mfryd
About
- Username
- mfryd
- Joined
- Visits
- 57
- Last Active
- Roles
- member
- Points
- 726
- Badges
- 1
- Posts
- 274
Reactions
-
Why Apple uses integrated memory in Apple Silicon -- and why it's both good and bad
An interesting idea. I suspect that many would complain that Apple was purposely throttling the performance of the third party socketed memory.dominikhoffmann said:The Mac Pro could be made to allow the use of PCI-based RAM as 2nd-tier memory. Then the SoC RAM would act as a rather huge cache. Alternatively, code could be programmed to perform certain less time-sensitive operations in that 2nd-tier RAM. This would then give users the opportunity to upgrade their RAM with aftermarket hardware.
In any case, that socketed memory would not provide the same level of performance as the main memory. This means it may not meet the needs of those that really need huge amounts of memory. The current M2 chips use the SSD as backing. This "virtual memory" scheme allows programs to address more memory than physically exists. Memory above and beyond what exists is stored on the SSD, and the actual memory is used a "cache" for the memory currently in use.
-
Why Apple uses integrated memory in Apple Silicon -- and why it's both good and bad
Actually, moving the memory further away from the CPU does add latency. Every foot of wire adds about a nanosecond of delay.melgross said:Ok, so the writer gets it wrong, as so many others have when it comes to the M series RAM packaging. One would think that’s this simple thing would be well understood by now. So let me make it very clear - the RAM is NOT on the chip. It is NOT “in the CPU itself”. As we should all know by now, it’s in two packages soldered to the substrate, which is the small board the the SoC is itself soldered to. The lines from Apple’s fabric, which everything on the chip is connected with, extend to that substrate, to the RAM chips. Therefore, the RAM chips are separate from the SoC, and certainly not in the CPU itself. As we also know, Apple offers several different levels of RAM for each M series they sell. That means that there is no limit to their ability to decide how much RAM they can offer, up to the number of memory lines that can be brought out. This is no different from any traditional computer. Every CPU and memory controller has a limit as to how much RAM can be used. So, it seems to me that Apple could, if it wanted to, have sockets for those RAM packages, which add no latency, and would allow exchangeable RAM packages. Apple would just have to extend the maximum number of memory lines out to the socket. How many would get used would depend on the amount of RAM in the package. That’s nothing new. That’s how it’s done. Yes, under that scheme you would have to remove a smaller RAM package when getting a larger one, but that's also normal. The iMac had limited RAM slots and we used to do that all the time. Apple could also add an extra two sockets, in addition to the RAM that comes with the machine. So possibly there would be two packages soldered to the substrate, and two more sockets for RAM expansion. Remember that Apple sometimes does something a specific way, not because that’s the way it has to be done, but because they decided that this was the way they were going to do it. We don’t know where Apple is going with this in the future. It’s possible that the M2, which is really just a bump from the M1, is something to fill in the time while we’re waiting for the M3, which with the 3nm process it’s being built on, is expected to be more than just another bump in performance. Perhaps an extended RAM capability is part of that.Then there is the issue of how many wires you run. When the memory is physically close to the CPU you can run more wires from the memory to the CPU, this allows you to get data to/from the CPU faster. It's not practical to run a large number of wires to a socket that might be a foot or more of cable run away. That means you transfer less data in each clock cycle.
Generally socketed memory is on an external bus. This lets various peripherals directly access memory. The bus arbitration also adds overhead.
Traditional CPUs try to overcome these memory bottlenecks by using multiple levels of cache. This can provide a memory bandwidth performance boost for chunks of recently accessed memory. However, tasks that use more memory than will fit in the cache, may not benefit from these techniques.
Apples "System on a Chip" design really does allow much higher memory bandwidth. Socketing the memory really would reduce performance.
-
Why Apple uses integrated memory in Apple Silicon -- and why it's both good and bad
Yes. No single solution will serve all users. Apple is not trying to sell the Mac as a solution for every possible situation. Apple is targeting the huge mainstream part of the computer market. The vast majority of computer owners never upgrade components. They stick with the original RAM/storage, and eventually just replace the whole computer.hmurchison said:I see a lot of this stuff as a stopgap solution. UMA is nice but the industry realizes that no single memory solution is going to cover all bases
...
Apple is not going after the low end of the market. Nor is Apple going after the ultra high end where users really do need terabytes of RAM, or multiple high performance GPUs.
Think of the car market. Most car owners stick with the original rims, suspension, radio, radiator, engine, etc. Even though many cars can be upgraded with performance boosting chips, fancier rims, even more powerful engines. Most car owners are not interested in these sorts of upgrades, and will keep the car as they bought it, until they replace it.
The Mac is not the best solution for everyone. However, Apple is trying to make it the best solution for most.
-
The new Apple Silicon Mac Pro badly misses the mark for most of the target market
You make a reasonable point that Apple does not offer suitable products for certain market segments.cgWerks said:
I'd question that.mfryd said:
The big issue that people are complaining about is the lack of support for external video cards. Apple's built in graphics are quite impressive, but there exist video cards out there that are faster. So the market that's excluded here is that portion that needs more than what Apple provides, but can get by with what third part cards can provide. A large part of that market is video gamers. I don't think they are generally considered to be part of the "professional" market.
Another market segment looking for the fastest GPUs are those mining for crypto currency. These people generally are buying commodity computers, and not Macs.
The bottom line is that the "pro" market is only a very small percentage of the total Mac market. Only a small percentage of that pro market needs more than 192GB of RAM and/or third party GPUs.
So while it's true that the Mac Pro is not ideal for every professional who wants a Mac, it certainly meets the needs of most professionals.
While it is true that the new Mac Pro is excellent for certain markets (video editing), I think the people who need RAM, and especially GPU-power, are a fairly high percent of that Pro market. It would be interesting to see some actual data!
I also think there is confusion here over video I/O and GPU? The video cards out there aren't just faster, they are a LOT faster. They probably can't get by with what Apple provides, in this case. It isn't just gaming. Anyone working in CAD/3D disciplines needs that GPU power. I think there is also quite a bit of scientific work that relies on GPUs.
As for crypto-mining, that I do know quite a bit about. GPU-mining is essentially over (at least for now, unless another coin that is GPU-mineable rises dramatically in price). And, what they were buying were special systems that could run a bunch of GPUs. What they are buying now are ASICs.
To be fair, if you're using those slots for things like high-speed storage expansion, or crazy-level I/O, TB isn't going to cut it. But, this brings up a point about GPUs and the rebuttal Apple keeps taking about in regards to RAM/GPU and 'optimized performance.' Sure, some off-chip RAM or a GPU in a slot or TB isn't 'optimized' but my eGPU on TB works just fine in the 3D apps and games I run on my Intel Mac. I have to think they just didn't want to go that direction, not that it wasn't possible, or even realistic.spliff monkey said:
Here's what's being ignored though. I have to throw out a new tower ever few years. A $7,000 tower EVERY 2-3 YEARS? No thanks. Anyone inclined to purchase a new Mac Pro Tower definitely need to be able to upgrade RAM and video cards on a machine that costs that much. Otherwise just buy a studio and a really nice TB chassis for your cards and call it a day for $5-6k. Hopefully your TB expansion chases will still be compatible when you replace the studio. If not you'll still save $$$
Literally the only advantage the new Mac Pro offers is PCIE in a chassis which tends be more reliable and perform better than connecting the same PCIE cards of TB. That's it. The new tower is a complete mystery if they couldn't include upgradeable ram and video cards. IF Apple didn't want to work with NVIDIA or AMD or couldn't come up with a way of upgrading RAM they shouldn't have bothered with the new tower and left it as the only intel model. The market the tower was intended for was better served with x86 for at least the next few years. This feels like a rush job.
Yes, but you're talking about one segment in the professional market, here (and then a generalization about the high-end pro market and budgets*). The problem is Apple has no solution for the other areas of the pro market, now, regardless of budget.mfryd said:
... If you are editing, color correcting, or otherwise working on major motion pictures with budgets well over $100 million, you likely fall into this category. You also fall into the category, where you expect to replace these machines every year or two with faster and better ones. $7K for a new machine is small potatoes.
Remember, these are people who are happy to pay $5k for a monitor as they are used to paying over $20K for monitors.
Now if you are a home hobbyist, and want to incrementally upgrade your machine as budget allows, then the Mac Pro may not be for you. But then, a home hobbyist is not the "professional" market.
...
They actually have a better solution for the home-hobbyist on the whole with the mini Pro or Studio, except for the lack of GPU-power. Those people weren't likely in the Mac Pro market since like early 2000s (like me, I once owned a Pro Mac... it's way out of my budget, now). I'm thrilled with the Mac Studio in theory. I never really needed a Pro besides there being no real middle in the Mac lineup much of the time.
But, in my discipline, I need reasonable GPU power. Apple has insanely increased the GPU power across the lineup (for gamers, this is good, as even an entry level Mac could potentially do some gaming now, not just a few high-end models), but at the same time, they have massively lost GPU-power on the high end. A Mac Pro with 4+ GPUs is so far ahead of anything in the current lineup, comparisons are absurd. But, even a setup like my 2018 mini w/ eGPU is faster than anything Apple currently offers. IMO, this is a fairly big problem.
* Yes, totally. I used to work for a near Fortune 50 in IT. Every year, the manager would come and say something like, "OK, we've got to spend $70k on some new equipment within the next couple of weeks. What kind of things do we need?" So, we'd go on server shopping trips, or everyone would get new laptops, etc. The equipment would be passed down or repurposed. But, if the hardware can't do the basics when new, it isn't going to get purchased in the first place. (And, we were just a small division within the bigger company. I imagine for some, it's like, 'we have to spend $200k, or $1.5M', etc.)
An important question, is what percentage of the overall computer market is being left out.
My suspicion is that the consumer market is likely at least 20 times larger than the high end pro market. I suspect that the high volume consumer market has a much higher total profit potential than the much smaller, high end professional market. If Apple lost 90% of the pro market, and gained 20% of the consumer market, I think that would represent an increase overall sales.
I may be off base, but ease of use advantages are likely more important to non-technical consumers, than high end professionals that have the support of their corporate IT department. A consumer benefits from the easy setup of Time Machine backups. A pro with an IT department, may not care, as backups are handled by the IT department.
Remember, Apple is running a business. They don't need to win every market segment. They can come out way ahead by capturing only the larger market segments. According to their marketing, Apple doesn't enter a market segment, unless they feel they can offer a much better product than the competition. The Apple Silicon based systems offer very good price/performance in the consumer market. Not so much at the high end. From a business perspective, it may not be necessary to develop a version of Apple Silicon that supports third party memory and GPUs.
Of course, I could be wrong.
-
The new Apple Silicon Mac Pro badly misses the mark for most of the target market
Obviously, these machines don't magically stop working after 2-3 years. Many people get long life from these machines.spliff monkey said:
Here's what's being ignored though. I have to throw out a new tower ever few years. A $7,000 tower EVERY 2-3 YEARS? No thanks. Anyone inclined to purchase a new Mac Pro Tower definitely need to be able to upgrade RAM and video cards on a machine that costs that much. Otherwise just buy a studio and a really nice TB chassis for your cards and call it a day for $5-6k. Hopefully your TB expansion chases will still be compatible when you replace the studio. If not you'll still save $$$mfryd said:
...entropys said:...
Literally the only advantage the new Mac Pro offers is PCIE in a chassis which tends be more reliable and perform better than connecting the same PCIE cards of TB. That's it. The new tower is a complete mystery if they couldn't include upgradeable ram and video cards. IF Apple didn't want to work with NVIDIA or AMD or couldn't come up with a way of upgrading RAM they shouldn't have bothered with the new tower and left it as the only intel model. The market the tower was intended for was better served with x86 for at least the next few years. This feels like a rush job.
You are correct that there is a market segment that needs incredible power, and the need expands each year. That can mean swapping in new CPUs, new SSD, new I/O cards, new GPUs, more RAM, etc. on an ongoing basis. If you are editing, color correcting, or otherwise working on major motion pictures with budgets well over $100 million, you likely fall into this category. You also fall into the category, where you expect to replace these machines every year or two with faster and better ones. $7K for a new machine is small potatoes.
Remember, these are people who are happy to pay $5k for a monitor as they are used to paying over $20K for monitors.
Now if you are a home hobbyist, and want to incrementally upgrade your machine as budget allows, then the Mac Pro may not be for you. But then, a home hobbyist is not the "professional" market.
I once had a conversation with someone who wanted to start a small business. They were asking me about how to accept credit cards. This was a number of years ago, and things like Square were not available in their country. I explained that they would need to open a merchant account. They would likely pay $20/month in fees for the account. They replied that they couldn't afford $20/month. My response was that if they couldn't afford $20/month, they had a hobby and not a business.
