Software defined silicon is Intel selling you DLCs to unlock hardware features on products you own. They intend to ship this for upcoming Xeons. No idea why would you as a customer be exited about that kind of stuff.
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.
Intel Alder Lake vs. Apple M1
- Thread starter leman
- Start date
- Sort by reaction score
Software defined silicon is Intel selling you DLCs to unlock hardware features on products you own. They intend to ship this for upcoming Xeons. No idea why would you as a customer be exited about that kind of stuff.
Don't pay too much attention to silly comments since an article also mentions:
"To meet the immediate needs of its clientele, Intel wants to offer them CPUs in the configurations they need now, but with SDSi it can leave the door open for future software upgrades should a client need additional functionality or just decides to repurpose a machine. Such upgradability ensures that Intel's clients do not go to AMD if they need an extra feature or two and will still pay Intel for its technologies."
How would Intel predict what new features customers want in the future with predefined features from the past?
"To meet the immediate needs of its clientele, Intel wants to offer them CPUs in the configurations they need now, but with SDSi it can leave the door open for future software upgrades should a client need additional functionality or just decides to repurpose a machine. Such upgradability ensures that Intel's clients do not go to AMD if they need an extra feature or two and will still pay Intel for its technologies."
How would Intel predict what new features customers want in the future with predefined features from the past?
Last edited:
That would be an improvement to the current state of things. Today, if you don't buy the most expensive top-tier chip, you are likely to get a crippled chip with no option to re-enable the disabled cores. Intel, AMD, and Apple are all guilty of that.
These are not some “magical future features” Intel will be pushing to existing CPU, but real existing hardware features that have been disabled by the CPU firmware configuration. E.g. faster clocks, AVX512, matrix instructions, AI instructions, advanced virtualization features etc.
Apple seems to have left the door open for next gen Intel CPU like
alder lake . Yes because the latest version of macOS (Monterey)
works very well with alder lake…This may not be full support but may require
only “cosmetic patches”. I see this on hackintoshworld.eu
So Apple can easily make new computers (or upgrade) with intel cpus.
To understand this you have to go back in time when Apple goes to PowerPC.
Yes in this time PowerPC was two-times faster than best intel/amd CPUs…
Nowadays M1/pro/max chips are three times slower than best intel cpus…
And that’s why apple is leaving itself an “open door” for Intel.
This is reasonable because we don’t know if the next gen Mxxx chip can come
close to Intel desktop CPUs. if not Apple will have to left “workstation market” or
… equip new pro machines with next gen intel cpus (alder, new xeon etc).
Furtheremore note that monterey support new radeon series graphics 6xxxxx.
They are not designed for "silicon technology" and never will be...
So it seems unbelievable that apple will resign from top CPUs/graphics.
alder lake . Yes because the latest version of macOS (Monterey)
works very well with alder lake…This may not be full support but may require
only “cosmetic patches”. I see this on hackintoshworld.eu
So Apple can easily make new computers (or upgrade) with intel cpus.
To understand this you have to go back in time when Apple goes to PowerPC.
Yes in this time PowerPC was two-times faster than best intel/amd CPUs…
Nowadays M1/pro/max chips are three times slower than best intel cpus…
And that’s why apple is leaving itself an “open door” for Intel.
This is reasonable because we don’t know if the next gen Mxxx chip can come
close to Intel desktop CPUs. if not Apple will have to left “workstation market” or
… equip new pro machines with next gen intel cpus (alder, new xeon etc).
Furtheremore note that monterey support new radeon series graphics 6xxxxx.
They are not designed for "silicon technology" and never will be...
So it seems unbelievable that apple will resign from top CPUs/graphics.
I don't think Apple is worried that the M1 variants are 3 times slower than the best intel GPUs (xeons), which hare 3x more cores and consume up to 10x more power.
Or if you refer to laptop intel ships, I'm not sure where the M1 variants are 3x slower, unless maybe in running X86 code.
We considered doing this 20 years ago. We had ideas about a new type of floating point unit, cache sizes, etc. We rejected it as arguably too evil.
Yes we do. We know they are about to release high core count variations based on M1 cores. We know that M2 will be around 14% faster in single core performance. And we know that Apple has a long and very consistent history of increasing single core performance around 15-20% per year, whereas Intel doesn’t come close to that. It’s absolutely inevitable that with each generation, Apple will leave less and less of the high end to Intel.
Last edited:
Somehow you managed to top mitchy in posting scenarios that are far removed from reality, kudos.
PPC two times faster than Intel? Lmao what.
Apple Silicon being three times slower than Alder Lake? Did you just crawl out of the stockfish thread?
And you just assume that because MacOS is still compatible with x86 (they’re not gonna drop support for computers they’re still shipping), Apple’s going to do an about face and ship Intel again?
Especially when Intel managed to “beat” Apple Silicon by cranking clock speeds to nuclear furnace levels? When Apple’s been touting superior performance per watt since Apple Silicon debuted?
I have to be in some MKULTRA-level troll thread now. There’s no way anyone on this planet can be this attached to x86 that they gaslight themselves to this level.
And he’s defending it as revolutionary, somehow gonna beat Apple by disabling features on xeon cpus and selling them back to buyers.
This isn’t even relevant to his consumer-programmable fpga argument, since this isn’t consumer-programmable.
This is a positive in his world. Intel gonna spank Apple by disabling features on their own cpus.(?)
A promise of longevity from a company that makes a new cpu socket every couple of years means jack ****.
Intel makes processors with different power consumption. From 15 TDP up to 140 TDP.
Its not a problem for users who need max performance and dont care about mobile version.
We spend thousands of $$$ for our computers...What is the point of saving for example 100$ a year
on electricity bills ? Any point. If I can reduce the time of rendering my projects form two hours to
one hour thats the goal. The sooner you do something the more money you make )
What makes you so certain Apple won't have a similar range once the transition is complete? Actually, if Apple could pull it off, it would be even more impressive. They would have everything from the S-series chips in Watch/HomePod to the Mac Pro.
Intel seems to think it’s important, why else would the 12900k have 8 efficiency cores instead of performance cores?
Alder Lake efficiency cores are not power efficient. They are area efficient for Intel to manufacture- let's say they give half the performance of a P-core in 1/4 the area? Not at all like what Apple does.
I’m aware, but my point was that Intel wouldn’t put “efficiency” cores on their highest consumer product if efficiency didn’t matter.
It is not talking about "new" features for a given generation's implementation. It is talking about activating features a company might have thought they didn't need but their service levels and requirements change.
This is all multiple decades old stuff in the mainframe world. IBM Z-class mainframes have had this stuff for years. Part of this is Intel offering the same stuff that folks "trading in " mainframes are likely asking for. It doesn't come cheap. ( and if not doing a wide variety of workload aggregatoin on a fixed number of servers , not all that high utility. Probably not a huge threat to AMD server progression. It is a bigger moat builder around a some well defined subsets that Intel has more penetration in. )
Intel has far too many server SKUs. Intel has almost 2-4x as many as AMD has. This should help them simplify their inventories and give customers more flexibility to shift there "clusters of server CPUs" between different workloads at time change. Should also lead to incrementally less e-Waste also. (don't have to "throw out" a die that has some units permanently 'fused off' that doesn't match workload anymore. )
There is absolutely nothing to do with changing what the silicon shipped does. It is turning units on/off and/or setting run time parameters.
Relative to the Intel 'P' cores they are power efficient. There has to be a context applying that phrase to.
Apple's E cores are also area efficient. 4 E cores plus their L2 fits in area of a single P core. The 2 E cores of the M1 Max/Pro are less area than the a couple of the Thunderbolt controller complexes. Some folks complain about the 2 E cores and claim get rid of them and bring in some real P core or blank-blank wonder feature. If do a straight swap.... there is isn't much could add. All the other major function unit areas on a M1 Max/Pro die are much , much bigger.
So relatively inside of a single implementation by a respective company the 'E' cores are smaller and more power efficient. The size of the gap between P and E isn't the same on both metrics. But the utility ( use some E to save space and save lots of power when lightweight workloads is the basically
the same.)
Different companies have different approaches. My impression is...
Apple efficiency core is very low performance relative to performance core but benefits from low <1W power consumption at idle.
Intel's hybrid E-core is mid to 3/4 performance of P-core and more like essential core leaving out extraneous rather than efficiency core. Idle power consumption unknown.
AMD uses power/frequency scaling instead of E-core that doesn't take up die space but still manages to have low ~<1W idle power consumption.
On the surface, AMD's decision seems the most sound.
Apple efficiency core is very low performance relative to performance core but benefits from low <1W power consumption at idle.
Intel's hybrid E-core is mid to 3/4 performance of P-core and more like essential core leaving out extraneous rather than efficiency core. Idle power consumption unknown.
AMD uses power/frequency scaling instead of E-core that doesn't take up die space but still manages to have low ~<1W idle power consumption.
On the surface, AMD's decision seems the most sound.
They’re putting it in for multithreaded performance that they can actually manufacture. If they used P-cores there would be better performance but they would have no yields. It matters for their highest consumer product that Intel has the ability to actually make it.
Well there must be some benefit to the big.little design if AMD follows through with this patent https://www.hardwaretimes.com/detai...and-process-node-for-different-core-clusters/
Perhaps the power/frequency scaling isn’t the best method, which I believe is the case. As it pertains to mitchy, he only claimed it’s the best method because Apple doesn’t use it. Somehow Apple and Intels chip designers should’ve consulted him.
I’m aware of this argument too, but since Intel claims it’s for power efficiency as well , I’m going with power efficiency being a factor. Likewise the AMD patent above.
No, they aren’t power efficient compared to P-cores. Look at chart- at almost every power level, P-core offers greater performance. E-cores use lower power, sure, but also much lower performance. But they are area efficient “under given physical constraints”.