TSMC also seems to be taking a big gamble, pissing of existing good clients.
Edit: (1) https://semianalysis.com/apple-cpu-gains-grind-to-a-halt-and...
TSMC also seems to be taking a big gamble, pissing of existing good clients.
Edit: (1) https://semianalysis.com/apple-cpu-gains-grind-to-a-halt-and...
"somehow delay their competitors from launching new and better chips"
Both AMD and nvidia have designed and built on older processes long after newer process capacity was available to them. They aren't being delayed at all.
Apple committing to production capacity is neither "Apple first", nor is it Apple "delaying" competitors. There is finite production capacity, and customers spread across the available options. There is literally nothing notable or interesting in that.
But once you add words like "Mutiny", you pander to a very gullible crowd.
NVidia and AMD do not have access to the latest processes. There is not enough capacity left at TSMC because of the Apple orders getting priority, so they have to stay on older processes. NVidia Ampere was not designed for Samsung 8nm, it was designed for TSMC 7nm but they couldn't get by with the leftovers so they had to put the consumer line on Samsung 8nm.
Apple is absolutely first. They are using their warchest to ensure their competitors do not have access to high quality processes.
Reality: Apple had designs ready and committed to production. Indeed, the first 5nm production was Apple chips, and HiSilicon. Yes, HiSilicon with stock ARM cores somehow managed what AMD and nvidia (who are barely even Apple "competitors") couldn't.
"They are using their warchest to ensure their competitors do not have access to high quality processes."
Again, nvidia and AMD both tend to hang two processes behind (making their chips cheaper, higher yield, and because they have a slow design iteration for new processes), but let's pretend that they were really trying for the newest process (you know, the one that HiSilicon managed to get). Apple's evil here is apparently simply making products. Because you know that Apple isn't just reserving unused capacity -- they are literally making A15s and M1s and soon 5G models in the hundreds of millions. It's pretty amazing that simply making your products is now nefariously blocking competitors (if you're ignorant enough of actual fact to think that was actually happening).
And AMD/Nvidia are direct competitors to the entire M1 line.
By the way, AMD die size production only for the next-gen consoles in less than a year is equivalent to ~100 million iPhones as they use 4x more silicon, and that's in supply constrained conditions. AMD and NVidia absolutely can move more wafers than Apple, they just got outbid by a company that can buy them both.
The issue is that we're in a market where Apple can afford to pay suppliers so much that other companies literally cannot afford to compete. Their positions in the market is slowing down technological advancement.
Don't manufacture arguments and then soundly beat them down. They didn't prioritize the new process so they ended up at the back of the line. They made a choice (to virtually no market detriment, and improving their financials in the process). As mentioned, both companies often run on older processes because power simply didn't matter, and wafer yield just mattered more. nvidia often has monster chips because...eh. If your GPU or giant AMD processor uses 300W, eh, that's life.
This may come as a shock to some, but despite all of the nonsensical rhetoric posted on HN (by Apple naysayers who have to dismiss anything Apple does with One-Simple-Trick nonsense that has zero association with reality -- I recall when it was the magical "big cache", as if Apple was sneaky having big caches, cheating the system), process improvements are not as big as they are held on here. They are of course a benefit, but a process improvement yields power or performance or density benefits (yup, even that last one is conditional on other factors), but the fiction spinners declare that no really it's everything all at once. It doesn't work like that. So they opted not to prioritize it.
Yet read your other posts and not only would AMD and nvidia be world's better than Apple in every dimension (see above about the Apple naysayers and their ignorance), but also they couldn't afford to because I guess the $1500 GPUs and $5000 CPUs just can't afford the big bucks Apple can spend making a smartphone CPU (where apparently they're outbidding everyone on every component, yet also simultaneously having by far the highest profit margin in the industry...my normal brain cannot even comprehend the lengths of how ludicrous and contradictory this nonsense is). Oh, and HiSilicon making chips for discount, very low end smartphones also apparently has more money to blow on this.
"The issue is that we're in a market where Apple can afford to pay suppliers so much that other companies literally cannot afford to compete"
So the bill of materials for Apple products must be enormous, right? Oh wait, they're absolutely rolling in profit, with some of the highest profit margins in the business. No, it isn't that whatsoever. Absolutely nothing indicates that Apple used its "warchest", or that it is paying a penny more than anyone else. Can you point out a single authoritative source claiming that? Because actual economics say no, that's utter nonsense.
But sure, Apple is "slowing down technological advancement" in the same post where you declare that they're paying more for that technological advancement. The desperate lengths this rhetoric has to go, with laughable self-contradiction, is embarrassing.
nvidia is currently selling very low power control boards for robotics, computer vision, drones, cars, etc, built on 12nm. There simply was no compelling reason for them to have a higher BoM to go smaller, and they had proven existing designs. There is a world of excess 10nm fab capacity, and you can sign up for 7nm all day long. Nope. Because what random blowhards say on HN has little correlation with fact.
??
> [...] somehow delay their competitors from launching new and better chips while Apple scrambles to find new talents in the limited pool available.
Nonsense, Apple buying capacity because they actually need to manufacture that much stuff. Anything else would be completely idiotic.
> Apple has bet their future on Apple Silicon, and it badly needs to retain a lead on other competing CPU for a year or two, to capture sufficient market share (the perception of Apple Silicon leading over other CPUs is very important for this).
Share of what market are they supposedly wanting to capture? Share of PCs in general? They care about share of premium devices, where they are very well represented.
Share of Macs themselves? That'll solve itself, since Apple won't be offering non-Apple Silicon Macs at some close point in the future.
> TSMC also seems to be taking a big gamble, pissing of existing good clients.
They take the offer with the most money attached, no one's (neither AMD nor Qualcom) surprised by that.
Taking the offer with the most money attached is not a good idea in the long term. They're currently providing unprecedented leverage to Apple.
Apple Silicon needs the performance crown as the comment you're replying to mentioned. Having a good market share in the premium sphere is not good enough - if instead of trading blows but still ending up ahead of AMD, AMD processors also were on 5nm with DDR5 memory they'd be plainly slower. Then Apple would be in the position where their laptops are slower and have worse battery life than the competition and also Docker doesn't work right for developers and some of your VSTs don't work at all, and few people would take the pain of changing architecture for worse performance.
Their SoCs are pretty good. But they're only good because Apple has almost exclusive access to TSMC 5nm.
I'd love to hear about how the M1 isn't competing against AMD and NVidia.
You’re making claims about hypothetical AMD / Nvidia performance at different nodes which are impossible to verify.
And Apple is clearly taking a huge risk by investing massively in the latest node. At the minimum they deserve credit for that.
There is very little speed differential between desktop and mobile single-core performance nowadays. The M1 Max is nowhere even close to the best desktop systems which would be the 3990X or the upcoming 5990WX. Despite having much faster RAM - which is the biggest reason why it can perform so well in FP workloads - in integer workloads it cannot compete. I'm citing Anandtech on their data, not on their opinion.
I'll give them credit for investing a lot in 5nm.
In any case, if someone said that the M1 was unquestionably faster than anything else on Earth it was not me. It does not change the point that asserting that better is somehow worse is still stupid.
> I'd love to hear about how the M1 isn't competing against AMD and NVidia.
To whom are they selling their GPUs?
The point is not that it is not faster. A 20% performance advantage on average given 3x faster memory and an up to 50% more efficient process means that the architecture is slower on average but that it makes it up in (third-party) memory and process.
>To whom are they selling their GPUs?
A lot of professionals that would otherwise buy high-end AMD or NVidia GPUs for now, and eventually they will probably be used for virtualization.
Some of the conclusions you're presenting in this thread are correct. But you're doing a terrible job of justifying them despite the available data, and making unnecessary exaggerations. This discussion deserves a bit more rigor.
Actually looking closer at the data, the reported package power is often even higher than wall minus idle, which means it's almost certainly inaccurate.
If you want to respond to someone who specifically referred to package power reported on Apple's chip, or if you want to make comparisons against package power and TDP reported on an AMD chip, why do you choose to respond with the wall power measurement? Subtracting out idle power doesn't remove all the potential sources of error from measuring at the wall, and in particular it cannot remove the error introduced by inconsistently including all the inefficiencies of converting from wall power to the low voltage DC the chip actually runs on. You seem to be disingenuously cherry-picking by going with the 62W wall power rather than the corresponding 44W package power measurement that was published a few pixels above it.
Yes, using wall power masks the efficiency of the charger. That said, the charger is >94% efficient : https://www.chargerlab.com/apple-87w-usb-c-power-adapter-a17.... VRM efficiency at the middle-power range is generally >95%.
On the other hand, using wall minus idle also has some advantages - you don't take into account the baseline power consumption of the memory controller, the baseline power consumption of the GPU, or the PCI/storage controllers, or the baseline power consumption of the USB controllers, and so on and so forth - that's easily 2-4W even on an M1, on a Ryzen chip this is included in package power, whereas it's not for the M1 Max.
Because of that I judged the two numbers to be the closest to each other, and the minimal power consumption losses are more than offset by excluding a fair amount of power consumption that is reflected in the package for an AMD chip (and indeed any chip).
I didn't explain and justify all of this because I didn't have the time to do so and I don't think it was necessary to clutter up the text this much. I'm not cherry-picking either. There is a reason why wall-power is included for Intel and M1 reviews but very rarely for AMD, including in AnandTech's reporting, and it's obvious from reading the Apple data on idle power usage for the M1 Mac Mini, from everything that the M1 chip does on-package and from the 7.2W idle, from which only the screen, fans and NAND chips (not controller) are outside the package, that this is not a package power reading, but an estimate of what on the PC World would be called "Core+SOC", which is typically significantly lower than actual package power. If there was a Core+SOC number provided for the 5980HS I'd compare that but I don't have one and I didn't find any. In either case in the real world the two measurements are going to be at most 1-2W apart and I'm not sure if it advantages the M1 Max or the 5980HS.
So now you're casting aspersions, too? The AnandTech article you've referred to numerous times explains exactly why wall power measurements were not included for the AMD chip, and the reason is not what you're trying to insinuate.
As for your point about charger efficiency: there's more than one step of voltage conversion to get from AC wall power to the CPU die. The losses at each step add up.
> We believe Apple had to delay the next generation CPU core due to all the personnel turnover Apple has been experiencing ...
https://semianalysis.com/apple-cpu-gains-grind-to-a-halt-and...
> Chief Architect for all Apple CPU and SOC development. For CPU, lead the Cyclone, Typhoon, Twister, Hurricane, Monsoon, Vortex, Lightning and Firestorm architecture work. Chief Architect for Apple MAC hardware platform M1 Pro and M1 Max.
I guess they were in the works long before he left.