Discussion Intel current and future Lakes & Rapids thread

[uzzi38]

@msj10 There's a pretty significant difference between the kind of clock rates Apple and Intel target.

Making a core hit 4GHz and making a core hit 5.5GHz is two totally different ballparks.

 

[uzzi38]

@DrMrLordX (they're not, most of the compute tiles are N3B)

Only one of them isn't.

 

[mikk]

Sorry to repeat myself (for those that have read what I've written elsewhere), but it's been reported multiple times that Intel won't be taking any N3B wafers at all. Arrow Lake compute tiles are likely to be on 20a only.

Reported multiple times lol. This FUD. Keep repeating your FUD.

 

[nicalandia]

Intel 5th Gen Xeon Platinum 8592+ "Emerald Rapids" CPU Leak: 64 Cores, 448 MB Cache, Up To 922W TDP

Intel's top 5th Gen Emerald Rapids chip, the Xeon Platinum 8592+, with 64 cores and 448 MB cache has been leaked by YuuKi_AnS.

wccftech.com

so 128 Core system pulling near 1000 Watts. A single Bergamo pulls One Third of that.

 

[Goop_reformed]

I'm not new to AT if thats what you mean. And it's not foolish. You keep obsessing over that slide. But I'm moving on.

You are running away because what you said was utter non sense. That stuff belongs to wccftech comment section.

 

[Goop_reformed]

Intel 5th Gen Xeon Platinum 8592+ "Emerald Rapids" CPU Leak: 64 Cores, 448 MB Cache, Up To 922W TDP

Intel's top 5th Gen Emerald Rapids chip, the Xeon Platinum 8592+, with 64 cores and 448 MB cache has been leaked by YuuKi_AnS.

wccftech.com

so 128 Core system pulling near 1000 Watts. A single Bergamo pulls One Third of that.

1/3 is a little bit too much. It's more like half.

AMD EPYC 9754 Benchmarks For The 128-Core Bergamo Review - Phoronix

www.phoronix.com

350-400W vs 720W ++

 

[SiliconFly]

You are running away because what you said was utter non sense. That stuff belongs to wccftech comment section.

And I thought the biggest nonsense ever was uttered by you. No wonder you like wccftech comments section.

 

[Goop_reformed]

Of course the final product could perform differently, they're called projections for a reason. They're the targets Intel have set for themselves to hit effectively. They could exceed or fall behind those targets when it comes to the real deal, but thats something we'll only find out when retail silicon comes to market (or is close to doing so).

And I thought the biggest nonsense ever was uttered by you. No wonder you like wccftech comments section.

Yeah this is not running away at all. At least own up to it like a man.

 

[DrMrLordX]

Reported multiple times lol. This FUD. Keep repeating your FUD.

Apple Orders Entire Supply of TSMC's 3nm Chips for iPhone 15 Pro and M3 Macs

Apple has reportedly secured all available orders for N3, TSMC's first-generation 3-nanometer process that is likely to be used in the upcoming...

www.macrumors.com

Apple devices to occupy 90% of TSMC 3nm capacity in 2023

TSMC will see the baseline 3nm process technology (N3B) dominate 90% of its N3 family capacity in 2023, and nearly 90% of the N3B capacity reportedly has been booked by Apple for new iPhones, MacBooks, and iPads for the year, which will bolster the pure-play foundry's revenue growth in the...

www.digitimes.com

MacRumors says 100%, DigiTimes says 90% (with no indication of who else is getting N3B). Multiple Apple enthusiast sites are repeating this information. If they're wrong they're wrong, but . . .

It really doesn't look like Intel is taking N3B.

 

[dullard]

Apple Orders Entire Supply of TSMC's 3nm Chips for iPhone 15 Pro and M3 Macs

Apple has reportedly secured all available orders for N3, TSMC's first-generation 3-nanometer process that is likely to be used in the upcoming...

www.macrumors.com

Apple devices to occupy 90% of TSMC 3nm capacity in 2023

TSMC will see the baseline 3nm process technology (N3B) dominate 90% of its N3 family capacity in 2023, and nearly 90% of the N3B capacity reportedly has been booked by Apple for new iPhones, MacBooks, and iPads for the year, which will bolster the pure-play foundry's revenue growth in the...

www.digitimes.com

MacRumors says 100%, DigiTimes says 90% (with no indication of who else is getting N3B). Multiple Apple enthusiast sites are repeating this information. If they're wrong they're wrong, but . . .

It really doesn't look like Intel is taking N3B.

Just to be clear with those articles:

1) Digitimes says Apple gets ~90% of N3B and that N3B is ~90% of all N3. So, Apple gets ~90% * ~90% = ~81%. That leaves plenty of leftover capacity for low volume and small size Intel tiles.

2) The MacRumors article simply refers to that Digitimes article, so you are still basing everything on one rumor article. But, that said, MacRumors says "Apple has procured 100% of the initial N3 supply". Notice the bolded words, that still leaves capacity after Apple gets their initial batch.

You seem to be making conclusions that go beyond your sources. I'm not saying Intel gets that remaining capacity. I don't really know. But, Intel itself claims to use N3 (without specifying which version).

 

[jpiniero]

You seem to be making conclusions that go beyond your sources. I'm not saying Intel gets that remaining capacity. I don't really know. But, Intel itself claims to use N3 (without specifying which version).

That was the plan. It's possible they cancel it and just roll with Raptor Lake some more.

 

[DAPUNISHER]

@SiliconFly @Goop_reformed

Conduct yourselves with better decorum please. Put each other on ignore if necessary.

- CPU Mod DAPUNISHER

 

[DrMrLordX]

Just to be clear with those articles:

1) Digitimes says Apple gets ~90% of N3B and that N3B is ~90% of all N3. So, Apple gets ~90% * ~90% = ~81%. That leaves plenty of leftover capacity for low volume and small size Intel tiles.

And MacWorld says Apple gets 100% of N3B so . . .

@jpiniero

Possible. Or they could just lean more-heavily on 20a.

Again if we see N3B showing up in Intel chips then the leaks were wrong and Intel actually took some N3B wafers anyway. It's impossible to entirely dismiss the possibility that Intel changed their plans in flight yet again.

 

[dullard]

And MacWorld says Apple gets 100% of N3B so . . .

Again if we see N3B showing up in Intel chips then the leaks were wrong and Intel actually took some N3B wafers anyway. It's impossible to entirely dismiss the possibility that Intel changed their plans in flight yet again.

MacWorld also said Apple may soon switch to N3E, so ...

None of the rumors you posted are incompatible with Intel getting some N3B. Yes, Apple will get most of N3B. But no rumor you have posted has said that Apple will both (1) get all of N3B for long and (2) will stay with N3B for long.

 

[DrMrLordX]

MacWorld also said Apple may soon switch to N3E, so ...

How is that relevant? It's not like Intel is going to start taking N3B wafers after Apple switches to N3E when Intel is specifically skipping over N3B for N3E.

 

[dullard]

How is that relevant? It's not like Intel is going to start taking N3B wafers after Apple switches to N3E when Intel is specifically skipping over N3B for N3E.

You said that Intel couldn't use N3B because Apple is using it according to MacWorld. But then also according to MacWorld, Apple is leaving N3B. Logically then your statement is false if MacWorld is correct--Intel could use N3B.

I'm not saying that it is wrong that Intel may use N3E. It is just that your logic you are using is wrong.

 

[nicalandia]

Yuuki Ans leaking Intel Alder Lake Refresh HEDT Xeon 2500 Series.

https://twitter.com/x/status/1720386208440959452

 

[Goop_reformed]

Uh oh it doesn't seem promising for meteor and upcoming lakes

Nova Lake with LLC aka v-cache coming in 1H 2027. That is a whopping 6 years after zen3x3d

 

[Geddagod]

Uh oh it doesn't seem promising for meteor and upcoming lakes

Nova Lake with LLC aka v-cache coming in 1H 2027. That is a whopping 6 years after zen3x3d

The only relevant upcoming lakes is ARL until 2026 anyway, so its not that big of a deal lol.
(Edit: as in big deal, I mean its not a big deal that they wont have a v-cache competitor until 2027 because I doubt they would be very competitive even if they did lol)
Zen 5 is going to have 3d cache while ARL doesn't, and Zen 6 is going to have 3d cache while ARL-R doesn't. So it's not exactly news that they wouldn't have anything like 3d cache (or actually anything too competitive)until they start competing with Zen 7.
Speaking of Zen 7, if Zen 5 is coming 1H 2024, Zen 6 is likely to be 2H 2025, and Zen 7 in 1H 2027. So if NVL? 3D cache variants are coming 1H 2027 (while NVL vanilla comes in 2H 2026)... it's not too bad.

That's just for DIY desktop tho, I have different opinions for server.
And ye ik this is a super late reply lol, but I swear I've gotten ebola or something cuz I have been dying for the past weekend xD

 

[Tigerick]

The only relevant upcoming lakes is ARL until 2026 anyway, so its not that big of a deal lol.
(Edit: as in big deal, I mean its not a big deal that they wont have a v-cache competitor until 2027 because I doubt they would be very competitive even if they did lol)
Zen 5 is going to have 3d cache while ARL doesn't, and Zen 6 is going to have 3d cache while ARL-R doesn't. So it's not exactly news that they wouldn't have anything like 3d cache (or actually anything too competitive)until they start competing with Zen 7.
Speaking of Zen 7, if Zen 5 is coming 1H 2024, Zen 6 is likely to be 2H 2025, and Zen 7 in 1H 2027. So if NVL? 3D cache variants are coming 1H 2027 (while NVL vanilla comes in 2H 2026)... it's not too bad.

That's just for DIY desktop tho, I have different opinions for server.
And ye ik this is a super late reply lol, but I swear I've gotten ebola or something cuz I have been dying for the past weekend xD

Zen6 would most likely to be last CPU for AM5 motherboards. Zen 7 will likely get new socket and support DDR6, thus I am expecting AMD will try to launch Zen 7 in H2 2026 to compete with Apple and Intel. That's why rumors said that Apple, Intel, nVidia and AMD are among the first to get N2 process from TSMC.

 

[Geddagod]

Apparently SRF-SP is now sampling

 

[cebri1]

https://www.tomshardware.com/news/intel-unreleased-emerald-rapids-cpu-benchmarks-48-core-chips-big-gains-sapphire-rapids-predecessors?utm_campaign=socialflow&utm_source=twitter.com&utm_medium=social

 

[SiliconFly]

Hilarious!

Link: Intel - Core Truths

Snake Oil... Used car salesman...

This is a new low! (after the rear-view mirror thingy)

 

[Hitman928]

Hilarious!

Link: Intel - Core Truths

Snake Oil... Used car salesman...

This is a new low! (after the rear-view mirror thingy)

While they do kind of have a point, I doubt they want anyone calling them out on their own renaming shenanigans (what is most of their 14th gen again ) . I also can't imagine the OEMs (e.g., HP and Dell) really like this type of advertising from Intel since they are the ones who push for this kind of marketing to begin with. . .

With that said, marketing will do marketing things, lol.

 

[Saylick]

Hilarious!

Link: Intel - Core Truths

Snake Oil... Used car salesman...

This is a new low! (after the rear-view mirror thingy)

Wow, this is just... desperate from Intel. I could make fun of Intel for similar things they've done in the past (*cough* rehashing 10nm 10nm ESH Intel 7 for how many generations?? *cough*) but it's all in the rear view mirror, right Pat?