News Intel Engineer Outs Panther Lake Architecture on LinkedIn

[bit_user]

May 29 2022, Hardwaretimes references Panther Lake: https://www.hardwaretimes.com/intel...lunar-lake-reportedly-an-experimental-design/

That cites an anonymous leaker, rather than any official source. I'm not surprised there have been such leaks, but the LinkedIn leak that just happened seems to both have more details and come from a more authoritative source.

Whoever posted those details on their LinkedIn profile was simply wrong about what had been announced vs. what hadn't, and wasn't diligent enough to actually go back and check. That seems the simplest explanation.

Jan 12, 2023 WCCFTECH list s Xe3 (Celestial) as the GPU architecture for Panther Lake in this article on lunar lake:

Intel Lunar Lake To Feature A Brand New CPU Architecture Built From The Ground-Up, Perf/Watt Focused at Mobile

Intel is working on a brand new CPU architecture designed from the ground for its next-gen Lunar Lake family of processors.

wccftech.com

Heh, it seems someone has been going back and retroactively updating their news articles! If you look at this snapshot from when that article was published, it lists zero details about Panther Lake, besides a couple guesses about the chipset, socket, and PCIe.

Panther Lake (TBA)

TBA

TBD

TBD

TBD

1000-Series?

LGA 1851?

DDR5

PCIe Gen 6.0?

2026

Source: https://web.archive.org/web/2023011...om-the-ground-up-perf-watt-focused-at-mobile/
​

It's fine if they update their articles, but they should indicate that they've done so.
(note to self: need to increase my annual donation to web.archive.org)

 

[edzieba]

You assume too much young padiwan. Intel hasn’t merely joined them, they have broken through the glass roof of hypocritical naming schemes. Just because process node names have gone away from first gate length, then pitch, then smallest feature size, doesn’t mean changing measurement dimensions is par for the course. Switching to a short lived 10^-1 difference in dimension scale is nothing more than a marketing ploy. I can see it now, the mainstream media saying, “while Samsung and TSMC are struggling to break through the nanometer wall having just introduced 2nm, Intel is enjoying the lead already breaking into the Angstrom era with their new 20A process.” When in reality 2nm and 20A processes are the same measurement and are in the same process class since Intel rebranded their 10nm and newer processes to reflect parity with TSMC 7nm and newer processes.

The problem with peanut-gallery commentary being incorrect stretches back over a decade, and will continue regardless of nomenclature. Years of breathless commentary on TSMC's 'new' 14nm nodes (same scale as their 16nm) whilst poking fun at Intel 14/14+/14++ evidence that, and it's not going to change from replacing an "nm" suffix with an "A" suffix.
Getting incensed over moving the decimal point to keep process scales round numbers makes about as much sense as demanding everyone call 800nm "0.8 um" to avoid changing suffix.

 

[The Historical Fidelity]

The problem with peanut-gallery commentary being incorrect stretches back over a decade, and will continue regardless of nomenclature. Years of breathless commentary on TSMC's 'new' 14nm nodes (same scale as their 16nm) whilst poking fun at Intel 14/14+/14++ evidence that, and it's not going to change from replacing an "nm" suffix with an "A" suffix.
Getting incensed over moving the decimal point to keep process scales round numbers makes about as much sense as demanding everyone call 800nm "0.8 um" to avoid changing suffix.

Except that Angstrom is not an SI-unit of measure and is no longer an acceptable unit of measure by the National Institute of Standards and Technology (NIST) or the (International Bureau of Weights and Measures). In fact I haven’t seen a scientific journal article using Angstroms since the late 90’s. Intel adopting the officially obsolete and no longer supported angstrom measurement unit literally tells me that Intel’s motivation is complete marketing BS, IE “let’s use this obsolete measurement unit to give us a perception edge with non-science consumers and media types”. That’s like Pepsi changing their volume measurement from 12oz to 120 dollops so that it appears Pepsi gives you more liquid than Coke. It’s disingenuous marketing any way you look at it.

Listen, I get that you are an Intel Lover (heck, my favorite CPU of all time is my Sandy Bridge-E 3960x that has been rock solid stable with a 52% all-core overclock for 10 years now. I absolutely love Intel’s last planar transistor process in their 32nm node), but your not helping your case by comparing giving Intel grief for going through 3 generations of 14nm (that Intel in fact made larger by increasing the transistor gate design from 70nm to 84nm for 14nm++) because they couldn’t get 10nm working correctly to TSMC’s labeling of an enhanced version of their 16nm node with 20% greater transistor density as 12nm (TSMC never produced a node named 14nm).

Oh and I disagree with your last quip. 800nm and 0.8um are the same size and complies with SI-unit standards, so no problem there using either.

 

[PEnns]

Intel Engineer Outs Panther Lake Architecture on LinkedIn


"Outs"?? LOL!!

A leak by any other name. And an intentional one too!!

 

[edzieba]

Except that Angstrom is not an SI-unit of measure and is no longer an acceptable unit of measure by the National Institute of Standards and Technology (NIST) or the (International Bureau of Weights and Measures). In fact I haven’t seen a scientific journal article using Angstroms since the late 90’s. Intel adopting the officially obsolete and no longer supported angstrom measurement unit literally tells me that Intel’s motivation is complete marketing BS, IE “let’s use this obsolete measurement unit to give us a perception edge with non-science consumers and media types”. That’s like Pepsi changing their volume measurement from 12oz to 120 dollops so that it appears Pepsi gives you more liquid than Coke. It’s disingenuous marketing any way you look at it.
[Ad-hominems clipped]

Angstroms are alive and well in the semiconductor industry, just as Rankine hangs around in rocketry like a bad smell.

 

[SiliconFly]

I think I wasn't paying attention and assumed that Meteor Lake would have a Battlemage (Xe2) iGPU, so this is a bit of a reality check.

Actually, Battlemage was initially paired with Arrow Lake. But knowing AXG's efficient execution, Battlemage got pushed to Lunar Lake.

And it's impossible to pair Battlemage with Meteor Lake. Meteor Lake test chips are already out and will go for volume ramp in early Q3 this year. But Battlemage is still under development. Even under the best of circumstances, a Battlemage tape-out can happen only in Q3/Q4. And if paired with Meteor Lake, it'll drag meteor lake to Q2 2024 at least. Not good.

 

[SiliconFly]

Yeah, I really didn't know what to expect. I'm actually a little surprised it's not.


I think that's worth a fact-check, because I'm pretty sure the details about which CPU would use which graphics weren't detailed at those events. And, apparently, nobody on this site or one other I follow had heard of Panther Lake, either.


No... Intel gets a pass for unit-less names like "Intel 7" and "Intel 4". It's talking about Angstroms that gets them in trouble.

And it doesn't help that the author is mindlessly parroting their marketing on this, as if it's technically accurate.

It's ok to be a fan of TSMC/AMD. But history is very important my friend! Your arguments need to be rational and well-balanced. You keep continuously pointing out that the Intel node naming scheme is ridiculous. Maybe. But you have NEVER spoken one bit about how ridiculous TSMC's node names are!!!

TSMC was the first to break the cycle almost 2 decades back with ridiculous node names/numbers. Intel has just been following suite all these years. When AMD proudly announces that their products are coming out in TSMC N7, not a single fanboy questioned 7nm!?!?! When AMD recently announced their Zen 4 is on TSMC N4, there was a huge fanboys celebration! No one spoke about the ridiculous 4nm claim!!!

But when intel says it's coming out with Arrow Lake on Intel 20A next year, the fanboys immediately have huge problems with Intel's node naming, performance, power efficiency, density, powervia, frequency, and how fake the node is in general. When in reality, it's far superior to TSMC N3 which nobody wants to mention. And the most typical silly argument of all, Intel will never deliver. When in fact, Intel already has 20A almost ready to go and TSMC 2nm is still in drawing board for a 2026 launch!!!!!!!!!!

Intel has pulled all stops & is going to crush competition starting next year. Kindly deal with it. In fact, if AMD doesn't deliver a powerful Zen 4 refresh on TSMC N4 this year that can effectively compete with the measly Meteor Lake, AMD's downward spiral will start this year itself (as Zen 5 is coming out only next year).

Time to acknowledge the technology gains Intel has accomplished is such a short span of time.

 

[jeremyj_83]

Intel has pulled all stops & is going to crush competition starting next year. Kindly deal with it. In fact, if AMD doesn't deliver a powerful Zen 4 refresh on TSMC N4 this year that can effectively compete with the measly Meteor Lake, AMD's downward spiral will start this year itself (as Zen 5 is coming out only next year).

Intel is only ahead in performance because they have thrown power consumption out the window. Anandtech looked at the 13900k and the 7950X at different power levels. In all but 2 tests the 13900k had to be at 125W to beat the 7950X at 65W (88W PPT). Once the 7950X was moved to 105W the 13900k needed to be at stock 250W levels to win. Now do note that the 13900k scales better than the 7950X as power levels increased.

In the data center Intel is still at least 3 years behind AMD. Sapphire Rapids (SPR) is only about as fast as Zen 3 based Milan when it comes to core performance. Intel is banking on customers using the hardware accelerators to make SPR be faster. However, those accelerators will require writing software to be aware of it so odds are this is planting the seed for servers CPUs 2 generations from now.

From what I remember Meteor Lake (mobile in late 2023), Arrow Lake is desktop in 2024, is supposed to be a bigger change on the P core than Raptor Lake was from Alder Lake (12th Gen). Raptor Lake was like a tick instead of a tock. Zen 5 is supposed to be a big redesign whereas Zen 4 was more like an optimization of Zen 3. On the desktop things will be interesting but Intel still needs to up their power and not give as many big cores because of power draw.

 

[bit_user]

it's not going to change from replacing an "nm" suffix with an "A" suffix.

In case you hadn't noticed, both TSMC and Intel both dropped the "nm". That's why it's unfortunate Intel decided to introduce the "A" and start yammering on about angstroms.

It's ok to be a fan of TSMC/AMD.

I just call BS where I see it. If TSMC were talking about angstroms, I'd say the exact same about them.

When AMD proudly announces that their products are coming out in TSMC N7, not a single fanboy questioned 7nm!?!?!

I think TSMC's N-naming was actually progress. Similar to Intel's "Intel 7" and "Intel 4". Both represented a step away from fictitious units.

Intel has pulled all stops & is going to crush competition starting next year. Kindly deal with it.

I "deal with" facts, not predictions. I'll applaud Intel, if and when that happens.

Time to acknowledge the technology gains Intel has accomplished is such a short span of time.

The gain I'm prepared to acknowledge is that Alder Lake and Raptor Lake were surprisingly strong.

I have also given Intel credit for their bold move into the realm of hybrid CPUs. That was a risky move and they've sure taken criticism for it, but I think it's paying off.

 

[The Historical Fidelity]

It's ok to be a fan of TSMC/AMD. But history is very important my friend! Your arguments need to be rational and well-balanced. You keep continuously pointing out that the Intel node naming scheme is ridiculous. Maybe. But you have NEVER spoken one bit about how ridiculous TSMC's node names are!!!

TSMC was the first to break the cycle almost 2 decades back with ridiculous node names/numbers. Intel has just been following suite all these years. When AMD proudly announces that their products are coming out in TSMC N7, not a single fanboy questioned 7nm!?!?! When AMD recently announced their Zen 4 is on TSMC N4, there was a huge fanboys celebration! No one spoke about the ridiculous 4nm claim!!!

But when intel says it's coming out with Arrow Lake on Intel 20A next year, the fanboys immediately have huge problems with Intel's node naming, performance, power efficiency, density, powervia, frequency, and how fake the node is in general. When in reality, it's far superior to TSMC N3 which nobody wants to mention. And the most typical silly argument of all, Intel will never deliver. When in fact, Intel already has 20A almost ready to go and TSMC 2nm is still in drawing board for a 2026 launch!!!!!!!!!!

Intel has pulled all stops & is going to crush competition starting next year. Kindly deal with it. In fact, if AMD doesn't deliver a powerful Zen 4 refresh on TSMC N4 this year that can effectively compete with the measly Meteor Lake, AMD's downward spiral will start this year itself (as Zen 5 is coming out only next year).

Time to acknowledge the technology gains Intel has accomplished is such a short span of time.

Yeah I’ll acknowledge that the Intel Fab division has their heads screwed on right, but Intel’s future roadmap is in shambles based on insider sources saying desktop meteorlake chips are cancelled and the desktop i9 will be the top of the range laptop 6P + 8E cpu with extended power limits, their first multi-core chiplet server chips are cancelled, and graphics division just said they are revising their product roadmap for releasing new gpu series from every year to every 2 years. And according to a very reliable insider, the design teams can no longer demand the foundry division change process node rules to fix lazy designs and oversights which are commonplace in previous Intel architectures. It will be interesting to see how the design teams adapt to their personal former foundry play toy now needing to maintain non-specific design rules to maintain compatibility with the foundry’s 3rd party customers. Maybe they will actually design chips properly so they don’t need custom design rules to fix their mistakes.

 

[The Historical Fidelity]

Angstroms are alive and well in the semiconductor industry, just as Rankine hangs around in rocketry like a bad smell.

The Rankine cycle is still a very relevant and accepted thermodynamics process…not an obsolete unit of measure no longer supported by all authorities in measures.

 

[SiliconFly]

Intel is only ahead in performance because they have thrown power consumption out the window. Anandtech looked at the 13900k and the 7950X at different power levels. In all but 2 tests the 13900k had to be at 125W to beat the 7950X at 65W (88W PPT). Once the 7950X was moved to 105W the 13900k needed to be at stock 250W levels to win. Now do note that the 13900k scales better than the 7950X as power levels increased.

In the data center Intel is still at least 3 years behind AMD. Sapphire Rapids (SPR) is only about as fast as Zen 3 based Milan when it comes to core performance. Intel is banking on customers using the hardware accelerators to make SPR be faster. However, those accelerators will require writing software to be aware of it so odds are this is planting the seed for servers CPUs 2 generations from now.

From what I remember Meteor Lake (mobile in late 2023), Arrow Lake is desktop in 2024, is supposed to be a bigger change on the P core than Raptor Lake was from Alder Lake (12th Gen). Raptor Lake was like a tick instead of a tock. Zen 5 is supposed to be a big redesign whereas Zen 4 was more like an optimization of Zen 3. On the desktop things will be interesting but Intel still needs to up their power and not give as many big cores because of power draw.

Intel's Arrow Lake is coming out on a brand new, revolutionary & superior 2nm node while AMD's Zen 5 is coming on an old & inferior TSMC 3nm node.

For the first time in the last 6 years, Intel now has a clear node advantage over AMD. And remember, Zen architecture isn't that great either. For example, Zen 4 was dead on arrival inspite of being on a far superior node. Zen is that bad guys compare to Intel's Core architecture. AMD won't even have E cores till 2024. That bad. Zen 4 under-performed so grossly compared to Raptor Lake during last December, AMD has to release 3D VCache models 3 months later to effectively compete. Pls stop crying.

Intel is getting a 1 + 1/2 node performance gain (a massive & unprecedented 1.5 node jump) for free with Arrow Lake. And coupled with other performance/power/architecture optimizations, Zen 5 won't stand a chance. Intel's server side sucks, but their client group is on a roll now. Like i said, Zen 5 will be dead on arrival. AMD has lost it's unfair node advantage over Intel already. Starting 2024, Intel will now have an unfair node advantage over AMD. It's a fact fellas. AMD is finished starting 2024 guys. Deal with it & move on.

We'll have a condolence meet for AMD in Dec 2024 in Santa Clara, CA at 7pm. BYOB.

 

[SiliconFly]

Yeah I’ll acknowledge that the Intel Fab division has their heads screwed on right, but Intel’s future roadmap is in shambles based on insider sources saying desktop meteorlake chips are cancelled and the desktop i9 will be the top of the range laptop 6P + 8E cpu with extended power limits, their first multi-core chiplet server chips are cancelled, and graphics division just said they are revising their product roadmap for releasing new gpu series from every year to every 2 years. And according to a very reliable insider, the design teams can no longer demand the foundry division change process node rules to fix lazy designs and oversights which are commonplace in previous Intel architectures. It will be interesting to see how the design teams adapt to their personal former foundry play toy now needing to maintain non-specific design rules to maintain compatibility with the foundry’s 3rd party customers. Maybe they will actually design chips properly so they don’t need custom design rules to fix their mistakes.

Intel's Server & AXG groups suck! No doubts there. But their client CPU group is currently rocking. Alder Lake was revolutionary & on time. Raptor Lake beat Zen 4 in performance and was released 1 month in advance. It took Zen 4 3D VCache models to beat Raptor Lake 3 months later. Meteor lake is perfectly on track for a Q3 release. Arrow Lake is progressing very very nicely without any hiccups. There's no stopping Intel on the client side.

AMD's road map is now in shambles as they're stuck with inferior TSMC nodes compared to far superior nodes that Intel has starting 2024. AMD had an unfair node advantage over Intel all these (5+) years. Intel now has that same unfair node advantage starting next year. Zen 5 next year will be dead on arrival like Zen 4. AMD future is now in question. AMD might die sooner than you think. My condolences.

 

[bit_user]

Why are you pumping Intel so hard? It's truly bizarre. It's as if you're trying to single-handedly drive up the stock price (doesn't work; don't try it).

their client CPU group is currently rocking.

Not really. The professional reviews of their i9-13900K are using water cooling, which most people don't have. And if you don't, then it ends up thermal-throttling.

When I get one, I'm going to have to set a lower power limit, because I'm only going to use air cooling and I don't want my CPU fan screaming.

 

[SiliconFly]

Why are you pumping Intel so hard? It's truly bizarre. It's as if you're trying to single-handedly drive up the stock price (doesn't work; don't try it).


Not really. The professional reviews of their i9-13900K are using water cooling, which most people don't have. And if you don't, then it ends up thermal-throttling.

When I get one, I'm going to have to set a lower power limit, because I'm only going to use air cooling and I don't want my CPU fan screaming.

You're pumping AMD so hard you seem to be invested heavily in it! Bizarre considering they're at the end of the road!

You make it sound like as thought supporting Intel is a carnal sin!!! To be honest, when AMD fanboys trash talk Intel, you don't care!!! You think it's okay!!!!!!!!! But when the reverse happens, you immediately have major issues. A clear cut case of double standards.

The world is changing fast. Intel CPUs aren't the crap they used to be. And more importantly, AMD isn't the saving grace anymore! In fact, AMD's downward spiral is starting soon because they're stuck with inferior nodes. AMD is finished starting 2024. Please deal with it. And don't let your double standards make things "truly bizarre " (in your own words).

 

[TerryLaze]

Why are you pumping Intel so hard? It's truly bizarre. It's as if you're trying to single-handedly drive up the stock price (doesn't work; don't try it).


Not really. The professional reviews of their i9-13900K are using water cooling, which most people don't have. And if you don't, then it ends up thermal-throttling.

When I get one, I'm going to have to set a lower power limit, because I'm only going to use air cooling and I don't want my CPU fan screaming.

Every modern CPU thermal throttles to give you the best possible performance, including ZEN or the Pi 400 were they just slapped a big piece of metal on a Pi to make it run as fast as possible until it throttles.
If your ZEN or intel or anything other doesn't thermal throttle then you are wasting electricity on unnecessary cooling, on making the CPU cooler than it has to be for no reason at all.
Unless like you propose you lower the power limits outright.

 

[bit_user]

You're pumping AMD so hard you seem to be invested heavily in it!

I'm not. Throwing the same allegation back at me is not a valid counter-argument.

Bizarre considering they're at the end of the road!

This is speculation. You presume to know the future, but we cannot.

You make it sound like as thought supporting Intel is a carnal sin!!!

Excessive support and "boosterism" is commonly referred to as fanboyism, and that's not helpful. These are technology products, not sports teams. This site is about sharing information and helping people out. When enthusiasm for a brand becomes counterproductive to that end, it becomes a problem.

To be honest, when AMD fanboys trash talk Intel, you don't care!!!

That's not accurate. I take Intel's side when I think it's warranted, but the way you're talking about Intel and AMD is overly partisan. It's not a 50/50 thing. The issue is when someone is being too excessive and you're really being excessive.

AMD is finished starting 2024. Please deal with it.

I already said that I deal with events once they've happened. If you tell me to "deal with it" again, I'll consider it harassment and it will be reported.

You can present information and state your predictions, but repeatedly harassing people who don't agree with them is not acceptable behavior.

 

[bit_user]

If your ZEN or intel or anything other doesn't thermal throttle then you are wasting electricity on unnecessary cooling, on making the CPU cooler than it has to be for no reason at all.

Thermal throttling is done to protect the CPU, not for energy-efficiency reasons. Power limiting is what people use, when they care about energy efficiency.

 

[jeremyj_83]

Every modern CPU thermal throttles to give you the best possible performance, including ZEN or the Pi 400 were they just slapped a big piece of metal on a Pi to make it run as fast as possible until it throttles.
If your ZEN or intel or anything other doesn't thermal throttle then you are wasting electricity on unnecessary cooling, on making the CPU cooler than it has to be for no reason at all.
Unless like you propose you lower the power limits outright.

In a desktop environment you do not want to be thermally throttling. If you are that means your cooling solution is not good enough for your usage. Not to mention that having your CPU hit the thermal ceiling all the time can lead to premature aging/death of the CPU.

Intel's Arrow Lake is coming out on a brand new, revolutionary & superior 2nm node while AMD's Zen 5 is coming on an old & inferior TSMC 3nm node.

Arrow Lake isn't coming out until end of 2024/beginning of 2025 at the earliest. That said even if Intel's node is superior that doesn't mean anything. Intel still cannot get their power draw under control on their P cores. Do you really think Intel would have gone to a BIG.little design if they could have gotten 16 P cores to clock at a decent level all while using lower power? Of course not they would have stayed with nothing but P cores. However, they knew how much power their P cores used and needed to reign in the absolute power requirements so the went hybrid. AMD on the other hand has been able to put 16 big cores all on the same die while keeping power under control. This is because Zen is more efficient that Core. We see this in benchmarks that limit the power (which I referenced earlier). Intel needs double the power draw to equal AMDs performance.

Intel's Server & AXG groups suck! No doubts there. But their client CPU group is currently rocking. Alder Lake was revolutionary & on time. Raptor Lake beat Zen 4 in performance and was released 1 month in advance. It took Zen 4 3D VCache models to beat Raptor Lake 3 months later. Meteor lake is perfectly on track for a Q3 release. Arrow Lake is progressing very very nicely without any hiccups. There's no stopping Intel on the client side.

Again Intel needed to throw power limits out the window in able to beat Zen 4. As I had said earlier Core does scale better with additional power than Zen. Once Zen got to 105W the scaling beyond that was minimal. Core on the other hand kept scaling pretty well all the way up to 253W. While that isn't a big deal on the desktop, that can be an issue in laptops. The Intel laptops need more power to keep up with Zen but that has the other problem of reducing battery life. Since Zen 2 was released, on similarily configured laptops AMD has had the battery life advantage over Intel. We also knew that Zen 4 3D was going to be released long before Zen 4 was released. AMD released roadmaps with Zen 4 3D on them over a year ago. One other thing don't forget that Intel made the 13900KS in an attempt to beat Zen4 3D in gaming.

Client computing isn't where companies like Intel and AMD make most of their money. Most of their money is made on data center products. Intel not doing well in data center products is a big reason why they posted a net loss of more than $600M in Q4 2022. Intel needs to strengthen their data center products to make the company financially stable. They cannot rely on client computing to make a profit. Without profits they cannot afford the R&D on new process nodes or CPU designs. Right now Intel is at least 2-3 years behind AMD in the data center market. If AMD keeps executing like they have, Intel will stay 2-3 years behind.

I want Intel to do well same as I want AMD to do well. Competition is the market is the best thing for everyone. It leads to innovation, lower prices, and better performance. The fact that you seem to want AMD to fail outright is a scary thought. Had Zen 1 not been a success, they bet the entire company on the Zen line and it was a bet that paid off, AMD would probably have gone bankrupt. Had that happened do you really think the i3's would be 4c/8t and the i7's 8c/16t? Of course not. We would have been stuck with i7's still being 4c/8t, maybe 6c/12t with little need to Intel to innovate beyond Skylake.

 

[TerryLaze]

Thermal throttling is done to protect the CPU, not for energy-efficiency reasons. Power limiting is what people use, when they care about energy efficiency.

How is that relevant to what I said?!
If your ZEN or intel or anything other doesn't thermal throttle then you are wasting electricity on unnecessary cooling, on making the CPU cooler than it has to be for no reason at all.
I didn't say that you would do this to save energy, just that you will if you do it.

In a desktop environment you do not want to be thermally throttling. If you are that means your cooling solution is not good enough for your usage.

That was true the last time almost ten years ago when CPUs still had ridgid turbo tables that wouldn't change, if your cooling couldn't handle these fixed clocks then yes you had an issue.
Now each core clocks to whatever limit the core itself reaches within the system, this is true for both intel and amd, the more you cool it the higher it will go, if it doesn't go any higher anymore and instead it decreases temps then you are just wasting cooling.

Not to mention that having your CPU hit the thermal ceiling all the time can lead to premature aging/death of the CPU.

The thermal ceiling is 5-10 degrees below the shut off temp which is again 5-10 degrees below the absolute maximum the cpu could withstand.
For example 12th gen will throttle at 100 degrees while the thermal trip is at 130...
https://www.intel.la/content/www/xl/es/products/docs/processors/core/core-technical-resources.html

Thermal Trip: The processor protects itself from catastrophic overheating by use of an internal thermal sensor. This sensor is set well above the normal operating temperature to ensure that there are no false trips. The processor will stop all executions when the junction temperature exceeds approximately 130 °C. This is signaled to the system by the THRMTRIP# pin.

 

[jeremyj_83]

How is that relevant to what I said?!
If your ZEN or intel or anything other doesn't thermal throttle then you are wasting electricity on unnecessary cooling, on making the CPU cooler than it has to be for no reason at all.
I didn't say that you would do this to save energy, just that you will if you do it.

That was true the last time almost ten years ago when CPUs still had ridgid turbo tables that wouldn't change, if your cooling couldn't handle these fixed clocks then yes you had an issue.
Now each core clocks to whatever limit the core itself reaches within the system, this is true for both intel and amd, the more you cool it the higher it will go, if it doesn't go any higher anymore and instead it decreases temps then you are just wasting cooling.

The thermal ceiling is 5-10 degrees below the shut off temp which is again 5-10 degrees below the absolute maximum the cpu could withstand.
For example 12th gen will throttle at 100 degrees while the thermal trip is at 130...
https://www.intel.la/content/www/xl/es/products/docs/processors/core/core-technical-resources.html

Even though the max temp for shutdown of the CPU is 130C and Tjunction (thermal throttling) is 100C, that doesn't mean you want to run your CPU at 100C all the time. No matter what that will shorten the lifespan of the CPU. Your cooling solution should be able to handle at least 33% more than the max wattage. For example on the 13900k max wattage is 253W at stock settings. That means you want a cooling solution that can handle 336W or more to keep your temps lower and not shorten your CPU lifespan. You are not wasting anything by getting a cooling solution that can bring your temps down.

 

[bit_user]

Arrow Lake isn't coming out until end of 2024/beginning of 2025 at the earliest. ...

FWIW, I think it doesn't make sense to get in a knock-down, drag-out fight about unreleased products. There's too much we simply don't know and can't prove. Worse, it's guaranteed to be unproductive.

That said even if Intel's node is superior that doesn't mean anything. Intel still cannot get their power draw under control on their P cores.

I think that subject can't be separated entirely from one about manufacturing nodes. Case in point: Ice Lake's Sunny Cove burned a lot less power on Intel's 10 nm+ node than when they back-ported it to their 14 nm++++ node.

Do you really think Intel would have gone to a BIG.little design if they could have gotten 16 P cores to clock at a decent level all while using lower power?

Not sure, but E-cores do make a lot of sense for highly-threaded workloads. They're quite simply a better way to scale, in terms of both power and area (i.e. cost). I support Intel on this point.

We also knew that Zen 4 3D was going to be released long before Zen 4 was released. AMD released roadmaps with Zen 4 3D on them over a year ago.

Yeah, you can't make a product like that without planning several years in advance. Zen 3 launched with the TSVs needed for the 3D V-Cache die, and Zen 4 likewise. I might've even read someone claiming they found those same TSVs on the Zen 2 die...

Client computing isn't where companies like Intel and AMD make most of their money. Most of their money is made on data center products.

Actually... Intel's latest results had $6.6B in Client Computing Group and $4.3B in Data Center & AI:

https://www.tomshardware.com/news/intel-posts-largest-loss-in-years-as-sales-of-pc-and-server-cpus-nosedive

Intel not doing well in data center products is a big reason why they posted a net loss of more than $600M in Q4 2022.

Even looking to the year prior, revenue and operating income were greater in client computing.

Right now Intel is at least 2-3 years behind AMD in the data center market.

I'd put Sapphire Rapids only 1 generation behind. It fares well against EPYC Milan (Zen 3).

Source:

Intel Xeon Platinum 8490H "Sapphire Rapids" Performance Benchmarks Review - Phoronix

www.phoronix.com

Had that happened do you really think the i3's would be 4c/8t and the i7's 8c/16t? Of course not. We would have been stuck with i7's still being 4c/8t, maybe 6c/12t

If you plot the core count of Intel's server CPUs, they were scaling on a nice exponential curve. If you plot the core count of their consumer CPUs, they flatlined all the way from Sandybridge to Kaby Lake. And Kaby Lake launched right after Zen 1. For 3 generations after that, Intel added 2 more cores to their top spec. That was no coincidence.

 

[bit_user]

If your ZEN or intel or anything other doesn't thermal throttle then you are wasting electricity on unnecessary cooling, on making the CPU cooler than it has to be for no reason at all.

Excess cooling is not necessarily wasted, depending on how much. I slows transistor aging and electromigration. These compromise the maximum boost clocks of newer CPUs, and outright stability of older ones.

The thermal ceiling is 5-10 degrees below the shut off temp which is again 5-10 degrees below the absolute maximum the cpu could withstand.
For example 12th gen will throttle at 100 degrees while the thermal trip is at 130...
https://www.intel.la/content/www/xl/es/products/docs/processors/core/core-technical-resources.html

If they didn't throttle until the thermal trip point, then the result would be CPUs wearing out prematurely - even before the warranty period expires! The trip point is just there to avoid instantaneous destruction of the CPU, in situations like: "Oh noes!!!111 my heatsink fell off!"

 

[SiliconFly]

Why are you pumping Intel so hard? It's truly bizarre. It's as if you're trying to single-handedly drive up the stock price (doesn't work; don't try it).


Not really. The professional reviews of their i9-13900K are using water cooling, which most people don't have. And if you don't, then it ends up thermal-throttling.

When I get one, I'm going to have to set a lower power limit, because I'm only going to use air cooling and I don't want my CPU fan screaming.

FWIW, I think it doesn't make sense to get in a knock-down, drag-out fight about unreleased products. There's too much we simply don't know and can't prove. Worse, it's guaranteed to be unproductive.


I think that subject can't be separated entirely from one about manufacturing nodes. Case in point: Ice Lake's Sunny Cove burned a lot less power on Intel's 10 nm+ node than when they back-ported it to their 14 nm++++ node.


Not sure, but E-cores do make a lot of sense for highly-threaded workloads. They're quite simply a better way to scale, in terms of both power and area (i.e. cost). I support Intel on this point.


Yeah, you can't make a product like that without planning several years in advance. Zen 3 launched with the TSVs needed for the 3D V-Cache die, and Zen 4 likewise. I might've even read someone claiming they found those same TSVs on the Zen 2 die...


Actually... Intel's latest results had $6.6B in Client Computing Group and $4.3B in Data Center & AI:

https://www.tomshardware.com/news/intel-posts-largest-loss-in-years-as-sales-of-pc-and-server-cpus-nosedive

Even looking to the year prior, revenue and operating income were greater in client computing.


I'd put Sapphire Rapids only 1 generation behind. It fares well against EPYC Milan (Zen 3).

Source:

Intel Xeon Platinum 8490H "Sapphire Rapids" Performance Benchmarks Review - Phoronix

www.phoronix.com

If you plot the core count of Intel's server CPUs, they were scaling on a nice exponential curve. If you plot the core count of their consumer CPUs, they flatlined all the way from Sandybridge to Kaby Lake. And Kaby Lake launched right after Zen 1. For 3 generations after that, Intel added 2 more cores to their top spec. That was no coincidence.

I agree with all your points except one. Sapphire Rapids is bad. Very bad considering how awesome EPYC 96C cpus are. Intel may win in powerpoint presentations & benchmarks. But in real-world use cases, having 96C instead of 56C in a server makes a LOT of sense! AMD for the win in server space. Intel's server cpus suck so hard!

 

[SiliconFly]

I'm not. Throwing the same allegation back at me is not a valid counter-argument.


This is speculation. You presume to know the future, but we cannot.


Excessive support and "boosterism" is commonly referred to as fanboyism, and that's not helpful. These are technology products, not sports teams. This site is about sharing information and helping people out. When enthusiasm for a brand becomes counterproductive to that end, it becomes a problem.


That's not accurate. I take Intel's side when I think it's warranted, but the way you're talking about Intel and AMD is overly partisan. It's not a 50/50 thing. The issue is when someone is being too excessive and you're really being excessive.


I already said that I deal with events once they've happened. If you tell me to "deal with it" again, I'll consider it harassment and it will be reported.

You can present information and state your predictions, but repeatedly harassing people who don't agree with them is not acceptable behavior.

To be honest, you're the one who started it! And I have every right to report you too. If you have issues with my predictions, you should learn to ignore. Thats a mark of a gentleman, instead of engaging in personal attacks and saying hurting stuff like whether I'm invested in Intel!!!!!!!