I would buy the appropriate part capped at 45-65 watts. Might be quad-core or hexacore. I like things to run cool, especially after owning Apple Silicon systems.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
[CPU only] Apple M1/2(Max/Ultra) (TSMC 5nm) vs AMD Zen 4 (TSMC 5nm) - Technical Analysis
- Thread starter exoticSpice
- Start date
- Sort by reaction score
I would buy the appropriate part capped at 45-65 watts. Might be quad-core or hexacore. I like things to run cool, especially after owning Apple Silicon systems.
If your workload is well multithreaded and you can split it over a lot of threads, then it may not be bad running one of those at lower power. The 7xxx series is better priced to do that though. I remember the 5950x selling for about $1,300 when it launched.
I would be curious to see if any of the OEMs provide that level of support in the UEFI. I know it exists (mine calls it either ECO or Low Power mode I forget exactly which), the big issue being limited base and boost clock when limiting TDP/PPT (along with limited TDC & EDC).
I think you are not taking into account that the Studio was a replacement for the 27 iMac. At $1,999, it is already priced $200 higher than the old iMac 27" and it has no screen! Even higher and the Studio Display also being $200 higher than an Apple screen and that would make the Studio even more problematical coverage for the old iMac 27" entry point.
The 27" iMac survived just fine with a "better"/"best' iMac 21.5" price in the $1,400-1,500 range. There is zero rational motivation that presents as M2 Mini Pro in the $,1400-1,500 as being hugely flawed. It worked before well for Apple. It would work again at the same gaps between model offerings. It is the same thing only they have pruned the screens off the offerings. ( so the completed system price for users is even higher for those who need to buy a new monitor).
The Studio starting with at least 24 GPU cores and the M2 Pro likely toping at 20 GPU cores is a product differentiation gap. They can make the gap wider once get to M2 Max that has the same GPU core count bump. The Studio is going to run a bit quieter for folks who want to extra pay for that also.
M2 Pro will work because many will want to spend that gap money on more RAM , more Store , and/or something else because the M2 Pro is "fast enough". Apple 'pulling' folks into spending more on BTO options actually likely gives them higher profits.
Not sure why surprised. The iMac 27" was likely one of the better selling desktop models Apple had. The overall iMac ( non Pro 21.5-27") was by far the best selling desktop model. So a major part of that segment probably would be best (or at least minimally 'better' ) selling also.
What were iMac 27" folks suppose to do? Dump it for a M1 Mini ( with gimped Display out) or smaller screen 24" or way too expensive and increasingly dated/stale Mac Pro 2019? Probably not. [ the 16 core CPU + W5700 configuration that was most popular Mac Pro configuration is actually slower than the up configured Studio for thousands of dollars less. Why wouldn't that sell well? There are lots of folks who will gladly take the multiple thousand dollar savings. ]
Where Apple placed it was right on top of approximately the middle of the old iMac 27" configurations ( $1,999 . Not the cheapest but also not on the BTO upscale price either. It was a standard configuration that likely sold very well off the shelf. ).
P.S. The iMac Pro sold much better than many folks though it did. In its heyday it was about 5th or so best selling 27" iMac configuration at B&H ( could rank at their website). upscale Studio + Studio Display and back in iMac Pro zone price wise. Folks cross upgrading from iMac Pro -> Studio would be an easy pipeline also.
Last edited:
The 7xxx series isn't priced better at MSRP.
AMD Zen 3 Ryzen Deep Dive Review: 5950X, 5900X, 5800X and 5600X Tested
MSRP was $799 for the 5950X. Has been retailing for less now
" ....
| Ryzen 9 7950X | $699 | 16 / 32 | 4.5 / 5.7 | 80MB (16+64) | 170W / 230W | DDR5-5200 |
| Ryzen 9 5950X | $546 ($799) | 16 / 32 | 3.4 / 4.9 | 74MB (8+64) | 105W | DDR4-3200 |
..."
AMD Launches Zen 4 Ryzen 7000 CPUs, Arrives Sept. 27 Starting at $299
Into the 5nm era
$1,300 is hype price ( also a bit of Intel screwing up price).
If look at the chart there in the toms hardware article, AMD's prices are generally up over (and at best the same) for the 7000 series.
The current context is substantively different. Intel Gen 13 will be far more competitive. Some workloads AMD will have an edge and others Gen 13 will. That will put a huge damper on the hype price. More supply spread over two suppliers and more competition.
P.S. when the 3D cache options arrive the price point will be higher still. (will get more , but also pay more ) .
P.P.S. with PC sales dramatically slowing I doubt Intel and AMD are going to get into a pricing war on lowering these higher end chip prices. They have older inventory to push into a slowing market also. Folks pressed for cash will be doing older board upgrades. They can't discount the new CPUs around a new motherboard upgrade.
Last edited:
I have 3 27 inch iMacs so I do know a bit about them. I'm typing on one of them right now. I also have an M1 mini on my desktop. I run compute-intensive stuff on the M1 mini and memory intensive and office stuff on the iMac. It's a nice combination.
I see lots of 27 inch iMacs where they list the purchase price at $2,500 - $3,500 and the Mac Studio has twice the compute of the 2020 i7 iMac so they can charge those prices. Furthermore, the Studio can drive five monitors while the 2020 iMac can only drive 3. You'd have to get the iMac Pro if you wanted support for five monitors.
What do 27 inch iMac owners do? You get a mini or Studio and add a Studio Display or you hack together something like what I have.
I have a todo item to try to get 5k display out from the M1 mini to the Intel 5k iMac but it's not on my priority list as using the systems cooperatively is working really well right now.
I hope the M2/3 is running at 4ghz in the Mac Pro and they let it have some more juice to let it show its true full power.
The 7xxx series isn't priced better at MSRP.
AMD Zen 3 Ryzen Deep Dive Review: 5950X, 5900X, 5800X and 5600X Tested
www.anandtech.com
MSRP was $799 for the 5950X. Has been retailing for less now
" ....
Ryzen 9 7950X $699 16 / 32 4.5 / 5.7 80MB (16+64) 170W / 230W DDR5-5200 Ryzen 9 5950X $546 ($799) 16 / 32 3.4 / 4.9 74MB (8+64) 105W DDR4-3200
..."
AMD Launches Zen 4 Ryzen 7000 CPUs, Arrives Sept. 27 Starting at $299
Into the 5nm erawww.tomshardware.com
$1,300 is hype price ( also a bit of Intel screwing up price).
If look at the chart there in the toms hardware article, AMD's prices are generally up over (and at best the same) for the 7000 series.
The current context is substantively different. Intel Gen 13 will be far more competitive. Some workloads AMD will have an edge and others Gen 13 will. That will put a huge damper on the hype price. More supply spread over two suppliers and more competition.
P.S. when the 3D cache options arrive the price point will be higher still. (will get more , but also pay more ) .
It was the price in the era of crypto-mining. Intel CPUs could be had at MSRP back then.
I'm talking back in early 2020.
There likely is not "true full power". When Apple shows perf graphs they cut off the diminishing returns tail that AMD/Intel sell and market ( as overclocked turbo extension). Apple isn't hiding some huge performance uplift. Their design is fundamentally built on lower clocks and wider bandwidth. If crank the clocks and do not move the bandwidth they are not likely going to get much. They are already squeezing the max out of the speculative execution, cache, and branch prediction (specialized caching).
Pretty unlikely Apple is going to do a "redesign from scratch" CPU/GPU just for the Mac Pro. Likely going to be the same CPU and GPU cores precision tuned to laptops , tablets, and phones. The only power gap likely is that they will trade-off burning more power to open up the I/O bandwidth to non memory workloads. (e.g, more PCI-e , directly/indirectly SATA , maybe some secondary memory , etc. ). It will be more cores and bigger area budgets (bigger chips and bigger multiple chip packages ), but largely the same building blocks with better/expanded interconnect.
Bringing a giant GPU in close proximity to more than a handful of CPU cores is going to be a 250-350W package. Apple is not likely going to try to run that up into 400-500W range.
They were also better gaming chips than 10/11th gen intel ones. So that didn't help lower prices.
They were almost 100% over MSRP. AMD kept saying that they would increase supply; same with their GPUs. I just skipped the generation and went with Intel as they were a lot easier to buy.
The power may increase a bit but overall efficiency will be much higher, AMD released the data transfer (energy) estimate for IF some time back in terms of pJ/bit or byte & it has only improved since then. From what I remember IF was the most efficient interconnect at the time, this may have changed recently with Apple's Mx series but for larger power hungry chips, on desktop or servers, IF is still the most efficient.
You're talking just the x3d cache? If so yes it consumes more power. The overall power consumption will go down because you don't need to fetch that data, present in cache, from the main memory.
It's just to show how efficient 7950x is at lower TDP's you can easily get a 7600x & achieve similar levels of efficiency.
Eventually supply caught up. But yes for this generation of chips Alder Lake was much easier (and cheaper) to get if you were willing to pay extra for DDR5...
We're talking about late 2019 and early 2020 and that was tenth-gen Intel and Ryzen 5xxx. From recollection, 11th gen and 12th gen Intel chips were easy to buy at MSRP. I was pretty annoyed that my motherboard was only good for one more generation despite being a new motherboard. I much prefer AMD's approach of supporting chips for a few generations.
I've found that Intel boards have been easier to work with than AMD boards, mainly in installing the CPU. It is possible that AMD has gotten better in this regard.
Yeah AM5 is LGA instead of PGA, that will be nice (coming from Threadripper which is LGA going to AM4 felt like going backwards).
Zen 3 didn't launch until late 2020. It was really hard to get until mid to late 2021.
Pretty much the SFF community. Almost doubling of performance at 65W is a compelling upgrade from 5950x. Now, to decide between 7950x and fat cache 7950x3D.
I guess I was off by a year. I think that we did two separate builds. My i7-10700 and a Ryzen 5 something and my daughter bent some of the pins putting the CPU in. Trashed the MB and the CPU.
The question is more, if Apple is willing to produce a desktop-only SoC. In this case they can easily go beyond 4GHz.
AMD said their TSMC N5 node is 'custom' and optimised for HPC. Seems the desktop processors are taking a free ride for performance.
AMD also said their TSMC N6 based new I/O die for clients consumes 20W under load. Hope this settles the speculation.
Independent of the question if the cell library is custom, even the standard TSMC N5 cell library offers cells with vastly different performance and power characteristics for the same cell function. The general rule here is, the slowest cells have the lowest leakage. In a mobile design you try to minimize leakage and tend to favor slow cells - in an HPC desktop design you can shift your cell distribution easily towards the high leakage cells.
The words Papermaster actually used was a "specialized , N5 , HPC process that enabled additional frequency past the baseline process". It is rather a odd if AMD is using the original N5 fab process at this point. N5 is over two years old at this point. Why would they jump from N6 to a two year old process? TSMC rolls out a couple of letter 'modifier' process specialized increments after the baseline gets stable anyway to their customer base as another option.
There is pretty good chance that this is technically N5P that they are moving to. N5P is over a year old. . So it isn't new either. (no high risk of initial growing pains or too few wafers to go around) N5P get +10% power or +5% performance over the baseline N5 process. So something as simple as just using N5P fits his basic description without going into a zone where only AMD can get it. AMD could easily just skip using the letter adjectives to make a simpler , clearer presentation without going into the weeds of TSMC fab tech.
Also would not be surprising if Zen 4c was on N4P. And that N4P is a small contributor to making the 4c cores smaller. ( 6% isn't huge chunk of a 50% reduction but if can get it for 'close to free' why leave it on the table. Getting 44% is a bit easier. ). 4c arriving closer to June 2023 would make sense as N4P ramping now.
From Papermaster's description of balancing power , area , and high frequency 4c could have a heavy contribution of a library shift that tosses out the high frequency ability and chooses area and power saving options instead. (plus loose some re-write registers and/or cache . Can juggle design libraries, but some part of that 44% is likely just a chunk of something being dropped. ). Can still have the 'add' function, but perhaps not three 'adds' in parallel ability.
N4P takes the same baseline design rules as N5P. And if swapping cell libraries anyway to remove the "turbo" frequency trade-off anyway there is a re-flow design overhead anyway. Adding shift to N4P on top of that isn't going t be that much more expensive. ( And they can probably re-use chunks of this work on a laptop part later ). AMD's roadmap diagram says that Zen 4 family is spread over N5 and N4.
" ... The new I/O die uses the 6nm process and houses the PCIe 5.0 and DDR5 memory controllers along with a much-needed addition for AMD — the RDNA 2 graphics engine. The new 6nm I/O die also has a low-power architecture based on features pulled in from AMD's Ryzen 6000 chips, so it has enhanced low power management features and an expanded palette of low-power states. AMD says this chip now consumes around 20W, which is less than it did with Ryzen 5000, and will deliver the majority of the power savings we see in Ryzen 7000.
Surprisingly, the new I/O die appears to be roughly the same size as the previous-gen 12nm I/O die. However, given that the 6nm die is far denser than the 12nm die from GlobalFoundries, meaning it has far more transistors, i."
https://www.tomshardware.com/news/a...ications-pricing-benchmarks-all-we-know-specs
The desktop 7nm tees up the 6nm laptop which sends contributions into the 5nm desktop product. AMD doing a steady set of base hits just to move the products around the bases to get to a 'homerun'. Intel keeps going to the plate and trying exotics stuff to hit some 500ft homerun shot over the center outfield wall. Apple can beat up on Intel from time to time with the strategy they are following. It is going to be lot harder to beat AMD when they are on the same process node with the non-over-reach strategy they are using. Apple is somewhat lucky AMD is mainly trying to build server focused chiplets as opposed to going "all in" on laptop product. But AMD putting laptop SoC last on the priority queue is still getting numerous iterations at this point (at least at the high end of the laptop SoC product line).
Last edited:
Thanks, very helpful! Hope to see some in-depth reviews once the products are out.
Does anyone have some recommendations which tech review outlets to follow now that both Andrei and Ian aren’t with Anandtech anymore?