more, and in terms of density Intel's current 10nm chips are competitive with TSMC and Samsung's 7nm.

Looking beyond 7nm, Intel is targeting an aggressive release schedule with major product updates happening annually. We're expecting their Alder Lake chips this fall, which will mix high and low-powered cores, followed by now-4nm Meteor Lake chips that will move to a "tile" (chiplet) design, and incorporate Intel's 3D stacked-chip technology, Foveros. 

Beyond that, Intel has technology mapped out for an EUV-based 3nm node that will use the high-energy manufacturing process to streamline chip creation, and a "20A" for angstrom node. This is one ten-billionth of a meter (meaning it's 2nm), and will be followed by a 18A node that Intel hopes to start moving into production in 2025 for products sometime in the 2nd half of the decade. Again, while node measurements don't really correspond to physical structures any more, a silicon atom is in the area of 2 angstroms wide, so these are seriously tiny transistors.

This release schedule seems aggressive, and Intel does not have the best track record of meeting targets for new nodes, but if it can even come close to these goals, expect your laptops and desktops to get a huge performance boost in the next few years. 

For even more info on Intel's plans, and details on its EMIB interconnect technology and two new versions of Foveros, be sure to check out the video above. You can see our sources here.

Intel has a plan to go beyond 3nm chips

Earlier this year, Intel announced they were planning to retake the CPU manufacturing lead and "unquestioned leadership" in the PC world. These were impressive goals, but what was missing was any sense of how they'd actually achieve them. Now, we finally know Intel's plan.

Intel's CEO Pat Gelsinger and SVP of Technology Development Dr. Ann Kelleher, laid out the company's plan for the future. For starters, Intel is renaming its manufacturing nodes. What used to be 10nm "Enhanced Superfin" is now just "7." This may feel a little duplicitous — "just wave a wand a you've got better technology!" — but to be fair to intel, the nanometer measurements of process nodes don't really correspond to anything physical any more, and in terms of density Intel's current 10nm chips are competitive with TSMC and Samsung's 7nm.

Looking beyond 7nm, Intel is targeting an aggressive release schedule with major product updates happening annually. We're expecting their Alder Lake chips this fall, which will mix high and low-powered cores, followed by now-4nm Meteor Lake chips that will move to a "tile" (chiplet) design, and incorporate Intel's 3D stacked-chip technology, Foveros. 

Beyond that, Intel has technology mapped out for an EUV-based 3nm node that will use the high-energy manufacturing process to streamline chip creation, and a "20A" for angstrom node. This is one ten-billionth of a meter (meaning it's 2nm), and will be followed by a 18A node that Intel hopes to start moving into production in 2025 for products sometime in the 2nd half of the decade. Again, while node measurements don't really correspond to physical structures any more, a silicon atom is in the area of 2 angstroms wide, so these are seriously tiny transistors.

This release schedule seems aggressive, and Intel does not have the best track record of meeting targets for new nodes, but if it can even come close to these goals, expect your laptops and desktops to get a huge performance boost in the next few years. 

For even more info on Intel's plans, and details on its EMIB interconnect technology and two new versions of Foveros, be sure to check out the video above. You can see our sources here.