The Wall Street Journal broke the initial report on Project ACDC. This project has been in the works for years, which does not surprise me. Ever since Apple’s M-series chips emerged, both performance leaders and giant performance-per-watt leaders, it seemed somewhat obvious that Apple’s own data centers would begin to change at some point in the future.

AI for Inferencing

The ACDC AI chips are reportedly aimed at inferencing, not training AI models. This suggests that Apple is initially planning to partner with AI LLMs from the likes of OpenAI, Anthropic, or others. A month ago, Apple hired former Google, IBM, and Nvidia executive Sumit Gupta as director of products for Apple Cloud.

Apple’s latest M-4 chip delivers a next-generation GPU architecture with faster hardware-accelerated ray tracing and brings it to the iPad Pro ahead of future Macs. The M4 Neural Engine is Apple’s most powerful AI chip technology publicly announced.

Apple’s plans seem to aim at creating custom server systems for its data centers around the world. Bloomberg’s Mark Gurman refuted this idea, suggesting that Apple prefers on-device AI processing. But Apple has a long history of playing two hands of cards at once. A report suggests that Apple may never reveal these chips publicly.

Meanwhile, earlier this week, Apple released its M-4 chip inside the new iPad Pro.

The post Apple and TSMC working in AI chip for data centers appeared first on Architosh.

New Mac Pro

The new Mac Pro comes with the M2 Ultra SoC chip. This chip was also announced for availability inside the latest new Mac Studio computers also introduced today.

The new M2 Ultra chip is Apple’s most powerful custom Apple Silicon. ARM architecture-based and a complete SoC (system on a chip), the M2 Ultra combines the CPU, GPU, Neural Engine, caches, and system memory all on the same piece of silicon. While rumors have speculated about Apple delivering a monster-sized version of one of the M-series chips as the main chip on the Mac Pro, the company hasn’t taken that approach.

Apple finally brings Apple Silicon to workstation-class PCIe expansion. While not a machine for CAD and BIM users, the new 2023 Mac Pro combines the sheer power of the M2 Ultra chip with loads of cutting-edge expansion options and up to 192 GB of unified memory.

This will no doubt disappoint hardcore professional Mac users who were hoping for something truly extraordinary. What Apple has delivered instead is much more prosaic. The good news, however, is that finally has Apple Silicon inside its full-featured and expandable Mac Pro workstation.

Key Specs for Mac Pro

Here are some of the key performance specs of the new Mac Pro:

• M2 Ultra SoC chip (24 cores)
• 64 (Base) to 192 GB total system memory
• 16 Performance Cores
• 8 Efficiency Cores
• 60 – 76 GPU Cores
• 30 – 32 Neural Engine Cores
• 1 TB (SSD in Base model) to 8TB
• 6x full-length PCIe Gen4 slots
• 1x half-length x4 PCIe Gen3 slot with Apple I/O card installed
• 300W aux power

The new Mac Pro also features a Media Engine with the power of seven Apple Afterburner cards on silicon for hardware-accelerated H.264, HEVC, ProRes, and ProRes RAW. There are 2x video decode engines, 4x video encode engines, and four ProRes encode and decode engines.

Apple’s new Apple Silicon powered Mac Pro comes in a rack-mounted version. This machine will likely find many customers in film and special effects studios, science and medicine, and government agencies running specialized software.

The tower model starts at USD 6,999, and the rack model at USD 7,499. Prices rise with upgraded specs. The Mac Pro ships starting 13 June 2023.

To learn more about the Mac Pro, you can go here.

Architosh Commentary and Analysis

OTOY’s Octane X renderer is a GPU-accelerated rendering software application. The Mac Pro with M2 Ultra and 60-core GPU (not even the 76-core GPU) is up to 7.6x faster than the previous version with 8-core Intel Xeon W and Radeon Pro W5500X GPU. If you configure the previous Intel Mac Pro with a 28-core Intel Xeon W and outfit the workstation with the available Radeon Pro W6900X, the same Apple Silicon-based new Mac Pro is still 2.7x faster on the same OTOY Octane render. Other performance differences can be found here. The Autodesk Maya comparison is also quite substantial in the performance difference. 

In terms of raw performance, the M2 Ultra doesn’t yet have Geekbench processor scores up on the site (as of this writing). However, the M2 Max found in the 2023 16-inch MacBook Pro has a single core (SC) score of 2729. As a point of comparison, the Geekbench score for the last Mac Pro model (2019 Mac Pro with Intel Xeon, 28-core) has a score of 1377. The fastest processor on the Geekbench board is the Intel Core i9 (13900 KS), with a score of 3083. The AMD Ryzen 9 7950X has a score of 2874. 

Apple states that the M2 Ultra is 40 percent faster than the M1 Ultra found in the first generation of the Mac Studio. That chip at Geekbench has an SC score of 2378. If we increase that by 40 percent, we can estimate the Geekbench SC score at 3329. That would mean the new M2 Ultra would be the Geekbench industry leader for single-core performance. 

Architects and other CAD and BIM users may feel disappointed that this machine is so expensive. They need to realize that this machine isn’t built for their industry. The big trade-off with macOS versus Windows in the AEC industry is now coming down to a single question: do you want upgradable GPU cards for the future?

It isn’t a question about performance anymore. Apple Silicon Macs today easily compete with the fastest PC computers. This new Mac Pro features the same exact M2 Ultra that can be found inside the new Mac Studio, which starts at USD 1,999 (for M2 Max) or USD 3,999 (for M2 Ultra). 

The Mac Pro user is spending an additional USD 2,000 for the workstation tower enclosure with state-of-the-art I/O (input/output) and expansion capabilities. Given the price of the kinds of PCIe cards that might go in a new Mac Pro, that cost is trivial. 

The post Apple Intros Apple Silicon-based Mac Pro appeared first on Architosh.

A Game Changer?

Gurman says Apple is entering a highly competitive market already led by Meta Platforms Inc but will take a “novel approach to virtual meetings and immersive video” as the Cupertino company aims to shake up the industry. (see: Xpresso No. 43, which is focused on XR/MR/AR/VR, including news on what Meta is doing in the space.) The Apple Technology Development Group has been working on the headset for seven years.

So how will the Apple Reality One be different?

Aside from offering a stunning Apple-designed operating system that feels a part of its Apple ecosystem—which offers benefits over rivals on multiple levels—the Reality One will have advanced eye and hand-tracking capabilities. Gurman also says it will feature FaceTime-based videoconferencing and [virtual] meeting rooms.

An External Monitor[s?]

Reality One will also be able to function as an external monitor for a Mac computer, plus replicate numerous functions on an iPhone and iPad. Like Meta’s newest headset, the Reality One will function as both a VR and AR device with pass-through video, enabling the user to wear it and see the real world around them. Varjo’s devices already deliver this feature and have for a while now. (see image below).  From this perspective, Apple and Meta are only playing catchup to Varjo. Even the word “reality” seems a bit borrowed from Varjo’s Reality Cloud.

This screen capture image from Varjo’s technology video illustrates that the two XR-3-wearing users can occupy the same space, the second person being teleported to the first person’s location where the XR-3 captures its surroundings and streams them via Reality Cloud to anywhere in the world. Apple’s Reality One will cost as much as the high-end Varjo headsets but with consumer market aspirations. Varjo is directly centered on high-end industrial use cases.

Interestingly, the Apple Reality One will feature a Digital Crown like the Apple Watch (likely much bigger, of course). Users will rotate the crown to switch from VR to AR video pass-through mode. Gurman says that the Digital Crown will be a highlight of the product. I suspect that’s because, in the switching (moving the digital crown), the VR display will “fade out and fade in” over the real-world video pass-through imagery.

Such a function would allow direct user interaction between the AR and VR imagery in a highly synchronized way and ultimately provide a type of x-ray vision potential for numerous uses and functions. It would benefit architects and designers who design and shape the real world, not to mention those who construct the world like general contractors.

Starting Slowly

Gurman says Apple’s plans are to release the Reality One this year, later in the year, and possibly announce it as early as this spring. The Reality One isn’t expected to ship in mass volume for quite some time as Cupertino takes on this market very slowly (think the Apple Watch or Apple TV+).

Gurman says that immersive video watching will be a core feature of the device, and the company is apparently discussing VR content production plans with the likes of companies like Walt Disney Co. and others. Apple will also update Apple TV+ content for the Reality One.

A version of the M2 processor will power the Reality One, plus a dedicated graphics and XR chip called the Reality Processor.

Impact for AEC

It is far too soon to understand the impact Reality One may have on the AEC/CO market and markets in general. What we know about Varjo is that their device was so amazing and capable that it ended up being a hit in high-end markets like aerospace and simulation and far less important than in AEC.

The AEC market will need a mass-market device, and Apple is already thinking of a second version that will be half the price of the estimated USD 3,000 Reality One.

To learn more details, visit Bloomberg’s report here. 

Architosh Analysis and Commentary

The feature described by Gurman about the Apple Crown fading in and out the VR to AR views seems incredible in terms of its potential. We can immediately see the industrial functional use for such features. This would apply strongly to the AEC/O industries and numerous other industrial applications and fields. 

The AEC/O Mac side of the VR market has been missing out on all the good action. Apple’s Reality One will finally give Mac users a really excellent pathway into immersive technologies. It is unclear if the Reality One will upstage the rest of the field (especially at the very high end where the Varjo devices sit). What we imagine is this will be an inflection point in the VR/AR/XR industry at the consumer awareness and usage level. 

The post Gurman says Apple Reality One Headset will have iOS-like Interface appeared first on Architosh.

New MacBook

The new MacBook features either a new M2 Pro with up to 12-cores on the CPU, 19-cores on the GPU, and up to 32GB of unified memory. That configuration features a 200GB/s memory bandwidth. The M2 Max keeps to 12-cores on the CPU side, delivers up to 38-cores on the GPU, and up to 96GB of unified memory, plus doubles the memory bandwidth to 400GB/s.

The new MacBook Pro in 14-inch and 16-inch configurations feature brand-new M2 Pro and M2 Max chips.

Apple is touting the new M2 Max, the most powerful and efficient chip ever in a professional laptop computer. The New MacBook comes in 14-inch and 16-inch models, and each can be configured with the two chip options.

The new 16-inch model has the longest battery life ever in a Mac—22 hours of video playback. You can watch movies on the battery on a flight from the US to Singapore with no problem.

We won’t cover all the other details on this new MacBook, as you can find them all here at Apple.

Architosh Analysis and Commentary

The new M2 Max configurations are truly compelling in terms of what a mobile workstation could be. With nearly 100 GB of unified memory and up to 8TB of storage when paired with the M2 Max SoC Apple Silicon processor, you have all the firepower you could ever want to do truly heavy work with CAD, 3D software, BIM, photography, and film and special effects editing solutions. 

You can take the unified memory from 32 GB to 64 GB for USD 400 or double that for double the cost. Both configurations would be ideal for CAD?BIM or 3D users on the Mac platform, not to mention having enough memory for running Windows in a virtual machine. 

The post Apple Intros new M2 Pro and M2 Max MacBook Pro appeared first on Architosh.

The M2, like the M1, is built on a 5-nanometer node but is built on enhanced second-generation process technology. Despite this node “enhancement”, Apple has eeked out significant performance improvement over the M1 processor. Let’s dig into what this means for professional apps like CAD and 3D.

First the Macs

So the M2—which will likely be the first in a series of M2-derived chips—goes into the Apple MacBook Air and MacBook Pro 13-in models. Apple is largely following a logic first deployed with the M1, taking new Apple Silicon into the company’s most popular computer products (their lightest and most affordable laptops).

Apple’s new MacBook Pro 13-in. model features the new M2 processor.

The logic of this makes further sense because the new M2 while over-performing the M1 does not over-perform higher-end models based on the M1 Pro and M1 Max. Recall that the M1 Pro offers up to 32GB of unified memory and up to 200GB/s memory bandwidth.

The new M2 offers up to 24GB of unified memory and 100GB/s memory bandwidth, still below the M1 Pro and M1 Max.

M2 Performance

Apple testing shows the M2 to be 1.4x faster than the M1 chip. It has a significantly faster GPU in particular and 50 percent faster memory bandwidth. Now that its memory bandwidth is at 100GB/s it exceeds the theoretical maximum memory bandwidth for Intel Core X-series processors per this Intel calculation.

Apple Silicon has now moved to the M2 chip. Although just based on a process node enhancement, Apple yields non-trivial performance improvements.

Apple says the M2 is 18 percent faster at the CPU level than the M1. In our own M1-based Mac mini tests from a year ago, we achieved a Geekbench 5 score of 1748. An 18 percent improvement on that is another 314.6 (315 Geekbench 5 points, yielding a grand total score of 2,063.

Apple doesn’t use Geekbench 5 to confirm that 18 percent faster CPU performance, but rather Final Cut Pro tests. So we really don’t know if the M2 achieves up to 2,000 on the single-core Geekbench test. If it did, however, it would essentially tie the top Intel i9 12900K chip. (see our Guide to Selecting Optimal CPU for Your Workstation section inside this Xpresso newsletter issue).

New MacBook Pro 13 for CAD/3D

Originally, the M1 while lauded for its single-core and multi-core CPU performance, did not necessarily win over the critics for its GPU performance. So the 35 percent improvement in GPU performance by the M2 is definitely appreciable.

As far as a mobile CAD or BIM laptop is concerned, an M2-based MacBook Pro 13-in model would make an excellent machine to have in the home office or to carry around to job sites. Since most CAD, BIM, and 3D modeling tools like SketchUp are single-threaded codebases, the M2 would run everything from SketchUp to SolidWorks as fast if not faster than just about any chip out there. Now SolidWorks is not written for macOS much less Apple Silicon, but if it was, it would perform very well on the M2. Sames goes for Revit, another frequency-bound, dominantly single-threaded BIM tool that is also not written for macOS.

As we learned from the Vectorworks folks, Apple’s system on a chip (SoC) approach with Apple Silicon has advantages that are simply not there to be had on Intel X86 architectures where the memory for the GPU and the memory for the CPU are separate things with an interconnect between them. As Steve Johnson, senior vice president of Vectorworks told Architosh, “With Apple Silicon, we don’t have to copy so much data around memory because of the System on Chip’s integrated memory architecture.” (see: Architosh, “The Gift That Keeps On Giving—Vectorworks Talks Apple Silicon,” 3 May 2022)

The M2 is significantly larger than the M1, with 20 billion transistors versus 16 billion for the M1. A key difference is making room for 10 GPU cores versus the 8 on the M1.

It especially means for BIM projects in AEC, those large models move around much faster when data can be cached. As Johnson says about the BIM program Vectorworks Architect, which is fully native to Apple Silicon, “we love it because the caches make these transitions from 3D views to 2D plan views much faster.”

Apple’s new M2 Macs mark the first step toward a new set of upgrades across their Apple Silicon-based product line. The one Mac not yet on Apple Silicon is the new Mac Pro. We now expect that machine to be introduced only after Apple prepares M2 Pro, M2 Max, and M2 Ultra-based Macs. In other words, Apple must go through its entire Mac product line first and this is because of the modularity of Apple Silicon. The upcoming Mac Pro is rumored to feature 2x Ultra chips (or 4x Max chips).

Architosh Analysis

For a more detailed look at the differences between the M1 and M2 chips, AnandTech has taken a deep dive. 

The M2 does feature “architectural” changes, including possible new “Avalanche” performance cores and “Blizzard” efficiency cores in the M2 versus the “Firestorm” and “Icestorm” cores of the M1. AnandTech suggests this because of the timing (how long it has been since M1). The publication also notes that the M1 versus the M2 is similar to the A14 versus the A15. 

They state, “the M2 looks a lot like a derivative of the A15 SoC, similar to how the M1 was derived from A14. As a result, at first glance, the M1 to M2 upgrade looks quite similar to the A14 to A15 upgrade. 

The M2 takes the total transistor count from 16 billion in the M1 to 20 billion. This is a significant increase. The M1 Pro has 33.7 billion transistors or essentially 2x times what the M1 has. That provides a bit of comparison for folks on the line between a larger MacBook Pro and the new M2-based MacBook Pro 13-in. 

The post Apple Unveils new M2 Chip in New Macs appeared first on Architosh.

Dr. James Thompson, Qualcomm’s chief technology officer, made the announcement this week at the company’s investor day event.

Nuvia Is Key

As has been written months ago by Architosh, Nuvia is the company whose expertise is critical in Qualcomm’s chip plans to go toe-to-toe with the Apple M-series SoC chips in the personal computer space. Qualcomm acquired Nuvia earlier this year for USD 1.4 billion. Nuvia itself was formed in 2019 by notably Apple’s chief chip architecture Gerald Williams III along with two other star chip designers with stints at Google and Apple.

Qualcomm says its Nuvia chip team is very far along in the development of an ARM-based SoC aimed at performance leadership in the PC space. (Image: The Verge)

While Nuvia’s original goals were to attack the server chip space, which has long been dominated by Intel and where AMD has made huge market share gains in recent years, Nuvia’s expertise is now being directed at the general PC market where Apple has stolen the performance per watt crown in a big way from Intel and AMD.

Bigger Ambitions

Qualcomm is looking to get chip samples to customers in about nine months’ time with products shipping next year with Qualcomm’s yet-named SoC (system on chip) silicon. And get this, Qualcomm isn’t just aiming at dominance over Apple at performance per watt, they are aiming at sustained performance.

Skeptics will say that Qualcomm tried entering the PC chip market before and failed, with the Snapdragon 8cx line of processors, but this time the company literally has the advantage of the talent of Nuvia’s chip team. Williams III was a significant part of the success of Apple’s A-series chips driving the iPhone. This may be the critical piece that lends Qualcomm success this time around.

Architosh Analysis and Commentary

The implications of the rise of ARM into the same PC territory as Intel and AMD’s X86 architecture have been detailed in our article from Xpresso #31, in the special feature titled, “Chip Technology, Geopolitics, and the CAD Industry.” 

We make the point that today’s CAD software contains millions of lines of code with hundreds of dependencies, including links to modeling kernels, rendering engines and so much more. Moving the CAD industry so it can support both X86 and ARM chip architectures will be a bigger task for some players than others. Some, like Nemetschek’s daughter company, Vectorworks, Inc., have already completed the task and are now on both X86 for Apple’s macOS and X86 for both Windows and macOS. 

In the past, some companies, like PTC (Parametric Technology Corporation) failed to acknowledge a sea change in platforms fast enough and lost huge ground to SolidWorks when the Windows PC began to dominate and take market share away from UNIX CAD platforms. This same process can repeat again for some of the biggest CAD companies in existence. If Qualcomm is truly just nine months away from shipping SoC chips that can rival Apple’s M-series processors—and why would they not be able to when the team responsible holds leaders from Apple’s chip team that brought about the A and M-series Apple Silicon?—then the Windows world will soon be aflush in foundational change. 

The post Qualcomm Aims to Take Chip Leadership with new ARM SoCs for Windows appeared first on Architosh.

Fixing the Multi-Core Issue

I don’t want readers to misunderstand that last sentence, however. The M1 is an incredibly powerful chip for its cost and it divides its 8 CPU cores into 4 Firestorm and 4 Ice-Storm cores. Balanced symmetrically between performance and energy efficiency, the M1 made sense for the types of Mac computers Apple had targetted the chip for.

Now with the MacBook Pro, Apple is thinking that pro users are going to bias workflows that put slightly more emphasis on powerful time-saving functionality over battery life. The new MacBook Pros still get incredibly battery life, but these machines are for power users. Where the M1 MacBook Air is ideal for say a journalist or author in a remote location who needs to get a lot of writing done and needs maximum battery-life, the M1 Pro or M1 Max-based MacBook Pro is for the film editor, architect, 3D animator, or coder who needs to get work done faster even when unplugged.

Apple’s new MacBook Pro comes in 14-in. and 16-in. models, and are powered by either the M1 Pro or M1 Max SoC (chip), offering industry-leading performance per watt across both CPU, GPU, and machine learning functions.

Apple has rebalanced things. The M1 Pro and M1 Max feature 10 cores and 8 of them are Firestorm cores, while only 2 of them are Ice Storm cores. The result is these new MacBook Pros deliver double the number of Firestorm cores (8) and that makes a big impact on multi-core workloads like 3D rendering workflows, film editing workflows, science and engineering workflows, and 3D CAD/BIM workflows in AEC where more processing these days is multi-threaded. In short, Apple has totally fixed the multi-core issue, giving pro users whose apps use more “multithreaded” the power they really want.

next page: on the next page we cover the critical Geekbench numbers and review in our analysis section Jade C-Die and Jade C-Chop notes from Bloomberg’s Mark Gurman. 

The post Thoughts on Apple’s M1 Pro, Max and MacBook Pros appeared first on Architosh.

One Chip, Not Two

It appears that there is a noticeable “cut line” in the M1 Max chip. If this chip is actually the die coming off the 300mm TSMC wafer for Apple, it would explain Gurman’s Jade C-Die version of the chip, which he implied would be more powerful than Jade C-Chop.

Apple’s three M-series chips for Macs. The M1 Pro is actually a version of the M1 Max that is cut across the dark lines in the lower center of the image far right.

The M1 Pro appears to simply be the northern part of the M1 Max chip as seen in the image. A close look reveals perfect symmetry about the cut-line shown in the lower image below.

MORE: Apple Will Introduce M1X Chiplet Technology On Monday

That cut line halves the possible GPU cores from 32 to 16 cores. Both chips have the same maximum 10 CPU cores. Gurman spoke of future chips having 20 and 40 cores, which is why we wrote yesterday that a chiplet strategy would allow Apple to place this chip onto a “chiplet” package and double and quadruple the chip to arrive at ever more powerful versions of Apple Silicon for desktop Macs.

Apple’s M1 Pro is “Jade C-Chop” per Mark Gurman’s late spring report. (click image for larger view).

Ironically, Apple could still go this way but given how they approached this M1 update, I’m beginning to question my analysis on how Apple will arrive at 20 and 40 CPU core options. On the other hand, what is Apple doing with the chopped 16 GPU core portion of the Jade C-Die? Would such an approach possibly make sense? How could this factor into the edge of the wafer with the circular shape intrudes on rectangular chips? Could these chopped-off segments be utilized in a chiplet strategy for the next M chips for the Apple Mac Pro and larger iMac so they are not wasted?

Have any thoughts? Email us at anthony {at} architosh {dot} com. Or comment below by logging in.

The post Apple’s new M1 Pro is “Chop” version of M1 Max Die (?) appeared first on Architosh.

This graphic is a telltale sign that Apple intends to blow past the entire industry in chip performance. So how do they intend to do that exactly? That’s what we focus on in this article. We will first summarize and then we will provide details for those who want to understand the tech behind the story.

The Summary

Apple’s A-series and M-series chips are SoCs, otherwise known as “system on a chip.” Its latest SoCs include the CPU, GPU, Neural Engine, and Memory, and much more directly on the same piece of silicon.

However, Apple will almost certainly jump to a SoIC type of semiconductor product tomorrow. SoIC stands for “system on integrated chips.” This technology is also known as “chiplets” and AMD is already first to the market with chiplet (SoIC) technology with its Ryzen 5000 series CPUs.

Chiplet technology is a new way to package a processor. If you have been wondering how Apple was going to scale up the unified memory on the M1 processor to meet the memory demands of its future professional Mac computers, chiplet technology is how you can get there. It is also the pathway to many more cores.

On Monday Apple will likely introduce the M1X (our name not Apple’s) as a chiplet. In one version the expected 10-core M1X will sit on an interposer layer in a SoIC package. The interposer is an electrical interface for routing between various components of the SoIC. Take a look at the patent graphic above.

Apple tomorrow will likely enter the chiplet (SoIC) era for Apple Silicon. Manufacturing challenges at smaller nodes are forcing the industry in the chiplet direction, and wafer yields favor small chips anyway. AMD has already led the way ahead of both Apple and Intel with a chiplet CPU implementation.

When the M1X is doubled up in chiplet design you can double both the GPU and CPU cores. If you place four M-series chiplets on a larger SoIC package, you have the processing power you need for the new Mac Pro towers. We believe that the Mac Pro, however, will feature the M2 chip in a chiplet package.

Details

Mark Gurman of Bloomberg has been one of the most reliable sources for the future of Apple products. In May of this year, Gurman wrote the new 14-inch and 16-inch MacBook Pro would be powered by two new chips code-named Jade C-Chop and Jade C-Die. Each will be 10-cores (8 high-performance cores and 2 energy-efficient cores). Both chips will come in 16 and 32 GPU core variants with up to 64GB of RAM.

But these are not different chips but rather TSMC will chop off half the GPU cores to arrive at a chiplet that is smaller. Or these will be binned chips that had defects with Apple disabling GPU cores or simply turning good cores off like in the M1 MacBook Air with its 7-core GPU.

Smaller chips are superior for better wafer yields. All wafers tend to have some small percentage of defective or low-performing chips—see red and purple squares. When a defect occurs on a large monolithic CPU chip, like Intel’s recent chips, the total wafer yield is far poorer than when the same number and type of defects occur on a wafer with a far greater number of smaller chips. The chiplet strategy will enable Apple to get optimal chip yields out of each TSMC wafer. And “binned” chips can still be used for throttling M1X performance options, allowing Apple to appear to be utilizing many CPUs when in fact they are actually only manufacturing one.

Therefore, the M1X may have several binned chip variants, enabling Apple to maximize wafer yields by offering versions of the chip with disabled defective cores. The M1X may come in 10-core and 8-core versions, with the latter still performing better than the M1 due to Apple moving to the TSMC 5NP process node (P is for the performance version of the 5nm node). Variants of the Jade C-Chop may come with 10, 12, and 14 GPU cores. We believe the Jade C-Die chip is a chiplet (SoIC) package with two (2x) M1X chips. Again, Apple can offer disabled binned variants with versions made for the larger 16-inch MacBook Pro offering 20, 24, or 28 GPU cores and possibly a 16 CPU core option.

With the new modularity of chiplet (SoIC) design, Apple can offer the market what seems like a plethora of chip and price options while only architecting and manufacturing a single M1X chip. The flagship version for the new MacBooks would likely be the Jade C-Die with 20 CPU cores, 32 GPU cores, and 64 GB of unified memory.

Mac Pro on M2

Gurman also writes that the new Mac Pro will have new chips dubbed Jade 2C-Die and Jade 4C-Die, in 20 and 40 core variants. What’s intriguing about these codenames is the numbers 2 and 4. Do they imply a doubling and quadrupling of the basic M1X chip? That would be the most logical notion. But what would the performance difference be between a Jade 2C-Die versus a Jade C-Die? The answer may lie in chip frequency.

Since the Mac Pro has completely different thermals as the enclosure is vastly larger than all the other types of Mac computers, Apple could afford to increase the clock frequencies of the M1X. Another explanation could be that Jade 2C-Die and Jade 4C-Die are chips built on the next 4nm or 3nm TSMC process. Since the next Mac Pro is due next year I think the Jade 2C-Die and 4C-Die are not built from M1X chiplets but rather M2 chiplets. That’s my guestimate based on the timing of the Taiwan Semiconductor Manufacturing Company (TSMC’s) 4nm node process.

Lightspeed

Bottom line? The new chiplet era for Apple Silicon is likely to begin tomorrow. They are following in AMD’s footsteps and the entire industry will ultimately have to go in this direction as well. But just imagine how much faster the M1X will be in its basic binned version, let alone its 10-core CPU / 16-core GPU and with 32GB of unified memory.

A basic indication is to look at the speed up of the A-15 Bionic over the A-14, both on 5nm cores. The A-15 however was built using the TSMC N5P node and gained, in the end, a larger performance increase than was first imagined. That iPhone 13 chip is faster than the M1. The M1X will benefit from the same TSMC N5P node improvements over the M1. It will also be larger than the M1 with likey about 30-40 percent more transistors. It will likely yield a 25-35 percent performance improvement in single-core performance. The real performance story, however, is the use of two of these M1X chiplets on a single die. This will scale multicore performance up dramatically north of 200 percent over the M1. And then later with the M2 chiplet in 4x formation, we may see Apple’s next Mac Pro become the most powerful workstation computer in the world by a long shot. This is what seems possible now.

The post Apple Will Introduce M1X Chiplet Technology On Monday appeared first on Architosh.

We consider this one of our best special reports published in the past few years. The long-form article makes a sweeping arc through the semiconductor industry’s present situation, including the heightened tensions between the United States and China and the emergence of the US Chips Act.

The article also mentions what China is lacking in technology and native industry in order to gain full semiconductor independence. Even if China invaded Taiwan that country has zero native manufacturing of critical tools used to make semiconductors.

The point of the article is to educate the reader on the state of semiconductors and their impacts on the CAD market.

Rising ARM and Falling Intel

The article looks at Moore’s Law and highlights the exact year timeframes when both the ARM architecture began to accelerate (in many ways thanks to Apple) and when the Intel X86 architecture began to falter. An entire section delves into the recent history of Intel’s chip manufacturing problems and answers the question “what specifically has been Intel’s chip manufacturing issues?”

Xpresso No. 31 features a long-form article on the semiconductor industry and its impacts on the world of software, especially CAD industry software. Readers can access this special feature by signing up for Xpress today.

The article also looks at Apple’s recent chip “brain drain” as its chief CPU architect left in 2019 to form Nuvia which was acquired by Qualcomm. Secondly, another wave, or, exodus occurred more recently heading to the RISC-V start-up Rivos.

The article ties together the PA Semi acquisition from 2007 and why Steve Jobs wanted to buy it. Even Jobs had just brought the Mac over to Intel a few years before, PA Semi showcased to Jobs a stunning high-performance, low-power PowerPC architecture chip that convinced Apple to acquire PA Semi and use their team as the foundation of Apple’s semiconductor group.

Apple currently crushes both Intel and AMD in world-leading single-core performance versus wattage, yet Nuvia has mapped out even more superior single-core performance aiming initially at the data center. Qualcomm has different plans for the Nuvia team, while Rivos’s plans for RISC-V architecture dominance center on the data center.

ASML and EUV

One of the most critical pieces of equipment for Apple, Qualcomm’s Nuvia team, Rivos, Intel, AMD, and NVIDIA’s future high-performance chips (CPUs and SoCs in particular) is ASML’s extreme ultraviolet lithography machines (EUV lithography). This equipment is critical to taking processor manufacturing nodes down to 5nm and beyond. Apple is already shipping chips using EUV from ASML, while AMD and Intel are not quite there yet. We discuss this Dutch company’s equipment and who has it and in which proportion.

ASML’s EUV lithography machines cost over 150 million USD and can take up to six months to install. (Image: ASML)

Lastly, we discuss the landmark tidal shifts in the global semiconductor industry and how the world of software—specifically CAD, BIM, and 3D software—may have to respond to these shifts. Many incumbents in the CAD industry may suffer the same fate as Intel which could not react fast enough to new types of challenges in its manufacturing. These CAD leaders are so entrenched around multi-decade “Wintel development models and technology” that the very thing that has given them streamlining and massive growth could now come to haunt them as new upstarts can target the emerging ARM-based platforms without the legacy code dependency issues that incumbents depend on.

Accessing the Special Feature

To read the feature please sign-up for our free monthly XPresso newsletter here. The system will email you and provide guidance to reading past issues, including issue No. 31 which contains this special feature.

To learn more about INSIDER Xpresso read here.

The post Xpresso #31—Chip Technology, Geopolitics and the CAD Industry appeared first on Architosh.

In a surprise move, AMD’s new series of cards include the Radeon PRO W6800X, Radeon PRO W6800X Duo graphics card, and the Radeon PRO W6900X.

AMD RDNA2 for Mac

AMD had just recently announced its new Radeon PRO W6000 series “workstation-class” GPUs for the general PC computer market back in early June. That announcement saw AMD boast of superior performance-per-price ratios compared to NVIDIA’s RTX Quadro cards, particularly offering twice the graphics memory of the NVIDIA Quadro RTX 5000 card for nearly 400 USD less cost.

AMD’s RDNA2 Architecture features hardware-accelerated real-time ray tracing technology to rival NVIDIA’s RTX technology.

“We have developed the AMD Radeon PRO W6000X series GPUs to unleash professionals’ creativity and help them bring more complex and compute-intensive projects to life, from animating 3D film assets to compositing 8K scenes to game development,” says Scott Herkelman, corporate vice president and general manager, Graphics Business Unit at AMD.

“The new AMD Radeon PRO W6000X series is packed with remarkable energy efficiency, enhanced compute units, and a new visual pipeline, enabling Mac Pro users to do more in less time across a broad range of pro applications,” he adds.

AMD’s new RDNA2-based Radeon PRO W6800X Duo for Mac.

Built on a 7nm node manufacturing process, the new AMD Radeon PRO GPUs bring to the macOS platform AMD’s RDNA2 Architecture—a chip architecture that brings up to 65 percent more performance per watt over AMD’s early generation RDNA Architecture. The RDNA2 Architecture features dedicated Ray Accelerators for real-time ray tracing, a first for an AMD GPU. The specialized ray tracing hardware handles the intersection of rays and produces an order of magnitude increase in intersection performance compared to a software implementation alone. The Ray Accelerators are just one of several ground-breaking performance enhancements for AMD’s graphics architecture.

Radeon PRO 6000 Series for Mac Details:

• AMD RDNA2 Architecture — the groundbreaking architecture first introduced at E3 in 2019 has finally arrived in the professional workstation market this June and now for the Mac Pro. RDNA2 features hardware-based Ray Accelerators (RAs) and there is one RA per Compute Unit (see below).
• AMD Infinity Cache — up to 256MB (total) of last-level data cache integrated on the GPU die is designed to reduce latency and power consumption
• AMD Infinity Fabric — Provides a high-bandwidth, low latency, a direct connection between local AMD GPUs, enabling high-speed GPU-to-GPU communications designed to satisfy today’s creative workloads. (this applies to Radeon PRO W6800X Duo card)
• High-speed GDDR6 Memory — Up to 64GB of GDDR6 memory with up to 512 GB/s bandwidth provides ultra-fast transfer speeds to power data-intensive professional applications.

The three-card options are detailed out below:

AMD Radeon PRO W6800X 

• 32GB of GDDR6 memory with 512/GB/s memory bandwidth
• 60 compute units
• 3840 stream processors
• 16 TF (32-bit) / 32 TF (16-bit)
• (see more details here)

AMD Radeon PRO W6900X

• 32GB of GDDR6 memory with 512/GB/s memory bandwidth
• 80 compute units
• 5120 stream processors
• 22.2 TF (32-bit) / 44.4 TF (16-bit)
• (see more details here)

AMD Radeon PRO W6800X Duo

• 64GB of GDDR6 memory with 512/GB/s memory bandwidth
• 120 compute units (60 per W6800X GPU)
• 7680 stream processors (combined)
• 30.2 TF (32-bit) / 60.4 TF (16-bit)
• (see more details here)

To learn more about AMD’s W6000X series GPUs for the Mac Pro please visit this page here.

Pricing and Availability

The new AMD RDNA2 Architecture-based Radeon PRO W6000X series cards are available to configure on the Apple Mac Pro workstation. Up to four Radeon PRO W6800X GPUs can be configured in a new Mac Pro using the Apple MPX Modules.

Mac-based professionals now have AMD’s new RDNA2 based Radeon PRO W6000X series GPUs for configuration inside the Mac Pro. The option shown above is for 2 – AMD Radeon PRO W6900X GPUs. Now the Mac platform has AMD hardware-based real-time Raytracing technology and a lot more for fewer costs than NVIDIA RTX cards.

Let’s look at these cards in terms of base-pricing and work ourselves up through these new cards. If you go to the Apple Store and start with the base price Mac Pro it begins at 5,999 USD. This machine is equipped with an AMD Radeon PRO 580X with 8GB of GDDR5 memory.

To upgrade to the new AMD Radeon PRO W6800X with 32GB of GDDR6 memory the price upgrade is 2,400 USD. Adding another one doubles the price essentially and a wee bit more to 5,200 USD. Adding a Radeon PRO W6900X GPU with 32GB of GDDR6 memory—a card formation that was not released back in June when AMD introduced the new RDNA2-based Radeon PROs—will cost you 5,600 USD. Adding two of those units essentially doubles the price. You can explore more price configurations, such as two Radeon PRO W6800X Duos set via the Apple MPX Modules for a combined total of 4 GPUs. Explore here.

Architosh Analysis and Commentary

The Mac-based workstation market has had a tough going of it since the failure of the 2013 Mac Pro. From the 2009 – 2012 perior and then up to 2019, there was nearly a decade-long vacancy of Apple putting in serious attention to the professional markets with a machine worthy of pro users. The one major company that was carrying about Mac professionals, however, was AMD and is Mac graphics division. They put out several GPUs aimed at Mac pros utilizing the 2009 – 2012 Mac Pro computers. We last reviewed such a card back in 2013. 

Now since the revival of Apple’s interest in pro markets and it’s stunning 2019 Mac Pro, AMD has had several GPU options for users. (a complete listing of GPUs for Mac Pros can be found here). One thing that it has not had is hardware-accelerated real-time ray tracing—something to match up against NVIDIA’s RTX technology. While RTX technology has been out for a few years (debuted in August 2018), a year ago the industry was still at a loss for what AMD’s plans were to match its rival. This isn’t to say this is AMD’s first GPU that can do ray tracing. AMD’s Radeon ProRender, aimed at the pro markets delivered targeted, professional real-time ray-tracing and did so months before NVIDIA RTX. But that technology never really took off, outside of some companies like Maxon which tapped into it. 

Hardware-accelerated real-time ray tracing has become an important technology in the pro markets—including AEC where a plethora of interactive renderers have hit the market over the past several years (Lumion, Enscape, Twinmotion, LumenRT, etc). A lot of these tools tap the power of NVIDIA RTX. AMD now has a powerful equivalent technology and will need to court pro developers to like this to boost and excite users in the AEC space, particularly those dedicated to rendering tasks.

Meanwhile, these new cards are sure to boost performance over the prior AMD Vega II and Vego PRO II GPUs in the previous Mac Pro lineup. Mac pro applications like Final Cut Pro, Cinema 4D, Octane X, DaVinci, Resolve and many others are already taking advantage of AMD GPUs. Real-time ray tracing aside, these new AMD GPUs provide great value for the dollar with their exceptional memory configurations. 

The post AMD Brings Radeon PRO W6000 Series GPUs to Mac Pro appeared first on Architosh.

Raffle for M1 Mac mini

Architosh will use a standard random number generator to select the winner of the Architosh-provided Mac mini with the new Apple M1 chip on Wednesday, 16 June 2021. The winner will be announced later in the week on Friday.

All beta group participants who completed all two phases of our work have also earned a life-time Architosh INSIDER Membership (Architosh’s subscription tier) enabling them unlimited access to all “premium-tagged” content and future INSIDER site features.

M1 25 Beta Group

The readers who joined the M1 25 Beta Group included professionals from five continents and more than nine countries, including spanning from Australia to Japan to the US to several European and South American countries.

The new Mac mini with Apple Silicon M1 chip. (Image: Apple / Architosh)

More than half of all readers have been reading the site for over 7 years. The group roughly fell into three primary reader types: (1) professional architects or AEC related readers, (2) BIM/CAD consultants and resellers, and (3) academics at Universities teaching in architecture or engineering schools.

The M1 25 Beta Group will receive a complimentary report and optionally be able to take part in future site development design reviews.

More Info

Readers can learn more about this group effort here.

The post M1 25 Beta Group – Winner of M1-based Mac mini announced this week appeared first on Architosh.

Typically each year Apple introduces major operating system updates, particularly iOS updates at each WWDC. Let’s look at a few predictions for what may be on tap for this year.

WWDC Predictions

1 – New OSs

We can certainly count on Apple previewing this year’s OS updates with the next version of macOS, iOS, iPadOS, WatchOS…you get the picture.

The folks at ZDNet have made the important distinction between “tik” and “tok” level releases with the former being major new features (a bigger update) while the latter focuses more on internal and performance improvements with less new features. Also, “tik” updates boast major new UI updates.

Apple’s Worldwide Developers Conference (WWDC) for 2021 and its keynote will likely unveil the next Apple Silicon Macs. A new 16-in. MacBook Pro will likely feature an M1X (or M2) chip purpose built for the needs of the larger pro-oriented MacBook Pro.

Both macOS and iOS will likely have true “tok” updates while iPadOS may get a major “tik” update, particularly one that takes advantage of the M1 chip inside the latest iPad Pro line.

2 – New Apple Silicon

We fully expect Apple to roll out updates to its Apple Silicon chips for the Mac product line at least on a six month pace per product in order to complete its full transition in two years. So here are the Macs remaining to be converted to Apple Silicon:

• 15-17 inch MacBook Pro
• 27-inch iMac
• 27-inch iMac Pro
• Mac Pro.

Each of these machines have TDP levels in the following ranges. Also, TDP stands for Thermal Design Profile and roughly corresponds to the maximum heat a computer chip can use in watts. Tom’s Hardware has a basic definition here, for more info. It is a basic indicator of power consumption and power-heat generation that must be accounted for in the computer design to keep the system operating at appropriate thermal levels.

• 15-17 inch MacBook Pro = 35 – 50 watts TDP range
• 27 inch iMac = 65 – 95 watt TDP range
• 27 inch iMac Pro = 125 – 150 watt TDP range
• Mac Pro = 150 – 250 watt TDP

When Apple introduced their plans for Apple Silicon they said they were working on new chips that would take advantage of every model’s TDP levels.

• Level 1 TDP = under 28 watts  (M1 chip)
• Level 2 TDP = 35 – 50 watts  (M1X chip)
• Level 3 TDP = 65 – 95 watts  (M2 chip)
• Level 4 TDP = 125 – 250 watts (M2X chip)

Apple is renowned for efficiencies and limiting the number of variants in products so we could easily see Apple consolidating their M-series chip designs rather than developing specific chips for every single Mac TDP. With the M1 included in the recent 21-in iMac line it suggests Apple will cross-deploy its chips when it makes economical sense and at the same time go full bore with chip design for its “pro” oriented Macs.

This is the current 16-in MacBook Pro shown here. The rumored Apple Silicon version will possibly operate at a sub-40 watt TDP, affording Apple the chance to dramatically lighten the machine’s chassis. As noted below, it may feature the M1X (or M2) with 10 CPU cores and be manufactured in two versions one with 16-core GPU and one with 32-GPU. Both units may feature 64GB of integrated memory, quadrupling the 16GB integrated memory on the M1.

Today at WWDC we are likely going to see the M1x chip but the name could be M2. This chip will likely be in the range of 35 – 50 watts but it could be also stretched to about 65 watts and serve double duty in the 27-inch iMac. The rumors noted thus far don’t necessary preclude that possibility.

3 – All new 16-inch MacBook Pro and Possible all new Mac mini

We are most certainly going to see one more Mac product move over to Apple Silicon today. That is most likely the 16-inch MacBook Pro.

A new M1X-powered 16-inch MacBook Pro is rumored at 10 CPU cores with Bloomberg reports suggesting it may feature 16 to 32 GPU cores. Apple should be on a “tok” update to TSMC’s 5nm process technology and 3nm chips should not be ready until the second half of 2022. We can expect the last quarter of 2022 to be the point at which the original M1 chip goes from a 5nm to 3nm process (a “tik” update). So the M1X (whatever Apple calls it) will be about 45 – 75 percent larger in transistor count with a TDP range estimated as noted above.

Finally, rumors about an all-new redesigned chassis for the Mac mini suggests that the M1X (or M2) would also fit into that machine making it a possible Mac mini Pro or Pro Mac mini. Both machines would be ideal for Apple developers and hence both are good targets for WWDC.

4 – Preview of New Mac Pro and iMac Pros

It is possible Apple will preview for us the new Mac Pro. WWDC has served as the venue for such preview introductions before for Apple’s top of the line computers. If such a thing happened Apple would announce the availability date for the new Mac Pros and iMac Pro. It might also update the Intel models in the present. This would be a surprise and most analysts don’t expect Apple to unveil these larger Macs this year at WWDC.

5 –  Big Music Related Updates

With the big announcements recently about Apple Music to support Spatial Audio format music files for all Apple Music subscribers as well as lossless file formats running from CD quality (16 bit at 44.1kHz) all the way up to Hi Resolution Lossless (24 bit at 192kHz and higher), which requires a DAC (digital to analog converter) we can see Apple introducing new music hardware in the form of an Apple DAC as well as new updates to its headphones to bring back “wired” models for audiophiles to enjoy the Hi Resolution Lossless music. Apple has promised its full Apple Music catalog of more than 75 million songs will be available eventually on lossless formats but at launch 20 million files will be ready.

The Spatial Audio update is so large a topic that Apple has announced a special event just for it that begins right after the WWDC keynote.

The post Apple WWDC 2021—A Few Predictions Before the Show appeared first on Architosh.

M1 iMacs

Let’s put all other aspects of the new iMacs aside for a moment and simply concentrate on the performance of the M1 chip with regards to CPU and GPU (graphics) performance. Recently we published an in-depth feature here on Architosh. (see, Architosh: “The M1 Mac mini vs iMac Pro 2017 vs Everyone (Part 1),” 12 April 2021). That article dives deep but let’s summarize some key items below.

All About Single-Core

Most CAD and 3D modeling software are “single-threaded” and not optimized for the utilization of multiple cores on chips. This fundamentally has to do with the mathematical realities that make 2/3D CAD software work. We’ve had recent conversations in the last year with the folks at Siemens Parasolid, for example, talking about their market-leading modeling kernel, and how much of it can be optimized for multi-threaded processing taking advantage of more CPU cores. The short answer is some parts—but of a limited nature—can be optimized for parallel processing (multithreaded) performance improvements.

Apple’s new 24-in iMac is powered by Apple’s new M1 chip, offering market-leading single-core processing and upper-midrange GPU Performance. (Image: Apple)

So whether one is talking about Revit models, SolidWorks models, or SketchUp models, the driver of performance in modeling activity workflows is primarily the single-core performance of the CPU in combination with the GPU’s performance to redraw the screen. And in this regard, we are talking about the M1’s high-performance Firestorm core(s).

Geekbench 5 

When it comes to the Apple M1 chip, the Geekbench 5 scores we have achieved and those others reported are at the top of the industry. No Intel chip matches the M1 in single-core (SC) performance and only AMD’s 5000 series (Zen 3) chips are close. In our tests, we achieved a Geekbench 5 single-core test score of 1748 compared to the 1251 reference score for a 2020 27-in. iMac i7. In our scores on our own iMac Pro (2017 model) with a Xeon processor, the result was 1100, far off the M1’s performance.

Apple’s new M1 Macs absolutely crush the Intel competition, especially on a performance-per-watts basis, enabling Apple to deliver a stunningly thin new iMac enclosure. The M1 is now in the new iPad Pro and that tells us a bit about how Apple aims to economize chip development and place chips into similar TDP enclosure criteria regardless of product line, thus iPad OS and macOS while different are tapping the same chip.

And although we found excellent material online from Nuvia—the chip company started by Gerald Williams, Apple’s former chief CPU architect, about how well Geekbench 5 accurately simulates not just PC workloads but server CPU loads of a “comprehensive general nature”—even they state that an ideal benchmark is one “that is most representative of the customer’s workload.” Therefore, it is good to test against a range of most representative benchmarks. That is why we tested with GFXbench and Cinebench as well. In the latter, the M1 essentially tied a top reference score (Intel i7 11th Gen) 4-core machine. (see image here). When you get a moment read our two-part feature as it goes deep into the M1’s performance.

next page: on the next page we discuss the importance of the GPU for CAD performance and where the M1’s GPU stacks up, plus the critical importance of RAM. 

The post Apple’s new M1 iMacs — Are They Fast For CAD/3D Users? appeared first on Architosh.

Drawings Re-positions App

The award-winning iOS app—now also available for macOS—shook up the industry with the first, truly powerful, and ultra-intuitive 3D CAD solution and famously required an Apple iPad Pro with Apple Pencil for its functionality.

With the Drawings launch, Shapr3D is now more a complete solution for companies that manufacture industrial goods. The CAD company says that Shapr3D “now supports the full concept-to-manufacturable-drawing journey.”

Shapr3D intros new Drawings features and new Business plan. Provides “full concept-to-manufacturable-drawing journey.”

Aimed at product and industrial designers, mechanical engineers, and design project managers, Shapr3D with Drawings enable these professionals to stay inside one program versus jumping back and forth between CAD platforms, in addition to saving costs on needing fewer digital tools.

New Business Plan

Drawings is exclusive to the new Shapr3D Business plan, which is unveiled on 1 March 2021. This new Business plan replaces the Pro package, for USD 499.00 per year or USD 59.00 per month. Business plan users will be able to create 2D views from a 3D body and save them as DWG or DXF files. The plan also includes unlimited CAD imports/exports, designs, and workspaces—as well as seamless integration with your desktop CAD.

“First, we created an unforgettable CAD design experience for CAD users,” says István Csanády, Founder and CEO of Shapr3D. “Now we’re going to do the same and more for businesses. Shapr3D will be the go-to app for design-focused companies looking to easily envision, design, and manufacture industrial goods. Every new feature takes us one step closer to rivaling traditional CAD giants, but with our direct modeling approach and an unwavering commitment to UX that sets us miles apart.”

MORE: Why Shapr3D Changes the iPad Pro Story From Toy to Tool

“Shapr3D is developing a tool that’s built from the ground up for capturing ideas and using the power of three-dimensional design to progress those ideas and concepts further – on a platform and with an input method that’s long underserved in the design and engineering software market. The addition of the ability to then document those ideas using the new drawing capabilities makes it even more useful,” adds Al Dean, Editor-in-Chief, DEVELOP3D.

“The addition of Drawings completes Shapr3D’s design capabilities to support concept to manufacturing. Now all key roles in the design process from Product Managers, Engineering Managers, Industrial Designers, Mechanical Engineers, and Manufacturing can take full advantage of Shapr3D’s intuitive user interface, with the flexibility of using any device they like,” says Michelle Boucher, Vice President of Engineering Software Research at research firm, Tech-Clarity. “Plus, the tiered pricing means they can take advantage of it at a very affordable price.”

“Shapr3D delivers a refreshing change for those frustrated by the complexities and user interface legacies of some other CAD tools. For those that love their Apple products, iPads, and Macs; they’ll feel particularly at home, possibly even delighted with the user interface. Their recently (and soon to be) released functionality is an intriguing evolution of the platform, aiming to be much more in tune with the needs of those involved in product design, engineering, and manufacturing.” Allan Behrens, Managing Director, Taxal Limited

Pricing Restructure

The new pricing comes into effect on 8 March 2021, and existing Pro users will be upgraded at no extra cost to the business package as long as their account is active. Prospective users can also take advantage of Shapr3D’s special promotion in the run-up to the new release: anyone purchasing the Pro version between now until March 8 will only pay USD 249.00 and will be eligible for the automatic upgrade, also.

To learn more visit them online here.

Architosh Analysis and Commentary

Shapr3D continues to advance towards more serious MCAD territory but with a bold vision and confidence about where—or what platforms and devices—the future of the CAD industry will be played out. Csanady and his young company have seen remarkable success developing its only product simply on the Apple iPad Pro with Apple Pencil as its only UX input device. If that isn’t evidence of disruptive potential underfoot, I don’t know what is. 

Recently they have taken their codebase to the new M1 chip-based Macs and are also thinking of Windows—that platform that, ironically represents the era of 90’s CAD establishment. Shapr3D is a company that is long on Apple. That puts them in the same category as Warren Buffett, which is always a comforting place to be.  (editor’s note: Apple is Berkshire Hathaway’s third-largest shareholding.).

The post Shapr3D Intros new Drawings Feature and New Packaging appeared first on Architosh.

To read the feature now, you can sign-up for Xpresso here. Once signed up you will be directed to the first issue and from that issue you can visit the newsletter’s Archive directory where you can open up the latest issue.

CAD and 3D Tested

Our in-depth feature goes through numerous benchmark suites testing both the CPU and the GPU, including specific GPU benchmarks applicable to pro applications like 2/3D CAD, BIM and 3D modeling and rendering programs.

Architosh has conducted an in-depth performance analysis (shoot-out) between the Apple Silicon-based Mac mini vs a powerful iMac Pro machine.

In addition to benchmarks, we have tested both machines against test files in SketchUp and Vectorworks Architect (two tools still in Rosetta mode only on the M1 Mac mini). Despite not having a native application available where we have well-developed test files, the M1-based Mac mini more than held its own against the far more expensive iMac Pro (2017).

We also include an in-depth analysis and discuss on ARM microarchitecture advantages over X86.

Click here to sign-up and gain access to this special feature inside the latest (February) issue of Xpresso.

The post Architosh publishes M1 Mac mini vs iMac Pro Shootout appeared first on Architosh.

Vulkan on Apple

Apple has forbidden Vulkan drivers on their macOS and other operating systems, eschewing a cross-platform graphics API framework for its proprietary Metal graphics API. This, of course, contributes to 3D graphics platform fragmentation which makes it harder and more expensive for software developers to write code and deploy it once to multiple operating systems and devices. The net result of that is less software choice. And on Apple’s minority share macOS platform, it has meant fewer games and professional 3D graphics applications than on Windows.

MORE: Apple Introduces Revolutionary New Macs with M1 Chip

The Khronos® Vulkan® Portability™ Initiative fights this fragmentation by encouraging and enabling layered implementations of subsets of Vulkan over Apple Metal, Microsoft DX12, and other APIs.

Khronos says that much of the initial interest in Vulkan portability has been centered on layering Vulkan functionality over Apple’s Metal API, as native Vulkan drivers are not permitted on macOS, iOS, or other Apple platforms.

Developing Vulkan applications for Apple Platforms on macOS.

There has been significant effort behind Vulkan Portability efforts for macOS, and there are multiple royalty-free, open-source Vulkan Portability implementations over Metal, including MoltenVK and gfx-portability. The former has recently added many new features, including support for Apple Silicon, Metal 3.0, Mac Catalyst, and more.

macOS SDK for Vulkan

As more developers use Vulkan Portability on Apple platforms the need to further improve associated development tools grows more urgent. LunarG, who develops and supports Vulkan open-source tooling, has previously ported the Vulkan SDK to macOS, enabling developers to build, run, and debug their Vulkan applications on Macs.

LunarG is now shipping Device Simulation (DevSim) and Validation layers for the Vulkan SDK on macOS in addition to Linux and Windows.

To learn more visit this blog post.

The post Khronos Group and LunarG—Vulkan SDK Ported to Apple Platforms appeared first on Architosh.

Up to 32 Cores

Apple’s intent on delivering a series of new Mac processors that are faster than Intel’s best processors. That might not be particularly hard to do given Intel’s various chip manufacturing problems and the already stunning performance of the new Apple Silicon M1 chip.

According to this Bloomberg report, Apple wants to be out of Intel by 2022. That suggests that the chip transition from Intel to Apple Silicon may come a bit faster than anticipated. When Apple made its last chip transition, that too went faster than expected.

Apple’s plans for the next versions of Apple Silicon Macs have been shared with a Bloomberg reporter by a source close to the matter. Apple aims to outperform the competition (Intel specifically) at every level of common computer configuration, from trim laptops like in the MacBook Air to powerful professional workstations. The future revised Mac Pro may feature a 32-core ARM-based Apple chip.

The new Mac chip (SoC processor) would be used in the MacBook Pro, the consumer level and pro-level of the iMac, and the very exciting upcoming redux of the already stunning Mac Pro workstation. The current Mac Pro ships with as much as 28 cores, so a 32-core Apple Silicon chip would make an amazing advancement.

16 Cores

Bloomberg reports that the iMacs and MacBook Pro may feature up to 16 power cores and 4 efficiency cores. And they are also developing graphics processors slated to be several times more powerful than what Nvidia and AMD have supplied to Apple in Intel Macs.

To read the full reports at Bloomberg go here and here.

Architosh Analysis and Commentary

None of this is surprising. We are especially not surprised by the fact that Apple appears to be more ambitious than other analysis expected. Apple’s recent statements in various post-M1 interviews strongly suggest—if you are willing to read between the lines—that the company is being coy about its plans, which suggests the possibility that they have even bigger and more stunning unveilings in their future. 

Apple is essentially moving its computer platform off the flatter trajectory of x86 processor performance onto the much steeper trajectory of ARM-based smartphone processors. This is the bigger point and shouldn’t be lost in the details over technical specifics and comparisons. Apple’s manufacturing partner, TSMC, is also road mapped for 3nm processors in a couple more years’ time. 

The post Apple’s future Mac Pro with Apple Silicon to sport 32-cores appeared first on Architosh.

Mozart was designed to future-proof AI software evolution at the silicon level, by introducing a radically different paradigm.

Composable Computing

SimpleMachines has introduced a new type of AI chip paradigm the company calls “composable computing,” addressing the fact that AI and machine learning algorithms are constantly evolving, making it difficult for chip designers to target their designs to optimize software that is intensely shifting. So fast is algorithms in AI that purpose-built custom AI chips are often obsolete on arrival because better AI algorithms have already been written.

Key Takeaway

SimpleMachines’ Mozart AI chip is a big departure from the current AI/ML chip options, including Apple’s Neural Engine silicon. It also boasts 35 8-bit TOPS at just 4 watts, compared to Qualcomm’s new Snapdragon 888 processor with its 6th generation AI Engine cranking out 26 8-bit TOPS. The chip sits on a PCIe card. Theoretically, this card can eventually enter Windows or even Mac Pro workstations, but today it assumes Linux backend servers in the cloud.  

At the same time, general CPUs and GPUs can take on AI and machine learning software applications but are highly inefficient. Think for example about why Apple has developed a “neural engine” in their own chips, as a way to specifically address machine learning applications for AI.

SimpleMachines, Inc.’s new Mozart AI chip and platform for faster AI and ML initiatives.

What the AI industry has been needing is a new paradigm that offers the performance benefits of custom silicon optimization and while offering the flexibility of general CPU and GPU processors. This is where Mozart comes in.

SimpleMachines’ breakthrough innovation is in the discovery of four “course-grained behaviors” that compose any algorithm. By ignoring application semantics at the silicon level and instead focus on extracting behaviors, SimpleMachines’ Mozart chip compiler can compile arbitrary programs into these four course-grained behaviors, embedded into TensorFlow.

Breakthrough Results

SimpleMachines’s AI chip with its composable behavior execution units moves source code through SimpleMachines’ behavior decomposition compiler and sorts machine code into four algorithmic behavior execution units. The resulting performance is application compute at 85 percent with minimal overhead at 15 percent, a near inverse of how the same code would work over an Intel Skylake CPU.

“The chip’s design can support very large models today and is capable of running up to 64 different models simultaneously,” said Greg Wright, SMI’s Chief Architect. “Our next-generation 7-nanometer design is expected to be ready to sample by the end of 2021 and will be 20x faster on a diverse set of workloads than current chips.”

“SimpleMachines’s solution is a radically new software-centric approach that deploys a programmable platform with a breakthrough software stack and compiler that enables the programmer to easily optimize the hardware on the fly and get the performance of custom silicon with a platform that supports hundreds of different use cases,” Wright said.

World-class Team

SimpleMachines, Inc. (SMI) is led by its founder, CEO, and CTO, Karu Sankaralingam, who is also a computer science professor at the University of Wisconsin. SMI’s chip team includes leading scientists and industry heavyweights formerly of Qualcomm, Intel, and Sun Microsystems.

“As fast, flexible computing becomes more accessible, AI will be used by more industries for more applications more frequently, so chip design must evolve accordingly,” Sankaralingam said. “We are disrupting the next wave of computing with our breakthrough technology and are excited about the market opportunity, especially for AI chips along the power spectrum.”

Chip Details and Applications

The chip, Mozart, is currently a 16nm design that utilizes HBM2 memory and is sampling as a standard PCIe card. It is manufactured by Taiwan Semiconductor Manufacturing Company (TSMC). The chip’s software interface includes direct support for TensorFlow as well as APIs for C/C++ and Python. Future versions of the chip will leverage its unique architecture to scale up and down the thermal power spectrum, from enterprise-class high-performance systems to 5-watt IoT devices.

Karu Sankaralingam told Architosh that the 16nm design helps cut costs at this stage of the company and product but future versions of the chip will move to more advanced and smaller processes, like the 7nm version due end of 2021.

Mozart’s architecture leverages the concept of Composable Computing, which abstracts any software application into a small number of defined behaviors. SMI’s novel compiler integrates into the backend of standard AI frameworks like TensorFlow to translate those programs and reconfigures the hardware on the fly to result in a chip that behaves as if it were originally designed for that application.

SMI is initially targeting companies in the public datacenter, network security, finance, and insurance industries for its Mozart Platform with plans to disrupt the edge and mobile device markets in future product generations.  According to Allied Market Research, the global AI chip market will reach $91 billion by 2025, with growth rates of 45% a year until then. Market drivers include a surge in demand for smart homes and smart cities, more investment in AI startups, and the rise of smart robots.

To learn more, visit them online here.

Architosh Analysis and Commentary

We had a chance to speak with Karu Sankaralingam just before this unveiling. The conversation was fascinating and his new company is offering an exciting direction for AI/ML chip design and application optimization. A very good AI chip roundup article over at Forbes notes that SimpleMachines is angling for the “inference market” in AI. It is not the only AI chip company taking this direction, but its first chip, Mozart (and at 16nm) at 35 8-bit TOPS (which is a standardized performance metric) places it near the top of the heap. Of course, there is debate about TOPS as a proxy for AI chip application performance, with the same pitfalls that Intel presented with higher MHz back in the 1990s. There are multiple different and more meaningful benchmarks to test SimpleMachines’s new Mozart chip. That is for another day and article. More to come! 

About SimpleMachines, Inc.

Founded in 2017, SimpleMachines, Inc. (SMI) is an AI-focused semiconductor startup that is disrupting the next wave of computing. The company’s flagship technology, Mozart Platform, is a first-of-its-kind, easily programmable high-performance silicon chip that can run on various classes of AI algorithms. SMI’s first-generation chip – available via a PCIe card called Accelerando, or Symphony Cloud Service – is designed to provide high-performance execution of software being developed for AI, machine learning, and big data analytics.  The product roadmap includes a second and third-generation chip design that will target the large and rapidly growing mobile and edge computing markets. For more information, please visit https://www.simplemachines.ai/.

The post SimpleMachines, Inc. debuts Mozart AI Chip—Disrupts AI Chip Landscape appeared first on Architosh.

ODA Apple Silicon Support

The leading provider of CAD and BIM interoperability solutions has announced the release of Apple Silicon support for its entire full suite of software development kits (SDKs).

ODA’s new Apple Silicon support allows ODA components to be incorporated into applications that run on the new Apple M1 processor, which brings stunningly new performance improvements to Apple’s latest Macs.

Apple’s new M1 Macs absolutely crush the Intel competition.

“ODA has a significant user base on Mac, particularly in the area of architectural design,” says Neil Peterson, ODA President. “Our new Silicon SDKs allow ODA members to easily port their existing Intel-based Mac applications over to Apple Silicon, to take advantage of the superior performance offered by the M1 chip.”

MORE: Apple Introduces Revolutionary New Macs with M1 Chip

“ODA has a long history of Mac support, dating back to the PowerPC in the late 1990s. We’re looking forward to a smooth transition to this latest generation of Apple processors.”

Availability and Pricing

Members can download ODA SDKs for Apple Silicon from opendesign.com, starting with version 21.11. Apple Silicon support is included as part of the standard ODA membership offering at no additional cost to members.

Architosh Analysis and Commentary

We have suspected that the ODA’s libraries are on the cusp of being ready since learning that Graebert’s ARES Commander 2022, available now in beta, ships with an Apple Universal binary installer. “Universal” Mac applications run on either Intel-based or Apple-Silicon Macs. We now expect that updates to several well-known CAD/BIM applications for Mac will soon be able to ship “universal” installers. 

The ODA’s platform consists of SDKs that enable data interoperability and software development for engineering CAD and 3D software developers. 

Did you enjoy this?

If you enjoyed this article and content like this, subscribe to our monthly email newsletter for FREE. You may also enjoy our INSIDER Reports series.

The post The ODA Releases Support for Apple Silicon Macs appeared first on Architosh.