M1 Ultra: Apple’s big step in chip design

Apple introduced what it calls “one final chip” to the M1, M1 Ultra family at Peek Performance Event. I’ve looked a little deeper into what this means.

Delivering the impossible – it’s a daily job

The chip was introduced by Johnny Srouji, Apple’s Senior Vice President of Hardware Technology. He describes the physical limitations of making chips larger than the M1 Max, explaining that the industry often chooses to use dual chips, but described the inefficiencies to that approach.

That’s why Apple chose to combine two M1 Mac chips in the M1 Ultra. “The M1 Ultra is another game changer for Apple silicon that will once again shake up the PC industry,” said Srouji. “

He also said the M1 Ultra “completes the line of Mac processors – at least for now.

Srouji went on to explain how the company has managed to combine two M1 Max chips into one M1 Ultra. He pointed to a deadly connectivity technology that has always been on the M1 Max, something Apple has never discussed before. (I seem to recall this being discovered by others at some point, but can’t find a reference.)

What is UltraFusion?

That connection uses a proprietary encapsulation architecture that Apple calls UltraFusion. Apple describes this as the industry’s “way ahead” and claims it can move data around the chip at 2.5 terabytes per second with fast, low-bandwidth, inter-processor bandwidth.

To do so, UltraFusion uses a silicon buffer with twice the connection density “than any other technology available,” says Srouji.

A silicon binder is essentially an in-package interconnect that bridges the two dies used in the M1 Ultra. (I imagine Apple could have used TSMC’s 3D Fabric technology to achieve UltraFusion.)

Apple has won what seemed to be its own First related design patent in March 2021, although work began earlier. Its approach is not unique; AMD and others have also developed silicon interposer technology for their PC chips. But Apple’s claim of achieving an out-of-the-box solution will resonate loudly.

Apple says its technology provides Quadruple the bandwidth of competing interposer technologies.

Another key element to Apple’s invention is that developers won’t need to rewrite their code to use this power; UltraFusion means that at the system level, the Mac treats the chip as a single processor, not two.

This means that in practice, applications can access twice the transistors and double the performance, efficiency, and GPU cores – without tweaking the code.

Meet the US family

Apple’s processor family currently offers four variants equipped with the following.

M1

The critically acclaimed M1 processor delivers:

  • 16 billion transistors and 119mm squared die size.
  • 4 Performance Cores, 12MB L2 Cache.
  • 4th efficient cores 4MB L2 cache.
  • 8 GPU cores.
  • 16GB DDR4x memory clocked at 68GB/s.

M1 Pro

The M1 Pro takes this higher, with:

  • 33.7 billion transistors on a 240mm square die.
  • 8 performance cores, 24MB L2 Cache.
  • 2 efficient cores with 4MB L2 cache.
  • 16 GPU cores.
  • 32GB DDR5 memory at 200GB/s.

M1 Max

M1 Max offers:

  • 57 billion transistors on a 420mm square die.
  • 8 performance cores, 24MB L2 Cache.
  • 2 efficient cores with 4MB L2 cache.
  • 32 GPU cores.
  • 64GB DDR5 memory at 400GB/s.

And the new M1 Ultra

M1 Ultra gives you:

  • 114 billion transistors on an 840mm square die.
  • 16 Performance Cores, 48MB L2 Cache.
  • 4 efficient cores with 4MB L2 cache.
  • 64 GPU cores.
  • Up to 128GB DDR5 memory at 800GB/s.

What does it all mean?

Since the M1 Ultra has twice as many cores, it can manage even the most complex computing tasks. Indeed, since the M1 Pro already offers what most Mac users need, this chip is sure to open the door to new, in-depth experiences. After all, the performance and power you can harness in 3D design, video, and audio applications will also benefit gaming.

The fact that the M1 Ultra also offers 64 graphics cores is remarkable.

Together, the combination of these technologies means that Apple has been able to create a chip that is 1.9 times more powerful than the latest 12 chip.order generation Intel Core i9-12900K CPU when both run at 60 watts. In other words, you get more raw CPU performance at a fraction of the energy cost.

That’s fine for individual users, but for companies that are running banks of high-end machinery, reducing energy consumption has major benefits to the cost of doing business.

When running an application, you can also settle for a large 128GB unified memory. With professional video production studios running banks of Macs performing compute-intensive operations, this results in significant budget savings. The fact that they also work so much faster means that the investment in these machines becomes less than “one rounding error” for some pro shops, like CCS . analyst Ben Wood noted.

Meanwhile, back to the real world

We still don’t know what the real-world impact of this performance will be, but first thing Geekbench Score only The Mac Studio running the M1 Ultra achieved a single-core score of 1,793 and a multi-core score of 24,055.

In contrast, the top-end Xeon W Mac Pro scored 1,152 and 19,951 in the same tests, while AMD’s much-advertised Threadripper 3990X was only marginally faster with 25,133 multi-core points and just 1,213 single points. core.

Of course, composite scores don’t make much sense when used in the real world.

While that’s true, I can say that what I’ve seen about M series chips so far strongly recommend these tests will translate to real-world application speed. With the switch to M1 quite a lot double the performance of Adobe apps on MacI’ll be interested to see what Adobe has to say about the M1 Ultra in the future.

Apple itself claims the processor will deliver 3.8x CPU performance, 4.5x GPU performance, and 3x faster machine learning than Mac systems, and the Intel iMac M1 also stands out. .

[Also read: One year on, developers still love Apple Silicon Macs]

What happens next?

Concluding its presentation, Apple said it had one more Mac to showcase – but said that would be for another time.

Many assume this will be the Mac Pro, but we can’t be sure if the company intends to use the M1 Ultra in that computer or has other plans. I think it’s likely that Apple will further optimize the M1 Ultra design and speed it up to exploit the additional thermal management available in the Mac Pro.

Apple’s top silicon engineer alluded to design improvements in his presentation, and there’s industry-wide recognition that we’re reaching a finite limit in terms of processor technology. That means future innovations will increasingly revolve around design, packaging, and architecture – and could be another reason why Apple developed UltraFusion: this constitutes a road map for innovation. New more chip design.

Unified memory and further expansion of what is placed on the SoC (such as adding Apple’s own 5G modem design) will open the door to further development, as well as the inclusion of a GPU on the same processor.

Apple should also be ready to introduce M-series processors built on the 3nm design process after starting the transition to 3nm manufacturing in 2023-24. This will certainly allow it to introduce more power, performance and efficiency improvements, even as the rest of the industry gathers information about the 5nm chip.

What do businesses think?

Pointing out that the M1 Ultra is 7 times faster than the M1, Wedbush analyst Daniel Ives, tell customers This represents “an amazing technological achievement that is another shot on the chessboard as chip giants like Intel or Apple beat chip companies at their own game on the desktop front.” “.

“Apple has set the bar quite high with the M1 Max, and now they are aiming to surpass it with the M1 Ultra. And if they can achieve those goals, then they will have twice set a new high for SoC design in the space of just 6 months. These are exciting times, really,” Anandtech.

However, one big question remains unanswered: why is the Mac Studio with the M1 Ultra 2 pounds heavier than the one with the M1 Max? Obviously, the answer is that they both have the same 370-watt power supply built-in. The extra weight reflects that the M1 Ultra model has a larger copper thermal module than the M1 Max Mac Studio, which uses an aluminum heatsink.

Follow me on Twitteror join me in AppleHolic’s Bar & Grill and Discuss Apple group on MeWe.

Copyright © 2022 IDG Communications, Inc.

https://www.computerworld.com/article/3652700/the-m1-ultra-apples-big-leap-forward-in-chip-design.html#tk.rss_all M1 Ultra: Apple’s big step in chip design

Fry Electronics Team

Fry Electronics.com is an automatic aggregator of the all world’s media. In each content, the hyperlink to the primary source is specified. All trademarks belong to their rightful owners, all materials to their authors. If you are the owner of the content and do not want us to publish your materials, please contact us by email – admin@fry-electronics.com. The content will be deleted within 24 hours.

Related Articles

Back to top button