If they are serious about gaming that would be a necessary move. At the current pace of Apple Arcade releases hardly any real AAA games wil be released.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Qualcomm Looking to Combat Apple Silicon With New Generation of PC Chips
- Thread starter MacRumors
- Start date
- Sort by reaction score
Apple has been working on ARM since the 1980s. The tech has been in mobile devices for 3 decades. It is not the new hotness at all. Anyone who says stuff like "the ancient x86 architecture" elicits a hearty guffaw from me (why not say "the ancient UNIX standard"? I know exactly why not).
Different architectures with different goals and use cases. ARM for peak efficiency, x86 for wholesale compatibility. Neither is going away (well one is on the Mac side, they don't have choices), so pick your need and accept your sacrifice.
Different architectures with different goals and use cases. ARM for peak efficiency, x86 for wholesale compatibility. Neither is going away (well one is on the Mac side, they don't have choices), so pick your need and accept your sacrifice.
I'm not sure Apple's chips are actually that much better than Qualcomm's. It's not a fair comparison because one is optimized for fewer than 10 models, one for thousands. I know that there's a similar situation in the macOS/Windows space but for some reason I still think it's not as bad and that Qualcomm could still get really close. The more competition the more progress and innovation so stories like this one make me happy.
depends on the competitor..
because we had badaOS as mobile competitor and did nothing...
this in pc world...probably will do nothing as well
Apple has an advantage that Qualcomm doesn’t: they ARE the computer manufacturer and OS developer so they don’t need to convince other companies to design and write apps/drivers around their ARM processor.
Doesn’t mean that Qualcomm will fail but it does mean that there are at least short term disadvantages that Qualcomm will need to overcome. And that leaves out getting an OS written and debugged. If Microsoft decides that they want in this market, without helping Apple by providing a non Apple compatible Windows ARM OS then the playing field could be more level. Could be, not automatically will be.
Doesn’t mean that Qualcomm will fail but it does mean that there are at least short term disadvantages that Qualcomm will need to overcome. And that leaves out getting an OS written and debugged. If Microsoft decides that they want in this market, without helping Apple by providing a non Apple compatible Windows ARM OS then the playing field could be more level. Could be, not automatically will be.
Last edited:
Computer gaming isn’t a huge market. With all of the consoles it is hard to get someone to spend a lot of money on a gaming PC. You face the “good enough “ problem that Apple ran into with the original HomePod. Yes the sound was a lot better than Alexa/Google. But the cost (and honestly, people’s perception of Siri) made the “ok” sound of Alexa preferable. And same with game consoles: you can’t get the frame rate or the shading and detail that a game machine can but you can get “pretty good” and there are more games actually available sooner. Especially now when games can be exclusive for a particular console for a while.
I see it mostly taking off for Chromebooks... a fast portal machine for students and residents in developing countries.
The article never said "Windows" only PCs, and the PC segment includes Macs, Linux, Chromebooks, Windows, etc.
If even Windows never runs on this chip (and it still might), there is an unexpected fragmentation of Windows marketshare that I never expected to see 20 years ago back when it was 93% Windows, 6% MacOS and 1% everything else.
It would be interesting to see them buy Unity, keep it cross platform, waive the license fees, and integrate it into their toolchain. This isn't quite the same as becoming their own content creator, but it would lower the effort necessary to port good content to Mac.
And have the added benefit of giving Epic a poke in the eye.
No, they won’t be even close. Apple‘s advantage in efficiency is due to the chip‘s design for the most part. So even on equal nodes Apple would win by 3-4x for sure.
Second - your assessment is based on the assumption that Apple stops to innovate and doesn‘t improve their process any further, what of course will be the case
i wonder what the optics would be if apple buys a studio that uses unreal for game development instead of their lackluster development frameworks.
Apple big speed enhancement is from using SoC so others are going to do the same. So Intel, Qualcomm, and other will make SoC too and they will get the same speed up that SoC architecture provides . Apple users are used to being told they can't update their hardware so for non-Apple user SoC will get speed but lose upgradable systems.
The future Mac Pro will be the key box to check out. The hardcore Mac Pro user need large banks of RAM and that's where SoC starts becoming an issue. Apple in one article said they are experimenting to a interconnect external memory at speeds close to SoC. What will be interest is to see how long Apple stays in this high end market. It is expensive in both time and money for a small customer base, that is something Apple has dabbled in, in the past and then backed away from. Apple is all about high profits margin in large markets, not high cost and small market.
The future Mac Pro will be the key box to check out. The hardcore Mac Pro user need large banks of RAM and that's where SoC starts becoming an issue. Apple in one article said they are experimenting to a interconnect external memory at speeds close to SoC. What will be interest is to see how long Apple stays in this high end market. It is expensive in both time and money for a small customer base, that is something Apple has dabbled in, in the past and then backed away from. Apple is all about high profits margin in large markets, not high cost and small market.
In the server market sure, there's a demand for low-power high performance SoC's... Outside of Apple, that's the only other market where those types of ARM designs have gained any traction... but I don't see it making any kind of dent in the WIntel market, when you already have extremely powerful discrete GPU's and CPU's... where there's already a sustainable volume of sales on a entrenched platform that EVERYONE seems afraid to abandon (and very few people seem to care about power usage).
Yes, of course there's a premium market outside of Apple... But there's a reason Apple has been the only company to get to this point so far... They have the volume of "premium" sales that's unmatched by any other company. Without that scale the cost per unit goes through the roof. Apple's R&D is basically 2 new CPU cores a year (p-core and e-core). Those are eventually used across the board, in hundreds of millions of high margin devices, from the watch up to the MacBook Pros, and soon enough in the Mac Pro. Same goes for their GPU cores, NPU cores, ISP cores, etc... Not to mention the absolute brilliant design of the M1 Max, where the same SoC can be halved, doubled and quadrupled (supposedly), saving a massive amount in fabrication costs by baking in redundancy so that defects can be closed off and rerouted to another processing unit. Here, Apple has only designed 2 SoC's for the Mac... The M1 and M1X (Max)... That will scale from entry level all the way to an extremely high-end Pro.
"3-4x"? It's nowhere near that even now with AMD's and Intel's node disadvantage.
It's not Apple that achieves process improvements, it's TSMC. At the moment Apple happen to be the first to use TSMC's 5nm process for a computer CPU, but that won't last.
Not really. Performance per watt is the measurement Apple targets. Boasting about 80 cores being more performant than a 28-core Xeon is not really a remarkable accomplishment when an M1 with 20 cores will more than likely do the same @ 60W - both of those others hit 240W and higher.
The PC motherboard market has so much fragmentation, that this is going to be a challenge for Intel to switch too.
Not only does Intel sell and manufacture the CPUs, but all of the motherboard chipsets, memory controllers, even going down to drivers for USB... often as separate hardware. Sure the chipset might support 4 USB ports but Motherboard X is better because it provides an additional 6 by using an additional USB controller.
I don't know if Apple built their own chipsets, used the default Intel ones or had Intel manufacture custom ones...
But all of that has completely changed now for M1+ Macs. There are tons of things previously controlled by specialized chipsets that are now being managed out of the new M1 SoC. A few things like wireless, SSD controllers, and some sound stuff related to the speakers might be handled by other chips; I don't know.
Apple is first because they pay for it which helps TSMC make the improvements at a faster pace. Do you think there is some other company that can outbid Apple? It is the guarantee of all those iPhone A series chips that allows TSMC to make the investment.
Microsoft is trying to divorce Intel as well and looking at making their own chips for a more seamless user experience like Apple has created. When they do that, all apps can work across their devices and platforms. Intel is buggy and have to patch another vulnerability. They are also very slow to innovate. Intel is done.
Nvidia Jetson AGX Orin looks very interesting.
Jetson AGX Orin specifications:
Jetson AGX Orin specifications:
- CPU – 12-core Arm Cortex-A78AE v8.2 64-bit processor with 3MB L2 + 6MB L3 cache
- GPU / AI accelerators
- NVIDIA Ampere architecture with 2048 NVIDIA CUDA cores and 64 Tensor Cores @ 1 GHz
- DL Accelerator – 2x NVDLA v2.0
- Vision Accelerator – PVA v2.0 (Programmable Vision Accelerator)
- AI Performance – 200 TOPS (INT8) @ 50W
- Video Encode – 2x 4K60 | 4x 4K30 | 8x 1080p60 | 16x 1080p30 (H.265)
- Video Decode – 1x 8K30 | 3x 4K60 | 6x 4K30 | 12x 1080p60| 24x 1080p30 (H.265)
- System Memory – 32GB 256-bit LPDDR5 @ 204.8 GB/s
- Storage – 64GB eMMC 5.1 flash
- 699-pin Molex Mirror Mezz connector with
- Storage – Single lane UFS
- Display – 1x 8K60 multi-mode DP 1.4a (+MST), eDP 1.4a, HDMI 2.1
- Camera
- Up to 6x CSI Camera (16 via virtual channels)
- 16 lanes MIPI CSI-2
- D-PHY 1.2 (up to 40Gbps) | C-PHY 1.1 (up to 164Gbps)
- Networking – 1x GbE, 4x 10GbE
- PCIe – 2x PCIe x8 (or 1x PCIe x8 + 2x PCIe x4), 1x PCIe x4, 2x PCIe x1 (PCIe Gen4, Root Port & Endpoint)
- USB – 3x USB 3.2, 4x USB 2.0
- Low speed IOs – 4x UART, 3x SPI, 4x I2S, 8x I2C, 2x CAN, DMIC & DSPK, GPIOs
- Power Modes – 15W, 30W, or 50W
- Dimensions – 100 x 87mm
It's a lot easier said than done. It requires one of the largest tech companies in the world to focus the majority of it's resources on this to pull off the "seamless user experience." Side projects that are not part of the core competency, like VOD, rarely achieve the same level of success. MSFT is 100% focused on Azure. If you need any more proof of this look at the recently leaked software engineering salaries across all of their different orgs. Their Azure comps are in a different universe, compared to Windows and Office orgs, as it's responsible for driving MSFT to its $2.5T valuation. The Surface HW products are a side hobby for them as is Pixel for Google or Numbers & Keynote for Apple. Surface will challenge MacBooks the day Numbers and Keynote challenges Excel and Powerpoint.
Last edited: