Rumored pro chip

[cmaier]

why do feel they only need 20% for this iteration though...? Also I’m assuming 20% is what they were able to achieve year to year with the knowledge that they would only be used in low power iPads and iPhone. Maybe they thought differently this time knowing they needed to also put them in the 16 inch MPB, iMacs and Mac Pros...?

They only need 20% single core improvement because they are already screaming - they destroy anything else at the same TDP. And they can turn up the TDP in the higher-end machines and still be lower than what they are with Intel parts. So it’s 20+that. Within a couple of years, they are blowing away anything anybody else has at *any* TDP.

 

[theorist9]

Imagine if this monster of an SoC powering a desktop in an iMac form factor. I think it’ll sell well for Apple’s customer base that need computing and graphics power without the need for expandability. My bank account will hate me if this comes to pass ?

Quite unlikely for this to go into a mobile form factor.

The rumor (which is from Bloomberg's Mark Gurman) is that that a 128-core GPU will be the upper-end graphics config for a Mac Pro. Alas, it's not expected to be seen in the iMac (which probably won't have the thermals to handle it).

Apple Readies MacBook Pro, MacBook Air Revamps

Apple Inc. is preparing to release several new Mac laptops and desktops with faster processors, new designs and improved connectivity to external devices, accelerating the company’s effort to replace Intel Corp. chips and leapfrog rival PC makers.

www.bloomberg.com

"Codenamed Jade 2C-Die and Jade 4C-Die, a redesigned Mac Pro is planned to come in 20 or 40 computing core variations, made up of 16 high-performance or 32 high-performance cores and four or eight high-efficiency cores. The chips would also include either 64 core or 128 core options for graphics."

 

[Rickroller]

Of course, higher memory bandwidth definitely helps improve thruput. For small computational workload that fits in L2 cache, the benefits will likely be less, which is what most benchmark measures. Workload that chews thru big datasets will definitely benefit.

Monstrous bandwidth definitely helps GPU performance tho.

The GPU angle is a good one. Even if they only used 8 GPU cores from the M1 without modifications or improvement, they’d get to 80,000 for a metal score with 32 on the Jade C chip. Aiming for 100,000 I would think...better than the Radeon Pro VII.

 

[quarkysg]

The rumor (which is from Bloomberg's Mark Gurman) is that that a 128-core GPU will be the upper-end graphics config for a Mac Pro. Alas, it's not expected to be seen in the iMac (which probably won't have the thermals to handle it).

Well, the iMac Pro could handle the thermals of a Xeon CPU plus an AMD dGPU but stays silent. It’s not out of the realm of possibility.

With the recent announcement of Apple Silicon Macs all using the same M1 SoC across different form factors, I think Apple likely will go with 3 tiers of compute performances across different form factors, catering to different market segments. This is probably Apple’s grand plan all along. ?

 

[theorist9]

Well, the iMac Pro could handle the thermals of a Xeon CPU plus an AMD dGPU but stays silent. It’s not out of the realm of possibility.

With the recent announcement of Apple Silicon Macs all using the same M1 SoC across different form factors, I think Apple likely will go with 3 tiers of compute performances across different form factors, catering to different market segments. This is probably Apple’s grand plan all along. ?

That's a good point about the thermals of the iMac Pro, and yes, if they made another iMac Pro*, it probably could handle some serious Apple Silicon. But we were discussing the new large iMac, which I think Apple will want to make thin (thinner than the 27") and consumer-priced, both of which will limit its computing power compared to what will be seen in the Mac Pro. So my prediction is there will not be a 128-core GPU in the next-generation of iMacs.

I think there will be either three or four lines of MacOS chips:

Four lines:

• low-end mobile/desktop (what we see now with the M1)
• high-end mobile (for the MBP)
• high-end consumer desktop (upper-end iMac)
• pro (with ECC and GPU's specialized for rendering)

Three lines: same as above, but combining the high-end mobile with high-end consumer desktop:

• low-end mobile/desktop (what we see now with the M1)
• high-end mobile (for the MBP) + desktop (upper-end iMac)
• pro (with ECC and GPU's specialized for rendering)

*The conventional wisdom, which could be wrong, is that they won't make another iMac Pro, since that was offered as a stopgap to keep pros who needed something better than the trashcan from fleeing to Windows while waiting for Apple to produce a new Mac Pro.

 

[cmaier]

The rumor (which is from Bloomberg's Mark Gurman) is that that a 128-core GPU will be the upper-end graphics config for a Mac Pro. Alas, it's not expected to be seen in the iMac (which probably won't have the thermals to handle it).

Apple Readies MacBook Pro, MacBook Air Revamps

Apple Inc. is preparing to release several new Mac laptops and desktops with faster processors, new designs and improved connectivity to external devices, accelerating the company’s effort to replace Intel Corp. chips and leapfrog rival PC makers.

www.bloomberg.com

"Codenamed Jade 2C-Die and Jade 4C-Die, a redesigned Mac Pro is planned to come in 20 or 40 computing core variations, made up of 16 high-performance or 32 high-performance cores and four or eight high-efficiency cores. The chips would also include either 64 core or 128 core options for graphics."

Agree that thermals in iMac probably couldn’t handle 128 GPU cores, at least not if each consumes the same power as on M1. But 64 should work.

 

[Rickroller]

They only need 20% single core improvement because they are already screaming - they destroy anything else at the same TDP. And they can turn up the TDP in the higher-end machines and still be lower than what they are with Intel parts. So it’s 20+that. Within a couple of years, they are blowing away anything anybody else has at *any* TDP.

I’m in the ‘want’ camp rather than the ‘need’. 20% Would be good/great, but at the same time it would be expected. Not by everyone, but a lot of the people who help create the news cycle Apple will get from the new stuff will be meh about it and the headlines would read as ‘Dog bites man’ instead of the more newsworthy reordered version.

As you said, ‘within a couple of years’ could be turned to a couple of months.

 

[cmaier]

I’m in the ‘want’ camp rather than the ‘need’. 20% Would be good/great, but at the same time it would be expected. Not by everyone, but a lot of the people who help create the news cycle Apple will get from the new stuff will be meh about it and the headlines would read as ‘Dog bites man’ instead of the more newsworthy reordered version.

As you said, ‘within a couple of years’ could be turned to a couple of months.

Some people are hard to impress. But apple doing in 26W what intel does in 80W is pretty impressive. And it’s likely that the M2’s will beat anything intel or AMD have at 100W (Still at 26W), and beat most anything in single core regardless of TDP. Then M3 is where you end up with a 26W M3 beating a 120W Xeon+++ (whatever the hell Intel is calling them then) in single core, and beating anything burning 2-digits of power in multi core.

 

[CWallace]

In terms of Performance CPU and GPU cores, I could see the iMac Pro possibly maxing out at 16/64, but I tend to think the Mac Pro will be the one at 16/64 and 32/128 and the iMac Pro will be 8/16 and 8/32 - which will still put it comfortably above the (24") iMac at 4/8.

 

[cmaier]

In terms of Performance CPU and GPU cores, I could see the iMac Pro possibly maxing out at 16/64, but I tend to think the Mac Pro will be the one at 16/64 and 32/128 and the iMac Pro will be 8/16 and 8/32 - which will still put it comfortably above the (24") iMac at 4/8.

I wonder if there will even be an iMac Pro. They only did it in the first place because the trash can Mac Pro dead-ended and they needed to fill the gap until the new cheese-grater arrived.

 

[EntropyQ3]

In terms of Performance CPU and GPU cores, I could see the iMac Pro possibly maxing out at 16/64, but I tend to think the Mac Pro will be the one at 16/64 and 32/128 and the iMac Pro will be 8/16 and 8/32 - which will still put it comfortably above the (24") iMac at 4/8.

I fear that they will do with a larger iMac what they did to the smaller - design it as a large iPad Air with severely restricted thermals. If they keep the same design language, I would be surprised if they allowed the SoC and memory to dissipate much more than 50W as even that level is difficult to get rid of silently in thin enclosures as demonstrated by Apples intel based laptops.
If so, 32 GPU units would seem to represent an upper end of the scale as 8 of them dissipate roughly 10W, and power scales pretty much linearly with units if your utilization is high - everything else being equal.

We’ll see. Maybe it’s OK if a desktop computer is an inch thick if it’s going to be a huge slab of glass anyway. Maybe not.

 

[cmaier]

I fear that they will do with a larger iMac what they did to the smaller - design it as a large iPad Air with severely restricted thermals. If they keep the same design language, I would be surprised if they allowed the SoC and memory to dissipate much more than 50W as even that level is difficult to get rid of silently in thin enclosures as demonstrated by Apples intel based laptops.
If so, 32 GPU units would seem to represent an upper end of the scale as 8 of them dissipate roughly 10W, and power scales pretty much linearly with units if your utilization is high - everything else being equal.

We’ll see. Maybe it’s OK if a desktop computer is an inch thick if it’s going to be a huge slab of glass anyway. Maybe not.

keep in mind that these would be fabbed on the new rev of the 5nm process which should reduce power a tiny bit, and will likely be a new microarchitecture as well. Even so, 32-64 is likely the upper bound in that form factor, at least for now, and assuming they don’t cheese-grater it like the XDR monitor.

 

[leman]

You’re double counting a bit. TSMC update is what enables the frequency update, etc.

But, yeah, I’m expecting a pretty big bump. The way I figure it is 20% from normal year-over-year improvements, plus the ability to increase TDP for these bigger form factors.

The thing is, we still don't know the real limit of Firestorm. Apple seems to do something different here, as they don't bin the chips by max stable frequency, only by power (presumably). It is entirely possible that already Firestorm is able to run at 3.5-3.7ghz and power consumption of 10 watt per core instead of 5 watts per core. So a prosumer Apple Silicon might end up bringing a bigger boost than what would be expected from the process improvement alone.

But yeah, I think that 20% in single core is a reasonable estimate. That would firmly put Apple on top of any current or expected CPU (overclocked models notwithstanding) until end of 2022 at least.

 

[Rickroller]

The thing is, we still don't know the real limit of Firestorm. Apple seems to do something different here, as they don't bin the chips by max stable frequency, only by power (presumably). It is entirely possible that already Firestorm is able to run at 3.5-3.7ghz and power consumption of 10 watt per core instead of 5 watts per core. So a prosumer Apple Silicon might end up bringing a bigger boost than what would be expected from the process improvement alone.

But yeah, I think that 20% in single core is a reasonable estimate. That would firmly put Apple on top of any current or expected CPU (overclocked models notwithstanding) until end of 2022 at least.

I think with Apple there is always the Rumsfeld factor, and especially with this next set of chips.

The known knowns...the known unknown...and the unknown unknowns.

Frankly the only thing we do know is that there will be a new chip.

 

[Rickroller]

I think with Apple there is always the Rumsfeld factor, and especially with this next set of chips.

The known knowns...the known unknown...and the unknown unknowns.

Frankly the only thing we do know is that there will be a new chip.

Also the reason we’re stuck on the 20% number, is because that’s what Apple has been giving us. I imagine if they had been giving us 25% we would be expecting that number instead.

 

[cmaier]

Also the reason we’re stuck on the 20% number, is because that’s what Apple has been giving us. I imagine if they had been giving us 25% we would be expecting that number instead.

Well, yes, but that’s a good thing. Relying on data from past examples is how engineers make predictions.

 

[EntropyQ3]

Å

keep in mind that these would be fabbed on the new rev of the 5nm process which should reduce power a tiny bit, and will likely be a new microarchitecture as well. Even so, 32-64 is likely the upper bound in that form factor, at least for now, and assuming they don’t cheese-grater it like the XDR monitor.

I hope for a cheese grate. Thermals aside, I kinda think it looks cool. ?

And I do think silent cooling is a major plus for a computer that is positioned a couple of feet in front of your face, so I hope they strike a good balance.

 

[sublunar]

I wonder if there will even be an iMac Pro. They only did it in the first place because the trash can Mac Pro dead-ended and they needed to fill the gap until the new cheese-grater arrived.

I think for marketing purposes if they can produce something that benches in the ball park of the iMac Pro rather than regular 27" iMacs they will be within their rights to call a 30" iMac 5.5k with miniLED panel an iMac Pro.

Obviously connectivity will be important - if Apple can match the ports on the existing iMac Pro for instance - unless they just want to brand a 30" iMac as a regular iMac like any other. An iMac Pro would come with 32Gb RAM and 1Tb SSD as base spec for instance if Apple want to start the range cheaper with lower RAM and storage by default they would just call it an iMac.

 

[CWallace]

I wonder if there will even be an iMac Pro. They only did it in the first place because the trash can Mac Pro dead-ended and they needed to fill the gap until the new cheese-grater arrived.

The main reason I believe it may be called the "iMac Pro" is for marketing - we have an iPhone and iPhone Pro, an iPad and iPad Pro, a MacBook and MacBook Pro so kind of makes sense that with the launch of the new iMac, the next model would be an iMac Pro. I mean heck, there are rumors we could have a Mac mini and Mac mini Pro.

I believe Apple is starting to drift back to the "consumer / professional" matrix like when Steve returned, though with more gradation (via products) within each matrix.

I fear that they will do with a larger iMac what they did to the smaller - design it as a large iPad Air with severely restricted thermals.

Well a fair number of folks want it to look like a 32" iPad Pro so if it does come with a chin to help with the cooling and speakers/connectors, expect to read much howling in this forum.

As cmaier noted, the M1 is ice cold compared to Intel and AMD chips so I believe Apple can increase the TDP a fair bit and still keep it operating at full-tilt with no issues. And as TSMC improves the 5nm process though 2022, those chips will run cooler still. And then they will move to 3nm in 2023/2024 and use even less power and generate even less heat.

 

[theorist9]

The thing is, we still don't know the real limit of Firestorm. Apple seems to do something different here, as they don't bin the chips by max stable frequency, only by power (presumably).

Could you explain this? The only binning I've heard of is to separate based on 7 vs 8 functional gpu cores.

 

[Joelist]

Something to remember here about Apple Silicon - at present it does roughly twice the amount of work per clock cycle as anything else out there. As a result it can run at much lower frequencies and still be "faster" in that it is processing more instructions in the same time frame. So they do not necessarily need to dial it up to high frequencies.

For those into CPU history, we saw this exact same thing happen when Intel debuted the Core 2 microarchitecture and had CPUs at low frequencies outperforming ones at much higher frequencies (it got to the point where an off the shelf air cooled Core 2 wiped the floor with the absolute top of the line AMD gaming CPU with liquid cooling).

 

[thingstoponder]

Well, you could look at it this way.

If Apple is confident enough to reduce the HE cores from 4 to 2, and assuming that 4 HE cores are actually required for the M1, then the new HE cores would be at least 100% more performant compared to the M1 HE cores to take the impact of a 50% reduction in cores to at least maintain the same performance level.

That’s assuming the cores are maxed out all the time when doing work.

You have to remember this chip was designed for the lowest common denominator, the iPad Pro. The rumored 10 core chip is the first chip designed only for the Mac. It’s possible that if the M1 was designed only for the Mac it would have 2 efficiency cores.

Either way I’m sure Apple knows what they’re doing with this whole silicon design thing. I suspect there will be no issues with 2 cores.

 

[cmaier]

That’s assuming the cores are maxed out all the time when doing work.

You have to remember this chip was designed for the lowest common denominator, the iPad Pro. The rumored 10 core chip is the first chip designed only for the Mac. It’s possible that if the M1 was designed only for the Mac it would have 2 efficiency cores.

Either way I’m sure Apple knows what they’re doing with this whole silicon design thing. I suspect there will be no issues with 2 cores.

I don’t think it’s fair to say the M1 was designed for the lowest common denominator. I suspect it was designed for the Macs and just happens to work fine in the ipad. I don’t think they said “gee, would be nice to put this on the chip for macs, but since the iPad Pro can’t make use of it never mind.”

 

[thingstoponder]

I am still expecting the next SOC to be based on M1 but with more cores and cache and RAM and additional controllers - so yes obviously the wafer will be larger even on 5nm. My thoughts are:

- 4 more Performance Cores
- 16 more GPU cores
- All caches doubled
- RAM doubled at all levels and a faster type of RAM too

We already have the info from Gurman. 8 performance cores, 2 efficiency cores, 16 or 32 GPU cores.

Gurman seemed to say these are two different chips but everyone seems to think it’s one chip that’s binned. I think it’s two chips. I think the 14” gets 16 GPU cores abd 16” gets 32.

I think it’s definitely based off the M1/a14 generation also. People who think they’re going to debut a new generation of architectural cores with this chip just have no clue what they’re talking about. That chip would be at least 9-12 months away, not right on the horizon for a possible WWDC unveiling.

 

[leman]

Could you explain this? The only binning I've heard of is to separate based on 7 vs 8 functional gpu cores.

I’m not an expert, only read about this stuff here and there but apparently a popular way to bun things in mobile is by power rather than by performance (which is what we know from desktop). Basically, as I understand it, one defines a performance limit and kind of tries to match it, even if it means that some chips will end up using more power and some end up using less power. The end effect of this is that some phones have slightly better battery life and some slightly worse, but it’s not something a user is likely to notice. Besides, mobile chips tend to have more complex power delivery (because efficiency matters more), and can dynamically tweak the needs of individual components better. So if you have a slightly inferior chip whose GPU needs more power to operate, yiu can usually supply that power without much problem.

For M1 the idea is that Apple takes chips where the GPU needs too much power to operate and disables one core to bring the power consumption to the expected numbers. I think this makes more sense than the defective core theory because M1 GPU cores are very small it’s it’s quite unlikely that a defect hits a GPU core and not something else.

Something to remember here about Apple Silicon - at present it does roughly twice the amount of work per clock cycle as anything else out there. As a result it can run at much lower frequencies and still be "faster" in that it is processing more instructions in the same time frame. So they do not necessarily need to dial it up to high frequencies.

Yes, but we still need to be a bit careful here, because this advantage does not apply to every situation. There are still pipeline delays and data dependencies, and M1 low frequency can become a measurable disadvantage when running code with low ILP.