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.
Apple Silicon A14X Bionic specs and results
- Thread starter Xenobius
- Start date
- Sort by reaction score
Something similar has already happened before. The first Intel MacBook Pro was faster than Power Mac G5.
Last edited:
I believe that’s what Apple is aiming for.
The CPU performance advantage is a given. The only questions remaining are GPU performance and RAM.
In terms of GPU, performance gains over integrated GPUs from intel are guaranteed. The comparison with dedicated GPUs even at laptop level is tougher (hence the surprise with MBP 16” rumors)
That's the point.
Remember that Tim, Jhony, Craig and the rest of the senior management team didn't come up with this idea over a couple pitchers of beer at BJs in January. Apple has been working on this for years.
Without a doubt, Apple has been running Mac-intended Apple Silicon prototypes in their labs for years. "Good enough" isn't. An alternative needs to be head-and-shoulders better than what their current supplier provides.
My guess is that Apple's in-house prototype silicon matched Intel in performance-per-watt several years ago; there was likely a similar performance growth in discrete GPU performance as well.
But it's not just peak performance that Apple is chasing. Per Jhony's segment in the WWDC keynote, the primary focus is performance-per-watt. He pounded this concept again and again.
We know that the Darwin kernel can support BIG.little (performance cores + efficiency cores). Windows 10 currently does not support BIG.little to my knowledge so there's little point in AMD and Intel in producing BIG.little CPUs since the primary OS can't utilize it.
By utilizing efficiency cores -- particularly in notebook computers -- Apple can achieve unrivaled battery performance with Apple Silicon compared to Intel/AMD powered hardware.
Much of Apple Silicon's potential superiority is in software.
We already see that in iPhones and iPads: Apple's homegrown ARM silicon is vastly superior to Samsung and Qualcomm ARM silicon. This is particularly apparent in the tablet market. This is one of the limiters for Windows on ARM: Microsoft does not have access to top-tier ARM silicon.
The advantage is that many of Apple developers are already quite familiar with BIG.little from iOS/iPadOS. Much of this is on Apple. The operating system (Big Sur and future iterations) should try to determine much of this so developers don't need to figure out how many performance cores are available versus how many efficiency cores are available on any given system (entry-level MacBook Air versus a maxxed-out MacBook Pro).
Maybe the desktop systems have a different ratio of performance-efficiency cores since there's no battery performance constraint.
Intel and AMD have no response to this approach. I have a custom built desktop PC and I'm running a 65W TDP CPU. It performs wonderfully when I max everything out. It's a big power hog when it's mostly idle.
I expect that Apple Silicon Mac performance will improve with updated chips but more importantly from future optimized macOS operating systems and better APIs for developers.
Last edited:
Same here though I've never used 100% CPU (i7-10700) - I wanted something with a lot of excess power that I will never use that runs cool. I really like this system but I could be convinced to go with a Mac Pro if the price isn't too nuts and it uses significantly less power running my normal workload.
I'm running macOS in my Windows system right now because most of my work runs better on Windows (one runs on WINE in macOS and I'm guessing it won't run well on macOS/AS). But I could get a weak AS machine and just VNC into it from my Windows system so that I have both on my desktop. It works fine with my 2015 MacBook Pro 15.
It would be really nice to only use macOS, even nicer with very low power draw. But I think that's asking a lot for the first round.
Last edited:
Handbrake encodes max out my 8-core AMD Ryzen 3700X. Same with my i7 Mac mini 2018.
My custom built Windows PC is quieter because I'm using an over-specced 240mm AIO cooler on the 65W TDP CPU. By contrast, the Mac mini 2018 is at the outer edge of the thermal envelope that Apple's mechanical engineers designed and the smaller, weaker Mac fan is louder.
Like you, I prioritized great cooling acoustics on my Windows PC build. Both the CPU and GPU have AIO coolers. Not including the PSU fan and the GPU MOSFET fan, there are five fans in my current PC case (micro-ATX). Whatever fan noise there is, it's inconsequential during gameplay or a full CPU load (the latter is far less common).
Both the Ryzen CPU and the RTX 2070 GPU top out at 65°C; they normally idle at 34°C and 29°C respectively.
My Windows custom build is superior to my Mac mini 2018 in terms of power and acoustics at the expense of size, power, and style.
The other big difference is the operating system.
My Windows PC is primarily a gaming system. I can play Control with Nvidia RTX hardware ray tracing and DLSS. I can't do that on a Mac, even with Bootcamp and the best Radeon RX 5700XT GPU.
My custom built Windows PC is quieter because I'm using an over-specced 240mm AIO cooler on the 65W TDP CPU. By contrast, the Mac mini 2018 is at the outer edge of the thermal envelope that Apple's mechanical engineers designed and the smaller, weaker Mac fan is louder.
Like you, I prioritized great cooling acoustics on my Windows PC build. Both the CPU and GPU have AIO coolers. Not including the PSU fan and the GPU MOSFET fan, there are five fans in my current PC case (micro-ATX). Whatever fan noise there is, it's inconsequential during gameplay or a full CPU load (the latter is far less common).
Both the Ryzen CPU and the RTX 2070 GPU top out at 65°C; they normally idle at 34°C and 29°C respectively.
My Windows custom build is superior to my Mac mini 2018 in terms of power and acoustics at the expense of size, power, and style.
The other big difference is the operating system.
My Windows PC is primarily a gaming system. I can play Control with Nvidia RTX hardware ray tracing and DLSS. I can't do that on a Mac, even with Bootcamp and the best Radeon RX 5700XT GPU.
Last edited:
My system is a trading rig and normal CPU load with my trading software is about 15-20% of max. It has a huge Arctic air cooler and three case fans. GPU is 75 watts - I just needed something to be able to drive 3 4k monitors. Normal CPU operating temps are around 30 degrees.
I have a friend that runs a similar trading setup on a 2018 Mac Mini and he reports that it's quite hot to the touch. That's my issue with MacBook Pros and iMacs. I have 2014 and 2015 MacBook Pros and those spin up the fans kind of loud when I run my trading software. If Apple can do MBP Intel 13 performance with great thermals, then people are going to buy them just to do Zoom conferences without the fan spinning up.
I haven't done any transcoding on this thing yet. I may have the opportunity to do it at some point but I'd probably throw it in a VM limited to two cores. I'm already thinking of my second build as it is nice to have a backup system.
We've discussed this before elsewhere.
Fidelity Active Trader Pro on the Mac is a Windows executable running in a customized version of CrossOver. It isn't optimized for macOS.
I have no idea why Zoom sucks so badly on Macs other than lack of priority by Zoom. It runs fairly well on a different crap $180 Windows PC. It seems to run adequately on my iDevices.
Fidelity ATP on Windows runs natively. I run a one window ATP instance on a crap $180 Windows PC (not my gaming system) with a wimpy Intel Celeron CPU.
I don't even bother with running ATP on my gaming rig. The GPU could easily drive three monitors and I'm sure it would run great.
Fidelity Active Trader Pro on the Mac is a Windows executable running in a customized version of CrossOver. It isn't optimized for macOS.
I have no idea why Zoom sucks so badly on Macs other than lack of priority by Zoom. It runs fairly well on a different crap $180 Windows PC. It seems to run adequately on my iDevices.
Fidelity ATP on Windows runs natively. I run a one window ATP instance on a crap $180 Windows PC (not my gaming system) with a wimpy Intel Celeron CPU.
I don't even bother with running ATP on my gaming rig. The GPU could easily drive three monitors and I'm sure it would run great.
ATP didn't work at all on Catalina. I think that Fidelity tried to rebuild it for 64-bit Windows but it took some time to get it running. ATP is like that - it can break on you if you upgrade something or run on a newer version of macOS which has issues with the version of WINE that they ship.
Zoom can do a lot more on PCs than on mobile devices.
I have not heard of Zoom complaints on Windows systems but most of the people I know that do Zoom use Macs.
I would guess that I'm doing a lot more on ATP than you are.
I run about 90 RealTime charts on ToS and about 12 on ATP but the ones on ATP have a lot of indicators. Running ATP on my 2014 MacBook Pro 15 was enough to spin the fans up. A couple of trading buddies use one or two Mac Pros to run Think or Swim. You can do quite a bit on these trading systems. ToS supports automated trading but I don't think that ATP does - or it may require an add-on.
You're right, while Fidelity would label me as an "active trader" I'm not pounding on ATP.
I can easily go a week without a transaction. These days I'm swing trading four equities. There are a handful of limit orders open at any given time but most of the time I'm just observing.
Tomorrow should be interesting. Nasdaq futures are +2.15%. Asia is already awake, Nikkei 225 is +1.95%.
Will Fidelity ATP ever run on ASi? Fidelity may need to run their own cloud services and stream this to client systems.
I can easily go a week without a transaction. These days I'm swing trading four equities. There are a handful of limit orders open at any given time but most of the time I'm just observing.
Tomorrow should be interesting. Nasdaq futures are +2.15%. Asia is already awake, Nikkei 225 is +1.95%.
Will Fidelity ATP ever run on ASi? Fidelity may need to run their own cloud services and stream this to client systems.
I received a Greenlight survey from Fidelity. The questions were mainly about Active Trader Pro and one of them asked me how I would feel if they discontinued or replaced it.
Schwab is laying off a bunch of employees after buying out TD Ameritrade. So they have two Pro Trading platforms now. It does not make financial sense to keep two of these things in development and I worry that they will toss ToS. I could always learn the other one but I'd rather not have to.
If Fidelity drops their Pro platform, then I might have to look at third-party solutions.
These brokerages need to move these pro trading platforms into the cloud. The transactions and data come from their servers anyhow.
If videogamers can play AAA PC games via GeForce Now/Google Stadia/xCloud/whatever, I don't see any argument why Fidelity ATP can't be run in the cloud and why ATP on the iPad can't be a thin cloud client.
That would make the ASi executable quandary irrelevant. The web browser ends up being the operating system.
If Fidelity drops their Pro platform, then I might have to look at other brokerages. I'm not married to Fidelity.
I've already moved from a full-service brokerage, I can move again. I've moved from Apple Motorola > PowerPC > Intel. I'll move to ASi if it makes sense but it appears that I have a few options today.
If videogamers can play AAA PC games via GeForce Now/Google Stadia/xCloud/whatever, I don't see any argument why Fidelity ATP can't be run in the cloud and why ATP on the iPad can't be a thin cloud client.
That would make the ASi executable quandary irrelevant. The web browser ends up being the operating system.
If Fidelity drops their Pro platform, then I might have to look at other brokerages. I'm not married to Fidelity.
I've already moved from a full-service brokerage, I can move again. I've moved from Apple Motorola > PowerPC > Intel. I'll move to ASi if it makes sense but it appears that I have a few options today.
The amount of data for trading platforms, though, is tiny. Much less than sending graphics data over the wire.
I've used third-part trading platforms in the past. I used to use QuoteTracker and then Medved Trader which worked with well over 100 different data sources from brokerages around the world. I also have an account at Etrade which I could use for quotes but I don't want to learn yet another platform.
C'mon, the newest Xbox and PS5 gaming consoles are doing 4K/120Hz.
UI response times on active trader consumer platforms is kiddie play compared to what gamers expect.
Fidelity ATP data only refreshes once a second.
UI response times on active trader consumer platforms is kiddie play compared to what gamers expect.
Fidelity ATP data only refreshes once a second.
I imagine that any data provider wants to minimize bandwidth. These platforms probably are quite similar to what they were doing in the 1990s.
One of the nice things about using a third-party trading platforms is that you can switch your data feed if there's a problem with what you're using. I could switch QuoteTracker between Etrade, Ameritrade and Fidelity for quote sources.
The data is all there. It's just that brokerage firms only refresh at limited/slow rates.
Videogames back in Nineties were probably running at 640x480p at 30Hz (max) back then. Today, it's 4K/60Hz (or more) with ray tracing and more.
I doubt any trader need technical indicator graphs refreshing at that rate but it's lame that the desktop software can't display this efficiently in 2020.
It's just a framebuffer that needs to be refreshed once a second.
Videogames back in Nineties were probably running at 640x480p at 30Hz (max) back then. Today, it's 4K/60Hz (or more) with ray tracing and more.
I doubt any trader need technical indicator graphs refreshing at that rate but it's lame that the desktop software can't display this efficiently in 2020.
It's just a framebuffer that needs to be refreshed once a second.
My daytrading is off of five-minute charts. That is my indicators update every five minutes. Prices update every second or two. I've seen futures traders that trade on shorter time-frames; mainly because you can get killed in a minute or two in futures trading. Most traders I know don't trade 1-minute charts.
So a cloud solution would use far more bandwidth than the vast majority of traders need.
You guys are hilarious, the A14X results that have leaked are the next iPad Pro and entry level Macbook/Air SOC with the 4 LP and 4HP core config. The 13" MacBook Pro will employ a variant with the 12 core configuration. The 16" we have no leak to go off yet but if I was to guess it will likely be a 16 core configuration with 8 LP and 8 HP cores. To clarify though that "12 core" SOC will appear in the 13" MacBook Pro.
I'm okay if other people here think I'm not paying attention.
I don't view these leaked "benchmarks" as legitimate. They might end up being accurate but I see no justifiable reason why I should consider these particular rumors to be definitive.
If there are 10 rumors published on this site, 9 are wrong/useless.
That's why this is MacRumors, not MacPressRelease or MacBenchmark.
The Windows PC sites have plenty of garbage rumors too. It's not Mac specific.
I don't view these leaked "benchmarks" as legitimate. They might end up being accurate but I see no justifiable reason why I should consider these particular rumors to be definitive.
If there are 10 rumors published on this site, 9 are wrong/useless.
That's why this is MacRumors, not MacPressRelease or MacBenchmark.
The Windows PC sites have plenty of garbage rumors too. It's not Mac specific.
Why would they add low power cores to the 16” chip? As I understand it they are intended for background processing of light tasks, and there’s only so much of that work to go around on any running system.
To save on power. It's not like Apple Mail checking your IMAP server needs the big core.
I'd say that much consumer PC activity can be handled by efficiency cores. We've already seen that on iDevices.
The activity that most consumer computer users do is web browsing but that's pretty much single-threaded. You'd be better off running 3-5 efficiency cores and one performance core when web surfing at most sites.
I'd say that much consumer PC activity can be handled by efficiency cores. We've already seen that on iDevices.
The activity that most consumer computer users do is web browsing but that's pretty much single-threaded. You'd be better off running 3-5 efficiency cores and one performance core when web surfing at most sites.
The "leaked" A14X benchmark, may be a synthetic one - we don't know if it came from a real machine, a prototype or a pre-production mode, nor whether it was for an iPad or a MacBook of some type.
We also have no factual data that tells us the MBP 13 will have a 12-core CPU. I've read that it could have the same 4+4 (LP/HP) cores as the MacBook Air, but a better GPU. We simply don't know.
So far *everything* that we have read are rumors...so we'll just have to wait and see. Not long now!
The leaked benchmark is not reliable. It's just a rumor number.
Geekbench is a joke measurement anyhow.
Geekbench is a joke measurement anyhow.