Intel Says New Core i9 Processor for Laptops is Faster Than Apple's M1 Max Chip

[Danfango]

The irony of this is Intel used to sell ARM SoCs… (look up StrongARM).

As for this release, no thanks. My corporate mandated Dell 5550 gets “thermal events” and hibernates every 20 minutes if you even stress it slightly. That’s only got an i7 in it. If you kick the RTX in and plug in the power brick it is louder than my washing machine.

Also I quite like not having baked genitals.

 

[Shirasaki]

As expected, all snarky comments, though I'd argue they are well justified. I only recently notice how hot the Intel Core-i5 9300H is, reaching 100C when room air flow is poor, though it quickly backs off to 75C without much throttling.

To be fair, M1 isn't energy efficient enough either, and 50C is definitely something that is noticeable, and 4hr battery life is far from advertised 10hr battery life, all while just running Numbers at the front and mostly Apple Silicon applications in the background. I can certainly believe Apple Silicon being much more efficient down the line, but software catchup (optimized for Apple Silicon) still has a long way to go.

 

[Shirasaki]

The irony of this is Intel used to sell ARM SoCs… (look up StrongARM).

As for this release, no thanks. My corporate mandated Dell 5550 gets “thermal events” and hibernates every 20 minutes if you even stress it slightly. That’s only got an i7 in it. If you kick the RTX in and plug in the power brick it is louder than my washing machine.

Also I quite like not having baked genitals.

Then I seriously question your company's software package, cause my Dell G5 can run 35 days straight without a single BSOD, rarely any noticeable slowdown (only in games after running 35 days). It is loud though lol.

 

[Bug-Creator]

Draw a vertical line at 35 watts, see which has more performance. It doesn’t have to go near 115 Watts in this test to surpass M1 Max performance.

Sure, draw one at 25W and the M1Max is still offering plenty performance while the Intel has to be in some level of sleep to get that low.

Your mileage may vary but having a computer idle or only doing light tasks while organic part of that workplace is trying to catch up isn't that unusual.

 

[downpour]

If this chart is true then Intel are providing significantly higher performance at a lower wattage than Apple (So not only faster but more efficient). This is not a good sign for Apple considering the M2 Max is probably a long way off.

 

[0924487]

They probably engineered the test to have this edge case where it’s actually presentable.

They didn’t even use the usual GeekBench etc., that probably means it’s worse than M1 on those generalized tests.

 

[Miha_v]

I guess Intel needed a kick in the rear (from Apple) to up their game ?
Competition is great for all of us consumers. I'd wait to see real-life performance and other standard tests though...

 

[Shirasaki]

Apple's best play would be to figure out how to marry both Intel silicon and Apple silicon onto the same motherboard, best of both worlds. When a MacBook Pro is plugged in, it would take full advantage of both SoCs. On battery power, it could fall back to only the Apple SoC.

With support for building Universal apps, a single app package already runs on both processors, so such a hardware architecture is not that far out of reach. Figuring out how they would cooperate is the biggest challenge.

This is simply not a thing, as apple has determined to kick Intel out, alongside any semblance of x86 on Mac platforms.

 

[kanedavid]

Well based on Intel's response to Apple's M1 SoC when it was released, clearly Intel have been known to cherry pick specific tests and stats to suit their marketing spin... We will obviously see what the real world tests reveal in due course, but I suspect that as per the previous Intel Core generation CPUs, their flagship model will likely eek out slightly better performance in a few specially-crafted tests, but overall, I suspect the M1Max will kick Intels butt... with M2 to make further gains.
Clearly moving to ARM based CPU's was a big win for Apple... the only minor blemish is that there are still a lot of people that need to still run some specialist Windows software - Hopefully Microsoft figure out a way forward to allow Apple users to run Windows on their Apple Silicon Macs!

 

[Bug-Creator]

Apple's best play would be to figure out how to marry both Intel silicon and Apple silicon onto the same motherboard, best of both worlds. When a MacBook Pro is plugged in, it would take full advantage of both SoCs. On battery power, it could fall back to only the Apple SoC.

So in short:
- increase the BOM by 200-300$
- decrease battery size by 20% (if keeping the same form factor)
- decrease battery life by another 20% (since you won't be able to turn off everything related to Intel in that thing 100%)
- confuse developers (aka giving them an excuse to not do their job)

All for what benefit ???

 

[theotherphil]

If you watch

particularly with high load stress 16" MBP M1 Max briefly hit AC 110 watts (fluctuating) at 05:20 time point, obviously its very efficient with other usage loads.

So, the entire 16” system, under full load draws the same as Intel’s CPU alone? That’s CPU/ GPU, SSD’s, 16” mini-LED screen etc, etc? LOL!

 

[chucker23n1]

Makes sense that both Apple and Intel are starting to make charts like this. For one, you can better capture the relationship between power and performance. And as a bonus, it's much harder to actually read much of anything into this, especially when your y-axis is "relative performance". (Relative to the 11980HK at 45W, I think is the idea.)

What they seem to be saying:

• at 35W, the 12900HK is the fastest, the M1 Max is about 90% that, and the 11980HK and 5900HX are about 82% and 80% that.
• at 45W, the 12900HK is another 20% faster, whereas the 11980HK and 5900HX are only about 5% faster each (but still not quite as fast as the M1 Max). At this point, the 12900HK is about 35% faster than either the 11980HK or the 5900HX.
• at 65W, the 12900HK is another 10% faster, and the 11980HK another 5% faster.

One question I would have is how they would measure the M1 at particular power draw. And I guess they didn't. They say "Apple M1 Max performance is estimated based on public statement made by Apple on 10/18/2021 and measurements on Apple M1 Max 16" 64GB RAM Model A2485." In fact, even their own performance is "estimated". Interesting CYA speak.

 

[johnnytravels]

I guess it's good that high-powered Windows machines will continue to exist. The use case for these is probably not mobile work per se but being able to have a high-powered machine with you when you travel from power outlet to power outlet (e.g. between office, field and home). I think it's a legitimate use case and I don't mind that this new generation i9 exists and is somewhat more powerful than the the M1 Pro Max in certain scenarios.

 

[podycust]

How does an i3/ i5 12th gen compare? as most consumers aren’t buying an i9 lol

 

[draugminaion]

M1 Max Duo and Quad (if that's what they will be called) will probably be here pretty soon.

 

[FlyingDutch]

That’s plain dumb advertising… they are speaking about performance, and their own graph is showing quite poor power efficiency. In a notebook power efficiency is a must.
M1 architecture is 1 year old and Intel still isn’t there (and won’t be there with Alder Lake).

 

[FlyingDutch]

If it's true then great, we as consumers win. Competition is great.

It is true, if we are speaking about raw performance, but what about power efficiency ? M1 and i9 aren’t even in the same league…

 

[Marx55]

Intel history is filled with delays, unfulfulled promises and wrong statements. I do not believe them any more. Remember the Intel Optane Xpoint fiasco? It was going to be 1,000 times faster than current SSD, yet it is even slower (only marginally faster on random read/write), requires a lot of energy, is not available at large sizes and is very expensive.

In the case of microprocessors, compare the performance and Thermal Design Power (TDP) of Intel x86 vs ARM-based Apple Silicon. The former may be more powerful on some very expensive microprocessors, with huge TDP (read huge energy consumption and low battery time) and huge price as well. Intel and AMD should switch to ARM. The writing is on the wall.

Update confirming the huge TDP of Intel microprocessors:

"That brings us to the new mainstream 65W desktop line which ranges from 2-8 cores and has a real TDP of up to 202W".

Source:
Intel releases lots of hardware at CES

Intel releases lots of hardware at CES

Intel is announcing a lot of goodies at CES ranging from CPUs to anti-competitive marketing plans.

semiaccurate.com

 

[marios_geo]

The amount of people that can't read and simple graph is flabbergasting.
There are two simple axes, power and watt. Still the majority keep complaining but "at what watt is the power dude".

Jeezz

 

[Wulf^2]

Makes sense that both Apple and Intel are starting to make charts like this. For one, you can better capture the relationship between power and performance. And as a bonus, it's much harder to actually read much of anything into this, especially when your y-axis is "relative performance". (Relative to the 11980HK at 45W, I think is the idea.)

What they seem to be saying:

• at 35W, the 12900HK is the fastest, the M1 Max is about 90% that, and the 11980HK and 5900HX are about 82% and 80% that.
• at 45W, the 12900HK is another 20% faster, whereas the 11980HK and 5900HX are only about 5% faster each (but still not quite as fast as the M1 Max). At this point, the 12900HK is about 35% faster than either the 11980HK or the 5900HX.
• at 65W, the 12900HK is another 10% faster, and the 11980HK another 5% faster.

One question I would have is how they would measure the M1 at particular power draw. And I guess they didn't. They say "Apple M1 Max performance is estimated based on public statement made by Apple on 10/18/2021 and measurements on Apple M1 Max 16" 64GB RAM Model A2485." In fact, even their own performance is "estimated". Interesting CYA speak.

Dear all,

let us bring this discussion back to documented facts. Anandtech did report the much discussed rate n integer Spec2017 scores for some of the processors of Intels chart. You find the numbers here: https://www.anandtech.com/show/17024/apple-m1-max-performance-review/5

The numbers are:
M1 Max: 53.38 @ maximal CPU power, which is around 35 W (see same anandtech report)
11980HK: 35.46 @ 35 W
11980HK: 38.82 @ 65 W

Thus, the plot intel shows does not agree at all with these independently published results from anandtech. 11980HK only reaches 72% of M1 Max maximal performance and in the Intel plot, it reaches more than 100%. If this number given by Intel, that everyone can verify, is indeed wrong, how much do you trust the numbers for 12900HK, which no one can verify at the moment?

Note that the M1 has a lead in integer performance in the anandtech report but it is the smaller lead compared to that of floating point, i.e. Intel did indeed cherry picking.

Anyhow, I guess it will take only weeks before we get first well documented tests from anandtech with numbers that can be trusted.

Regards,

Wulf^2.

 

[the8thark]

Typical Intel. Still stuck in the MHz wars.

The whole point of M1 is to have more grunt and more things too can do per mhz output by the chip. Add in the fact that a lot of the work is offloaded away from the main cores to secondary cores on the SoC.

Also faster seems to imply way more thermal constraints for this new i9 cpu.

I agree that competition is always good. I would just prefer the competition be a little more competent and learn from the mistakes of the past.

 

[chucker23n1]

The amount of people that can't read and simple graph is flabbergasting.
There are two simple axes, power and watt. Still the majority keep complaining but "at what watt is the power dude".

Jeezz

Yeah, easily half the comments don't seem to have looked at the chart at all.

 

[ahurst]

why are so many people hating on intel here? if they pull off something good, great. if not, then boo.

Can’t speak to others, but for me personally:

a) It’s always fun to watch an old complacent champion panic and flounder in the face of a newcoming challenger,

b) Intel’s been frustratingly slow to innovate for *years*, such that it wasn’t until the M1 that a new Mac could come close to doubling the single-core of my 2013 (!) i7 iMac,

c) Intel’s had firm control of the direction of the desktop/laptop market for years, owning most of the instruction set patents and supplying most of the chips OEMs use. Apple throwing them off-balance and highlighting a market for mid-to-high-performance ARM chips in the PC space would shake things up and loosen their near-monopoly. I’ll be happy to see improvements in Intel chips when AMD and/or Qualcomm both have comparable shares of the PC chip market, until then I’ll relish every PR blunder and product misstep.

 

[sartorius]

No chance this thing is not an overheating piece of **** in a laptop form factor, or has to be crippled so much that it is par/slower than M1.

I'll never go back to Intel. Never. I will rather move to the mountains and become a hermit than use their garbage.

My 2019 i9 MBP was the worst piece of hardware I ever owned, what a perpetually throttling, perma-fan piece of caca.

They're always happy to coast until someone lights a fire under their fat, lazy ****ing asses.

 

[MBAir2010]

WOAH .. I actually thought you were joshin everyone.

i josh not

and own a Dell XPS 13" (2019) when  could not make a laptop that actually worked and had a lifespan.
this new Dell XPS seems great and i hope the battery last 14 hours or so.

i don't understand the hate geared towards a company that want to succeed in their field.