Apple has already stated for the record that they are going to use their own GPU. Why is this written as speculation?
They have also stated that they are designing Mac specific SoCs. So no, it won’t be an A14X. Though it might use the same core design but the number of CPU and GPU cores are going to be Mac specific.
Because they haven't said that at least in the context of discreet GPU's. Some information was deciphered to suggest that that was the case but it can be seen that the information can be interpreted in more than one way. The discussion is a bit ridiculous anyways as every A series Apple SoC comes with a GPU, so all Macs using an A series chip will have an Apple GPU even if a discreet is included. So the question becomes does Apple use a discreet Apple GPU in addition to its integrated GPU in the initial Macs and frankly we can't say. There are still good reasons to stick with AMD at the high end, mainly because of performance, but even here Apple has options to mix AMD with Apple GPU's.
With respect to the A14X in a Mac Book revival; that is easy to understand and has no impact on the idea that Mac specific SoC's are coming. The very nature of Macs will require very different chips for the various classes of machines. Personally if Apple doesn't implement SMT they will need a 24 core laptop machine by the end of 2021 to remain competitive. On the Mac Pro they are screwed if they can't match what AMD is already offering which means offering at least 64 cores and 128 cores will be needed if they can't get SMT working. AMD has some of the most compelling workstation offerings on the market right now so a Mac Pro is going to really stretch Apples capabilities.
Your assertion that Apple will need 24 cores in the main CPU in 2021 (without SMT, perhaps 12 with) in a laptop to remain competitive in 2021 is laughable. In a Mac Pro it can make sense to need that many main CPU cores, because they’re going to be expensive non-portable machines used for dedicated server/workstation tasks that can reasonably make use of all those cores, but if and only if it isn’t constrained by running off of low power or a battery. Very few applications these days come close to using 4 cores in any meaningful sense short of the special types of applications you run on dedicated workstations and servers: most of the time most of the 24 (or even 8 cores now with Intel) will remain idle. If they weren’t idle most of the time, with that many cores going full-tilt, you’d be lucky to get more than an hour or so of battery life on a 16” MacBook Pro: it’s not rational to expect much different regardless of CPU architecture.
Let’s say Apple paid any attention to know-nothings claiming they needed that many cores in a laptop to be competitive: Apple would be pissing away power efficiency to even have all those cores exist even with them not running code most of the time. After all these years, it seems you’ve pointedly ignored Apple’s practical strategy in iOS devices of fewer but faster cores in comparison to Android devices. Sure, Android device makers could claim they had more cores, but that had no practical value beyond advertising copy. Anyone that has enough software development experience knows this.
Will Intel be selling any 12 core with SMT mobile-targeted chips? Perhaps they’d get sales from enthusiasts and those that think having that many cores will improve their game performance, but even that is dubious. It’s really hard to justify that due to what I mentioned above. The way it’d make sense to have that many cores is using the BIG/little strategy purely for power efficiency with (usually) only one half (all same type) being active, but not likely to use both types at the same time. Even then, Windows and typical applications you use in a laptop very rarely, all combined, will saturate that many cores.
I was watching Twit, the tech blog. And it was said by one of thev guests, and the others agreed, that most people don’t need more than 4 cores, and that 8 was the most almost anyone did need. I agree. Apple got away with just two cores for years, and very successfully. I’m not even counting the Efficiency cores, because all of those put together just barely equal one Performance core.
I can see 8 cores, maybe even the twelve we’ve been hearing about. But more than that for most computers is a laughable waste of real estate and power. Those extra cores will likely never be used properly.
I assume everyone here knows how to use Apple’s Activity Monitor? Open it up and strip out the core indicator. Then run software and do a lot of different things while that’s on the screen. You’ll notice that even when a lot of cores are being used, they’re just blips most of the time. Multitasking, encryption, unpacking, all of them just use cores for a fraction of the time, and usually only at a low usage. Read the CPU numbers for “user”. You’ll see what I mean. My Mac has 8 hardware cores, and 8 virtual cores. It’s amazing how rare it is for software to take advantage of it. Video rendering is where it matters most. even Fusion 360 doesn’t use all the cores most of the time.
my old Mac Pro from 2012 had two CPUs, each with 12 hardware cores and 12 virtual cores. It was even worse then. What mattered most was how fast an individual core is.
the problem is that we see synthetic tests that show multi core use, and give numbers. But that’s not the way the computer works normally.
That’s the typical usage model in most user environments.
Spinning up a lot of threads that you keep around and not actively use is just horribly inefficient, and most applications are waiting on the user more than 99% of their wall time. If browsers are using up the CPU cores with a large number of threads, that’s likely because of ads more than anything else, over a lot of pages opened at the same time. The more cores that are running different code and processing different data, the less efficient the entire system gets for effective CPU throughput and energy efficiency as a result of blowing away CPU caches and waiting for main memory accesses. Again, very wasteful to do it that way. I’ve found it makes sense to do most of my web references via iPad and use Xcode without Safari open to maximize battery life when not plugged in. All having more cores beyond a certain modest point for a laptop will do for you is eat the battery alive when they’re being used, and still waste power by merely existing when they aren’t.
Apple has already stated for the record that they are going to use their own GPU. Why is this written as speculation?
They have also stated that they are designing Mac specific SoCs. So no, it won’t be an A14X. Though it might use the same core design but the number of CPU and GPU cores are going to be Mac specific.
Because they haven't said that at least in the context of discreet GPU's. Some information was deciphered to suggest that that was the case but it can be seen that the information can be interpreted in more than one way. The discussion is a bit ridiculous anyways as every A series Apple SoC comes with a GPU, so all Macs using an A series chip will have an Apple GPU even if a discreet is included. So the question becomes does Apple use a discreet Apple GPU in addition to its integrated GPU in the initial Macs and frankly we can't say. There are still good reasons to stick with AMD at the high end, mainly because of performance, but even here Apple has options to mix AMD with Apple GPU's.
With respect to the A14X in a Mac Book revival; that is easy to understand and has no impact on the idea that Mac specific SoC's are coming. The very nature of Macs will require very different chips for the various classes of machines. Personally if Apple doesn't implement SMT they will need a 24 core laptop machine by the end of 2021 to remain competitive. On the Mac Pro they are screwed if they can't match what AMD is already offering which means offering at least 64 cores and 128 cores will be needed if they can't get SMT working. AMD has some of the most compelling workstation offerings on the market right now so a Mac Pro is going to really stretch Apples capabilities.
Your assertion that Apple will need 24 cores in the main CPU in 2021 (without SMT, perhaps 12 with) in a laptop to remain competitive in 2021 is laughable. In a Mac Pro it can make sense to need that many main CPU cores, because they’re going to be expensive non-portable machines used for dedicated server/workstation tasks that can reasonably make use of all those cores, but if and only if it isn’t constrained by running off of low power or a battery. Very few applications these days come close to using 4 cores in any meaningful sense short of the special types of applications you run on dedicated workstations and servers: most of the time most of the 24 (or even 8 cores now with Intel) will remain idle. If they weren’t idle most of the time, with that many cores going full-tilt, you’d be lucky to get more than an hour or so of battery life on a 16” MacBook Pro: it’s not rational to expect much different regardless of CPU architecture.
Let’s say Apple paid any attention to know-nothings claiming they needed that many cores in a laptop to be competitive: Apple would be pissing away power efficiency to even have all those cores exist even with them not running code most of the time. After all these years, it seems you’ve pointedly ignored Apple’s practical strategy in iOS devices of fewer but faster cores in comparison to Android devices. Sure, Android device makers could claim they had more cores, but that had no practical value beyond advertising copy. Anyone that has enough software development experience knows this.
Will Intel be selling any 12 core with SMT mobile-targeted chips? Perhaps they’d get sales from enthusiasts and those that think having that many cores will improve their game performance, but even that is dubious. It’s really hard to justify that due to what I mentioned above. The way it’d make sense to have that many cores is using the BIG/little strategy purely for power efficiency with (usually) only one half (all same type) being active, but not likely to use both types at the same time. Even then, Windows and typical applications you use in a laptop very rarely, all combined, will saturate that many cores.
I was watching Twit, the tech blog. And it was said by one of thev guests, and the others agreed, that most people don’t need more than 4 cores, and that 8 was the most almost anyone did need. I agree. Apple got away with just two cores for years, and very successfully. I’m not even counting the Efficiency cores, because all of those put together just barely equal one Performance core.
I can see 8 cores, maybe even the twelve we’ve been hearing about. But more than that for most computers is a laughable waste of real estate and power. Those extra cores will likely never be used properly.
I assume everyone here knows how to use Apple’s Activity Monitor? Open it up and strip out the core indicator. Then run software and do a lot of different things while that’s on the screen. You’ll notice that even when a lot of cores are being used, they’re just blips most of the time. Multitasking, encryption, unpacking, all of them just use cores for a fraction of the time, and usually only at a low usage. Read the CPU numbers for “user”. You’ll see what I mean. My Mac has 8 hardware cores, and 8 virtual cores. It’s amazing how rare it is for software to take advantage of it. Video rendering is where it matters most. even Fusion 360 doesn’t use all the cores most of the time.
my old Mac Pro from 2012 had two CPUs, each with 12 hardware cores and 12 virtual cores. It was even worse then. What mattered most was how fast an individual core is.
the problem is that we see synthetic tests that show multi core use, and give numbers. But that’s not the way the computer works normally.
It's what AMD is doing to be very obnoxious, basically, "forget about everything and lay cores on cores." While winning the performance war, it's not a good choice especially for professional workstations. Memory bottlenecked, stability sucks, little to no overclock potential, and most people don't even buy a 16-core to begin with.
To give some examples, Threadrippers have all 8 clusters to share two memory controllers: They also get unstable when you maxed out all PCIe lanes where Xeon just moves on. I believe LTT made a video about it. TR also stuck at 512GiB of RAM until recent release of the Pro variant. That was late by a year and more.
Say Intel sucks all they want, it's not like they're going to experience firmware issues or maxing all slots while laying down heavy workloads.
... Personally if Apple doesn't implement SMT they will need a 24 core laptop machine by the end of 2021 to remain competitive.
My opinion is that the choice will be thumbs down to SMT. Apple has avoided that up to this point. I'm guessing that the reason is heat generation - SMT concentrates it, more cores spread the heat. Some chips have been built with 54 billion features - the A12X only uses 10 billion. Only! There would seem to be lots of expansion space available. 24 cores sounds really fast!
I think the heat comment from SMT is a very astute observation. For Apple Silicon, I think Apple will make whatever decision that improves perf/Watt. SMT yields not much performance benefit for Apple's workloads, and it keeps high power consumption cores active. On the other hand, the efficiency cores likely yield about the same performance benefit as SMT, but uses much less power. Not sure if really true. If so, heterogenous CPU cores it is. And they have already said Apple Silicon will be a heterogenous CPU architecture. So moot point.
I bet the A14, for the phone, will have about 12b transistors. The A14X for the iPad Pros, and probably low end Macs, will have 15b transistors. The Mac silicon could be 20b to 30b transistors!
I'm betting on 32 performance cores as the top end for pro desktops. Who knows about the efficiency cores. The CPUs will only take about 20% of the transistors. The vast majority of the transistors will be for GPU and dedicated hardware units (neural engine, etc).
Well, I don't think modern benchmarks doesn't know how to execute SMT, claiming that ASi was no match for x86 SMT is ridiculous:
... Personally if Apple doesn't implement SMT they will need a 24 core laptop machine by the end of 2021 to remain competitive.
My opinion is that the choice will be thumbs down to SMT. Apple has avoided that up to this point. I'm guessing that the reason is heat generation - SMT concentrates it, more cores spread the heat. Some chips have been built with 54 billion features - the A12X only uses 10 billion. Only! There would seem to be lots of expansion space available. 24 cores sounds really fast!
I think the heat comment from SMT is a very astute observation. For Apple Silicon, I think Apple will make whatever decision that improves perf/Watt. SMT yields not much performance benefit for Apple's workloads, and it keeps high power consumption cores active. On the other hand, the efficiency cores likely yield about the same performance benefit as SMT, but uses much less power. Not sure if really true. If so, heterogenous CPU cores it is. And they have already said Apple Silicon will be a heterogenous CPU architecture. So moot point.
I bet the A14, for the phone, will have about 12b transistors. The A14X for the iPad Pros, and probably low end Macs, will have 15b transistors. The Mac silicon could be 20b to 30b transistors!
I'm betting on 32 performance cores as the top end for pro desktops. Who knows about the efficiency cores. The CPUs will only take about 20% of the transistors. The vast majority of the transistors will be for GPU and dedicated hardware units (neural engine, etc).
Well, I don't think modern benchmarks doesn't know how to execute SMT, claiming that ASi was no match for x86 SMT is ridiculous:
Yes. I agree with you. SMT isn't a big win for Apple, not a win for consumers, and not a win for gamers. It's mostly a win for HPC and server VMs if the memory bandwidth is there. For the MBP16 Apple Silicon equivalent, probably ranging from 8 to 16 performance cores. No idea on the efficiency cores, They are small and Apple can pile them up and not use much transistors.
My 32 core comment was in regard to the expensive desktop top-end CPU model. I think there is some chance it can make it into a MBP16 like laptop, or the Mac mini, though. It would have to be downclocked to hit a theoretical 60 to 70 Watt TDP.
I do contemplate segmentation and how much value that Apple would attach to additional CPU cores. Eg, the difference between 6, 8, 10 cores - the current segmentation in the iMac 5K - really isn't that much when increasing core counts by 2, and the value of 2 only decreases with more cores. With the current Mac Pro, why even bother with the 12 and 24 core models? Just use 8, 16, 28, or if the 14 core model is there, 8, 14, 28.
So, I've been wondering if Apple would segment with factors of 2 in numbers of cores, like 2, 4, 8, 16, 32. CPU cores will be really cheap at TSMC N5, unless Apple pulls a rabbit out of their and doubles CPU core sizes in the A14, which they could, but even then, it still won't be that expensive in terms of die area.
... Personally if Apple doesn't implement SMT they will need a 24 core laptop machine by the end of 2021 to remain competitive.
My opinion is that the choice will be thumbs down to SMT. Apple has avoided that up to this point. I'm guessing that the reason is heat generation - SMT concentrates it, more cores spread the heat. Some chips have been built with 54 billion features - the A12X only uses 10 billion. Only! There would seem to be lots of expansion space available. 24 cores sounds really fast!
I think the heat comment from SMT is a very astute observation. For Apple Silicon, I think Apple will make whatever decision that improves perf/Watt. SMT yields not much performance benefit for Apple's workloads, and it keeps high power consumption cores active. On the other hand, the efficiency cores likely yield about the same performance benefit as SMT, but uses much less power. Not sure if really true. If so, heterogenous CPU cores it is. And they have already said Apple Silicon will be a heterogenous CPU architecture. So moot point.
I bet the A14, for the phone, will have about 12b transistors. The A14X for the iPad Pros, and probably low end Macs, will have 15b transistors. The Mac silicon could be 20b to 30b transistors!
I'm betting on 32 performance cores as the top end for pro desktops. Who knows about the efficiency cores. The CPUs will only take about 20% of the transistors. The vast majority of the transistors will be for GPU and dedicated hardware units (neural engine, etc).
Well, I don't think modern benchmarks doesn't know how to execute SMT, claiming that ASi was no match for x86 SMT is ridiculous:
Yes. I agree with you. SMT isn't a big win for Apple, not a win for consumers, and not a win for gamers. It's mostly a win for HPC and server VMs if the memory bandwidth is there. For the MBP16 Apple Silicon equivalent, probably ranging from 8 to 16 performance cores. No idea on the efficiency cores, They are small and Apple can pile them up and not use much transistors.
My 32 core comment was in regard to the expensive desktop top-end CPU model. I think there is some chance it can make it into a MBP16 like laptop, or the Mac mini, though. It would have to be downclocked to hit a theoretical 60 to 70 Watt TDP.
I do contemplate segmentation and how much value that Apple would attach to additional CPU cores. Eg, the difference between 6, 8, 10 cores - the current segmentation in the iMac 5K - really isn't that much when increasing core counts by 2, and the value of 2 only decreases with more cores. With the current Mac Pro, why even bother with the 12 and 24 core models? Just use 8, 16, 28, or if the 14 core model is there, 8, 14, 28.
So, I've been wondering if Apple would segment with factors of 2 in numbers of cores, like 2, 4, 8, 16, 32. CPU cores will be really cheap at TSMC N5, unless Apple pulls a rabbit out of their and doubles CPU core sizes in the A14, which they could, but even then, it still won't be that expensive in terms of die area.
HT won't help when one have a vastly superior architecture. So does core counts. Mind you both 9900K and 3900X require active cooling to achieve above 4.5GHz, A13 is passively cooled.
Doesn't mean the final product have to crush both of them in multicore though, that depends on the target audience. Winning the core war doesn't guarantee it will be a good product, like their 16- and 64-core.
Apple has already stated for the record that they are going to use their own GPU. Why is this written as speculation?
They have also stated that they are designing Mac specific SoCs. So no, it won’t be an A14X. Though it might use the same core design but the number of CPU and GPU cores are going to be Mac specific.
Because they haven't said that at least in the context of discreet GPU's. Some information was deciphered to suggest that that was the case but it can be seen that the information can be interpreted in more than one way. The discussion is a bit ridiculous anyways as every A series Apple SoC comes with a GPU, so all Macs using an A series chip will have an Apple GPU even if a discreet is included. So the question becomes does Apple use a discreet Apple GPU in addition to its integrated GPU in the initial Macs and frankly we can't say. There are still good reasons to stick with AMD at the high end, mainly because of performance, but even here Apple has options to mix AMD with Apple GPU's.
With respect to the A14X in a Mac Book revival; that is easy to understand and has no impact on the idea that Mac specific SoC's are coming. The very nature of Macs will require very different chips for the various classes of machines. Personally if Apple doesn't implement SMT they will need a 24 core laptop machine by the end of 2021 to remain competitive. On the Mac Pro they are screwed if they can't match what AMD is already offering which means offering at least 64 cores and 128 cores will be needed if they can't get SMT working. AMD has some of the most compelling workstation offerings on the market right now so a Mac Pro is going to really stretch Apples capabilities.
Your assertion that Apple will need 24 cores in the main CPU in 2021 (without SMT, perhaps 12 with) in a laptop to remain competitive in 2021 is laughable. In a Mac Pro it can make sense to need that many main CPU cores, because they’re going to be expensive non-portable machines used for dedicated server/workstation tasks that can reasonably make use of all those cores, but if and only if it isn’t constrained by running off of low power or a battery. Very few applications these days come close to using 4 cores in any meaningful sense short of the special types of applications you run on dedicated workstations and servers: most of the time most of the 24 (or even 8 cores now with Intel) will remain idle. If they weren’t idle most of the time, with that many cores going full-tilt, you’d be lucky to get more than an hour or so of battery life on a 16” MacBook Pro: it’s not rational to expect much different regardless of CPU architecture.
Let’s say Apple paid any attention to know-nothings claiming they needed that many cores in a laptop to be competitive: Apple would be pissing away power efficiency to even have all those cores exist even with them not running code most of the time. After all these years, it seems you’ve pointedly ignored Apple’s practical strategy in iOS devices of fewer but faster cores in comparison to Android devices. Sure, Android device makers could claim they had more cores, but that had no practical value beyond advertising copy. Anyone that has enough software development experience knows this.
Will Intel be selling any 12 core with SMT mobile-targeted chips? Perhaps they’d get sales from enthusiasts and those that think having that many cores will improve their game performance, but even that is dubious. It’s really hard to justify that due to what I mentioned above. The way it’d make sense to have that many cores is using the BIG/little strategy purely for power efficiency with (usually) only one half (all same type) being active, but not likely to use both types at the same time. Even then, Windows and typical applications you use in a laptop very rarely, all combined, will saturate that many cores.
I was watching Twit, the tech blog. And it was said by one of thev guests, and the others agreed, that most people don’t need more than 4 cores, and that 8 was the most almost anyone did need. I agree. Apple got away with just two cores for years, and very successfully. I’m not even counting the Efficiency cores, because all of those put together just barely equal one Performance core.
I can see 8 cores, maybe even the twelve we’ve been hearing about. But more than that for most computers is a laughable waste of real estate and power. Those extra cores will likely never be used properly.
I assume everyone here knows how to use Apple’s Activity Monitor? Open it up and strip out the core indicator. Then run software and do a lot of different things while that’s on the screen. You’ll notice that even when a lot of cores are being used, they’re just blips most of the time. Multitasking, encryption, unpacking, all of them just use cores for a fraction of the time, and usually only at a low usage. Read the CPU numbers for “user”. You’ll see what I mean. My Mac has 8 hardware cores, and 8 virtual cores. It’s amazing how rare it is for software to take advantage of it. Video rendering is where it matters most. even Fusion 360 doesn’t use all the cores most of the time.
my old Mac Pro from 2012 had two CPUs, each with 12 hardware cores and 12 virtual cores. It was even worse then. What mattered most was how fast an individual core is.
the problem is that we see synthetic tests that show multi core use, and give numbers. But that’s not the way the computer works normally.
It's what AMD is doing to be very obnoxious, basically, "forget about everything and lay cores on cores." While winning the performance war, it's not a good choice especially for professional workstations. Memory bottlenecked, stability sucks, little to no overclock potential, and most people don't even buy a 16-core to begin with.
To give some examples, Threadrippers have all 8 clusters to share two memory controllers: They also get unstable when you maxed out all PCIe lanes where Xeon just moves on. I believe LTT made a video about it. TR also stuck at 512GiB of RAM until recent release of the Pro variant. That was late by a year and more.
Say Intel sucks all they want, it's not like they're going to experience firmware issues or maxing all slots while laying down heavy workloads.
AMD didn't have problems with more than 256GB of RAM [NOT 512GB] as Zen has had a 2TB available RAM support from the beginning. It's the OEMs who wanted to cap TR at 256GB.
AMD required their Pro Line to go up to 2TB as they do when they released them. They released them in agreement with their OEMs to give them 18 months worth of Consumer level Threadripper sales.
Anyone today not buying the Pro Model is a fool. TR Pro requires 8 Memory Channels instead of 4 which Threadripper is limited to and what you're alluding to about memory limitations.
Most OEM AMD Vendors won't sell the Pro. In fact, do some research. I've yet to find a motherboard specifically for the Threadripper Pro line thus supporting 2TB of RAM. They're all stuck on their pre-capped 256GB.
Apple has already stated for the record that they are going to use their own GPU. Why is this written as speculation?
They have also stated that they are designing Mac specific SoCs. So no, it won’t be an A14X. Though it might use the same core design but the number of CPU and GPU cores are going to be Mac specific.
Because they haven't said that at least in the context of discreet GPU's. Some information was deciphered to suggest that that was the case but it can be seen that the information can be interpreted in more than one way. The discussion is a bit ridiculous anyways as every A series Apple SoC comes with a GPU, so all Macs using an A series chip will have an Apple GPU even if a discreet is included. So the question becomes does Apple use a discreet Apple GPU in addition to its integrated GPU in the initial Macs and frankly we can't say. There are still good reasons to stick with AMD at the high end, mainly because of performance, but even here Apple has options to mix AMD with Apple GPU's.
With respect to the A14X in a Mac Book revival; that is easy to understand and has no impact on the idea that Mac specific SoC's are coming. The very nature of Macs will require very different chips for the various classes of machines. Personally if Apple doesn't implement SMT they will need a 24 core laptop machine by the end of 2021 to remain competitive. On the Mac Pro they are screwed if they can't match what AMD is already offering which means offering at least 64 cores and 128 cores will be needed if they can't get SMT working. AMD has some of the most compelling workstation offerings on the market right now so a Mac Pro is going to really stretch Apples capabilities.
Your assertion that Apple will need 24 cores in the main CPU in 2021 (without SMT, perhaps 12 with) in a laptop to remain competitive in 2021 is laughable. In a Mac Pro it can make sense to need that many main CPU cores, because they’re going to be expensive non-portable machines used for dedicated server/workstation tasks that can reasonably make use of all those cores, but if and only if it isn’t constrained by running off of low power or a battery. Very few applications these days come close to using 4 cores in any meaningful sense short of the special types of applications you run on dedicated workstations and servers: most of the time most of the 24 (or even 8 cores now with Intel) will remain idle. If they weren’t idle most of the time, with that many cores going full-tilt, you’d be lucky to get more than an hour or so of battery life on a 16” MacBook Pro: it’s not rational to expect much different regardless of CPU architecture.
Let’s say Apple paid any attention to know-nothings claiming they needed that many cores in a laptop to be competitive: Apple would be pissing away power efficiency to even have all those cores exist even with them not running code most of the time. After all these years, it seems you’ve pointedly ignored Apple’s practical strategy in iOS devices of fewer but faster cores in comparison to Android devices. Sure, Android device makers could claim they had more cores, but that had no practical value beyond advertising copy. Anyone that has enough software development experience knows this.
Will Intel be selling any 12 core with SMT mobile-targeted chips? Perhaps they’d get sales from enthusiasts and those that think having that many cores will improve their game performance, but even that is dubious. It’s really hard to justify that due to what I mentioned above. The way it’d make sense to have that many cores is using the BIG/little strategy purely for power efficiency with (usually) only one half (all same type) being active, but not likely to use both types at the same time. Even then, Windows and typical applications you use in a laptop very rarely, all combined, will saturate that many cores.
I was watching Twit, the tech blog. And it was said by one of thev guests, and the others agreed, that most people don’t need more than 4 cores, and that 8 was the most almost anyone did need. I agree. Apple got away with just two cores for years, and very successfully. I’m not even counting the Efficiency cores, because all of those put together just barely equal one Performance core.
I can see 8 cores, maybe even the twelve we’ve been hearing about. But more than that for most computers is a laughable waste of real estate and power. Those extra cores will likely never be used properly.
I assume everyone here knows how to use Apple’s Activity Monitor? Open it up and strip out the core indicator. Then run software and do a lot of different things while that’s on the screen. You’ll notice that even when a lot of cores are being used, they’re just blips most of the time. Multitasking, encryption, unpacking, all of them just use cores for a fraction of the time, and usually only at a low usage. Read the CPU numbers for “user”. You’ll see what I mean. My Mac has 8 hardware cores, and 8 virtual cores. It’s amazing how rare it is for software to take advantage of it. Video rendering is where it matters most. even Fusion 360 doesn’t use all the cores most of the time.
my old Mac Pro from 2012 had two CPUs, each with 12 hardware cores and 12 virtual cores. It was even worse then. What mattered most was how fast an individual core is.
the problem is that we see synthetic tests that show multi core use, and give numbers. But that’s not the way the computer works normally.
It's what AMD is doing to be very obnoxious, basically, "forget about everything and lay cores on cores." While winning the performance war, it's not a good choice especially for professional workstations. Memory bottlenecked, stability sucks, little to no overclock potential, and most people don't even buy a 16-core to begin with.
To give some examples, Threadrippers have all 8 clusters to share two memory controllers: They also get unstable when you maxed out all PCIe lanes where Xeon just moves on. I believe LTT made a video about it. TR also stuck at 512GiB of RAM until recent release of the Pro variant. That was late by a year and more.
Say Intel sucks all they want, it's not like they're going to experience firmware issues or maxing all slots while laying down heavy workloads.
AMD didn't have problems with more than 256GB of RAM [NOT 512GB] as Zen has had a 2TB available RAM support from the beginning. It's the OEMs who wanted to cap TR at 256GB.
AMD required their Pro Line to go up to 2TB as they do when they released them. They released them in agreement with their OEMs to give them 18 months worth of Consumer level Threadripper sales.
Anyone today not buying the Pro Model is a fool. TR Pro requires 8 Memory Channels instead of 4 which Threadripper is limited to and what you're alluding to about memory limitations.
Most OEM AMD Vendors won't sell the Pro. In fact, do some research. I've yet to find a motherboard specifically for the Threadripper Pro line thus supporting 2TB of RAM. They're all stuck on their pre-capped 256GB.
This should bring about a 30% retail price reduction in Apple products formerly reliant on Intel C and G PU's. Though I doubt Apple will ever pass on the savings to their customers.
Nor will they competitively OEM their processors to PC manufacturers. No, a quick shot to the foot is the Apple way.
This should bring about a 30% retail price reduction in Apple products formerly reliant on Intel C and G PU's. Though I doubt Apple will ever pass on the savings to their customers.
Nor will they competitively OEM their processors to PC manufacturers. No, a quick shot to the foot is the Apple way.
That's not how Apple builds its products. They start with a price, like $1000, then they build a product customers will perceive as worth that price. Since there are similar products on the market, the product's features are an amalgam of performance and features relative to competitor products and customers' perceived value at that price point.
So, maybe you are asking if Apple will build a $800 laptop. That's doubtful. A Macbook Air with a 13" display, A14 SoC, and 8 GB of RAM would make for a great product. A really great product for that price as it will outperform most x86 ultrabooks at that price. So, they'd really price it at $1000. Maybe an A13 with 8 GB of RAM might hit $800. An A13 laptop would make for a great product too, but Apple has to want to do it.
To give you a sense of Apple's pricing with their own silicon, like at the iPad Pros. The iPad Pro 12.9 has a 13" display, 6 GB of RAM and 128 GB of storage for $1000. Don't think Apple Silicon laptops of comparable performance and features will be lower.
This should bring about a 30% retail price reduction in Apple products formerly reliant on Intel C and G PU's. Though I doubt Apple will ever pass on the savings to their customers.
Nor will they competitively OEM their processors to PC manufacturers. No, a quick shot to the foot is the Apple way.
That's not how Apple builds its products. They start with a price, like $1000, then they build a product customers will perceive as worth that price. Since there are similar products on the market, the product's features are an amalgam of performance and features relative to competitor products and customers' perceived value at that price point.
So, maybe you are asking if Apple will build a $800 laptop. That's doubtful. A Macbook Air with a 13" display, A14 SoC, and 8 GB of RAM would make for a great product. A really great product for that price as it will outperform most x86 ultrabooks at that price. So, they'd really price it at $1000. Maybe an A13 with 8 GB of RAM might hit $800. An A13 laptop would make for a great product too, but Apple has to want to do it.
To give you a sense of Apple's pricing with their own silicon, like at the iPad Pros. The iPad Pro 12.9 has a 13" display, 6 GB of RAM and 128 GB of storage for $1000. Don't think Apple Silicon laptops of comparable performance and features will be lower.
But...but...Apple has a two trillion dollar valuation, matey. This mostly due to unethical Tax avoidance and criminal evasion trickery.
I was thinking more along the lines of a sub $500.00 MacMini with user expansion options.
This should bring about a 30% retail price reduction in Apple products formerly reliant on Intel C and G PU's. Though I doubt Apple will ever pass on the savings to their customers.
Nor will they competitively OEM their processors to PC manufacturers. No, a quick shot to the foot is the Apple way.
That's not how Apple builds its products. They start with a price, like $1000, then they build a product customers will perceive as worth that price. Since there are similar products on the market, the product's features are an amalgam of performance and features relative to competitor products and customers' perceived value at that price point.
So, maybe you are asking if Apple will build a $800 laptop. That's doubtful. A Macbook Air with a 13" display, A14 SoC, and 8 GB of RAM would make for a great product. A really great product for that price as it will outperform most x86 ultrabooks at that price. So, they'd really price it at $1000. Maybe an A13 with 8 GB of RAM might hit $800. An A13 laptop would make for a great product too, but Apple has to want to do it.
To give you a sense of Apple's pricing with their own silicon, like at the iPad Pros. The iPad Pro 12.9 has a 13" display, 6 GB of RAM and 128 GB of storage for $1000. Don't think Apple Silicon laptops of comparable performance and features will be lower.
But...but...Apple has a two trillion dollar valuation, matey. This mostly due to unethical Tax avoidance and criminal evasion trickery.
I was thinking more along the lines of a sub $500.00 MacMini with user expansion options.
This is very unlikely to happen. Apple Silicon Macs will trend more and more like iPads and iPhones where RAM and storage are not user upgradeable. The only machine that will maintain modularity is the Mac Pro, which is unfortunately 10x higher than your $500 price.
Comments
Spinning up a lot of threads that you keep around and not actively use is just horribly inefficient, and most applications are waiting on the user more than 99% of their wall time. If browsers are using up the CPU cores with a large number of threads, that’s likely because of ads more than anything else, over a lot of pages opened at the same time. The more cores that are running different code and processing different data, the less efficient the entire system gets for effective CPU throughput and energy efficiency as a result of blowing away CPU caches and waiting for main memory accesses. Again, very wasteful to do it that way. I’ve found it makes sense to do most of my web references via iPad and use Xcode without Safari open to maximize battery life when not plugged in. All having more cores beyond a certain modest point for a laptop will do for you is eat the battery alive when they’re being used, and still waste power by merely existing when they aren’t.
To give some examples, Threadrippers have all 8 clusters to share two memory controllers:
They also get unstable when you maxed out all PCIe lanes where Xeon just moves on. I believe LTT made a video about it.
TR also stuck at 512GiB of RAM until recent release of the Pro variant. That was late by a year and more.
Say Intel sucks all they want, it's not like they're going to experience firmware issues or maxing all slots while laying down heavy workloads.
By the way, most people still don't have 8 cores in their PCs:
https://store.steampowered.com/hwsurvey
Well, I don't think modern benchmarks doesn't know how to execute SMT, claiming that ASi was no match for x86 SMT is ridiculous:
My 32 core comment was in regard to the expensive desktop top-end CPU model. I think there is some chance it can make it into a MBP16 like laptop, or the Mac mini, though. It would have to be downclocked to hit a theoretical 60 to 70 Watt TDP.
I do contemplate segmentation and how much value that Apple would attach to additional CPU cores. Eg, the difference between 6, 8, 10 cores - the current segmentation in the iMac 5K - really isn't that much when increasing core counts by 2, and the value of 2 only decreases with more cores. With the current Mac Pro, why even bother with the 12 and 24 core models? Just use 8, 16, 28, or if the 14 core model is there, 8, 14, 28.
So, I've been wondering if Apple would segment with factors of 2 in numbers of cores, like 2, 4, 8, 16, 32. CPU cores will be really cheap at TSMC N5, unless Apple pulls a rabbit out of their and doubles CPU core sizes in the A14, which they could, but even then, it still won't be that expensive in terms of die area.
HT won't help when one have a vastly superior architecture. So does core counts.
Mind you both 9900K and 3900X require active cooling to achieve above 4.5GHz, A13 is passively cooled.
Doesn't mean the final product have to crush both of them in multicore though, that depends on the target audience. Winning the core war doesn't guarantee it will be a good product, like their 16- and 64-core.
“no motherboard vendor have that yet”
🙄️
Nor will they competitively OEM their processors to PC manufacturers. No, a quick shot to the foot is the Apple way.
So, maybe you are asking if Apple will build a $800 laptop. That's doubtful. A Macbook Air with a 13" display, A14 SoC, and 8 GB of RAM would make for a great product. A really great product for that price as it will outperform most x86 ultrabooks at that price. So, they'd really price it at $1000. Maybe an A13 with 8 GB of RAM might hit $800. An A13 laptop would make for a great product too, but Apple has to want to do it.
To give you a sense of Apple's pricing with their own silicon, like at the iPad Pros. The iPad Pro 12.9 has a 13" display, 6 GB of RAM and 128 GB of storage for $1000. Don't think Apple Silicon laptops of comparable performance and features will be lower.
I was thinking more along the lines of a sub $500.00 MacMini with user expansion options.
A 5G, Apple Silicon MacBook 12", with 16GB and 512GB
A 5G, Apple iPhone 12 Pro Max, with 8GB and 512GB
That's all!
Here are some cookies...