Intel is relaunching the Core i9-14900K... wait, it's the 6.2GHz Core i9-14900KS

[HardReset]

Most upcoming Arrow Lake CPUs will leverage TSMC nodes for compute, Intel 20A only for Core i5 and lower: Leak

This leak raises some questions.

www.tomshardware.com

You know, you can use Google too...

And regardless of which node they use, Intel 20a and TSMC 3 are pretty much the same anyways... the point that Intel won't be using an inferior node than AMD is valid.

That's just another "leak" that may or may not be true.

TSMC has already produced 3nm chips that are on sale. So far Intel 20A is nothing else than promises how it will be superior (we remember 10nm fiasco forever). So Intel may well have inferior node.

Problem with Intel is that they won't easily admit that TSMC makes better chips so don't expect top model to use TSMC.

 

[Squid Surprise]

That's just another "leak" that may or may not be true.

TSMC has already produced 3nm chips that are on sale. So far Intel 20A is nothing else than promises how it will be superior (we remember 10nm fiasco forever). So Intel may well have inferior node.

Problem with Intel is that they won't easily admit that TSMC makes better chips so don't expect top model to use TSMC.

Well, obviously Intel 20A hasn’t been seen yet - Arrowlake isn’t out yet! The point is that it will no longer be intel @ 14nm vs AMD @ 7…. The nodes will either be identical or at the worst, much closer.

AMD might still remain superior - but they will now have to rely on their architecture, not their node for that superiority.

 

[HardReset]

Well, obviously Intel 20A hasn’t been seen yet - Arrowlake isn’t out yet! The point is that it will no longer be intel @ 14nm vs AMD @ 7…. The nodes will either be identical or at the worst, much closer.

AMD might still remain superior - but they will now have to rely on their architecture, not their node for that superiority.

Again, we already know what TSMC 3nm is, since there are actual products launched. On other hand, we have no real information about Intel 20A, only promises and we all know what that means when talking about Intel's recent processes. At worst it will be another 10nm disaster. Another thing is that there is still no real information about clock speeds 20A process allows. Without clock speeds at least 5.5 GHz with decent power consumption, AMD has no problems staying ahead.

All in all, right now there is nothing concrete that says AMD no longer have node superiority when Arrow Lake is out.

 

[Squid Surprise]

All in all, right now there is nothing concrete that says AMD no longer have node superiority when Arrow Lake is out.

Yes - but there’s nothing concrete that says otherwise… why assume AMD maintain node superiority? And even if they DO have a slight node superiority this generation, it won’t be the same gap as last time - and it won’t be getting larger for the next gen…

Eventually, AMD will need to maintain superiority based on something other than their processing node…whether that’s late this year, next year, or in 2-3 years - that time WILL come.

Better for AMD to be prepared now.

 

[HardReset]

Yes - but there’s nothing concrete that says otherwise… why assume AMD maintain node superiority? And even if they DO have a slight node superiority this generation, it won’t be the same gap as last time - and it won’t be getting larger for the next gen…

Eventually, AMD will need to maintain superiority based on something other than their processing node…whether that’s late this year, next year, or in 2-3 years - that time WILL come.

Better for AMD to be prepared now.

Nothing says otherwise? We have:

- TSMC 3nm, up and running, making actual products, no major concerns.
- Intel 20A, non-proven, no existing products, nothing but Intel promises.

Additionally, Intel is actually using TSMC for their own products. What that means? Even Intel do not trust their own process.

Even so simple thing that 20A is not able to reach as high clock speeds (not known right now) as Intel 7 is more than enough for AMD to Extend their node advantage even more.

 

[Squid Surprise]

Nothing says otherwise? We have:

- TSMC 3nm, up and running, making actual products, no major concerns.
- Intel 20A, non-proven, no existing products, nothing but Intel promises.

Additionally, Intel is actually using TSMC for their own products. What that means? Even Intel do not trust their own process.

Even so simple thing that 20A is not able to reach as high clock speeds (not known right now) as Intel 7 is more than enough for AMD to Extend their node advantage even more.

That shows NOTHING . It MIGHT mean something - but not until we actually see it…

I understand you hate intel and must support AMD for EVERYTHING but…. Maybe calm down for a bit and actually wait until the product is actually released?

 

[HardReset]

That shows NOTHING . It MIGHT mean something - but not until we actually see it…

I understand you hate intel and must support AMD for EVERYTHING but…. Maybe calm down for a bit and actually wait until the product is actually released?

Exactly. TSMC 3nm is already released. Intel 20A, well, nothing but Gelsinger promises.

When was last time Intel delivered node that actually was something at least remotely good? That was 6.5 years ago. Based on that we should expect that 20A trounces TSMC 3nm?

So in fact YOU are one that should calm down a bit. Wait until we have at least Something Actual on Intel 20A.

 

[Squid Surprise]

Exactly. TSMC 3nm is already released. Intel 20A, well, nothing but Gelsinger promises.

When was last time Intel delivered node that actually was something at least remotely good? That was 6.5 years ago. Based on that we should expect that 20A trounces TSMC 3nm?

So in fact YOU are one that should calm down a bit. Wait until we have at least Something Actual on Intel 20A.

So you are basing this all on… nothing… fair enough… and Intel WILL be using TSMC’s node for at least their GPU, if not the CPU as well - not because they don’t trust their own fabs, but simply because they can’t produce enough (despite being “inferior” to AMD, they outsell them by a wide margin). Again, wait and see, but don’t just assume that AMD will have the superior node.

By the way, I’m not just an intel shill - my main PC is an AMD !

But my next PC will be whatever is better at the time - and there’s no reason to think it will be AMD forever. Intel has more money, and usually, that’s who wins in the long run .

 

[HardReset]

So you are basing this all on… nothing… fair enough… and Intel WILL be using TSMC’s node for at least their GPU, if not the CPU as well - not because they don’t trust their own fabs, but simply because they can’t produce enough (despite being “inferior” to AMD, they outsell them by a wide margin). Again, wait and see, but don’t just assume that AMD will have the superior node.

That's funny. Intel didn't outsource CPU production when 10nm was 3.5 years late and they really had capacity problems. And now when they shouldn't actually have production problems, they outsource. Does not compute. They just don't trust their own nodes enough, that's only explanation that makes at least some sense.

Again, TSMC 3nm is already up and producing actual products. Intel 20A is still nothing else than imaginary process that Perhaps will come Someday and Maybe will be something. Comparing TSMC 3nm against Intel 20A is comparing something that exists against something that only exist on Gelsinger dreams.

But my next PC will be whatever is better at the time - and there’s no reason to think it will be AMD forever. Intel has more money, and usually, that’s who wins in the long run .

Intel had more money than TSMC had but that didn't help too much.

 

[ddferrari]

So, spend and extra 100 bucks for a .2Ghz clock speed increase... or spend ten minutes overclocking the 14900K and get the exact same specs, if not higher?

Tough choice here.

 

[Disiw]

Arrow Lake will use 3nm TSMC for high-end chips, K models. Google exist.

Hard Reset no I won't be buying Zen 5 if Arrow Lake is superior. Luckily I can wait for both and watch 1:1 comparisons before I decide.

If AMD is still competitive without having node advantage, I will be amazed.

Afterall they did not beat Intel on performance while having node advantage.

 

[Disiw]

So, spend and extra 100 bucks for a .2Ghz clock speed increase... or spend ten minutes overclocking the 14900K and get the exact same specs, if not higher?

Tough choice here.

14900KS is better binned, you will not reach same speeds with a 14900K, which is why KS model took records on day one.

13900KS = 14900K pretty much.

 

[HardReset]

Arrow Lake will use 3nm TSMC for high-end chips, K models. Google exist.

Hard Reset no I won't be buying Zen 5 if Arrow Lake is superior. Luckily I can wait for both and watch 1:1 comparisons before I decide.

If AMD is still competitive without having node advantage, I will be amazed.

Afterall they did not beat Intel on performance while having node advantage.

FYI, Arrow Lake will be MCM. That means there may be different techs used on different chips.

Information available today says CPU is Intel, GPU is TSMC. So, how about a link?

AMD Threadripper 64 core consumes around same power as 14900K.

 

[erickmendes]

My 5800x3D give me 80% gaming performance at 1440p using 35w... My air cooler barely feel it...

What type of cooling do you need for this monster by the way? 380mm AIO?

I bet 240 can't but this 260 can do the job. 260 cubic miles btw.

 

[Disiw]

My 5800x3D give me 80% gaming performance at 1440p using 35w... My air cooler barely feel it...



I bet 240 can't but this 260 can do the job. 260 cubic miles btw.

View attachment 89505

Your 5800X3D is nowhere near in CPU bound gaming and in other workloads you will be absolutely crushed.

This is actually AMDs big problem with 3D chips. Good for gaming but lacks outside of gaming.

The regular chips are good outside of gaming but lacking for gaming.

Intel has it all in one chip and with Arrow Lake and 3nm TSMC, watt issues will be gone.

Lets just hope AMD will deliver decent performance with Zen 5 even tho they skimped on process node ... and hopefully 3D parts will be clocked higher.

 

[HardReset]

Intel has it all in one chip and with Arrow Lake and 3nm TSMC, watt issues will be gone.

How about finally giving source or stop putting this BS everywhere?

 

[Squid Surprise]

How about finally giving source or stop putting this BS everywhere?

It isn’t BS - we won’t know either way for weeks (or months) but there is NOTHING that disproves it available.
Yes, we all know you’re an AMD shill but no one is arguing that AMD doesn’t have a great CPU!

 

[HardReset]

It isn’t BS - we won’t know either way for weeks (or months) but there is NOTHING that disproves it available.
Yes, we all know you’re an AMD shill but no one is arguing that AMD doesn’t have a great CPU!

Saying Arrow Lake CPU is TSMC 3nm over and over again while there is no single source. I call that one BS as multiple sources say otherwise.

 

[Squid Surprise]

Saying Arrow Lake CPU is TSMC 3nm over and over again while there is no single source. I call that one BS as multiple sources say otherwise.

We already went over this earlier - there are ONLY rumours at this point… some sources say it might be TSMC for the GPU only, some say the whole thing… it’s NOT BS!

And regardless, the node will be far closer to AMD with Arrow Lake (if not identical) than it has been for many years.

 

[HardReset]

We already went over this earlier - there are ONLY rumours at this point… some sources say it might be TSMC for the GPU only, some say the whole thing… it’s NOT BS!

1. So far I haven't seen single source that says CPU tile is TSMC 3 nm.

2. If that is rumor, better not to say it as a fact but speculation. Especially after pointed out many times.

So once again: provide source for Arrow Lake CPU tile being TSMC 3nm or just admit it's all speculation at this point, not a fact.

 

[Squid Surprise]

1. So far I haven't seen single source that says CPU tile is TSMC 3 nm.

2. If that is rumor, better not to say it as a fact but speculation. Especially after pointed out many times.

So once again: provide source for Arrow Lake CPU tile being TSMC 3nm or just admit it's all speculation at this point, not a fact.

I think you suffer from a reading comprehension problem… there is no evidence for ANYTHING - it’s ALL speculation!

But that’s not BS!!

You don’t know if it’s TSMC or Intel fab… no one else knows either (or they aren’t sharing that info yet).

Once again - the difference between new Intel fab and new TSMC fab will be far narrower than previous AMD/Intel node differences…

So AMD will need to rely on something other than node to maintain superiority.

 

[HardReset]

I think you suffer from a reading comprehension problem… there is no evidence for ANYTHING - it’s ALL speculation!

But that’s not BS!!

Speculating something, not offering any valid arguments for claim and saying it as a fact despite multiple sources saying else is not speculation, but trolling.

Once again, provide single source from 2024 that Speculates (yeah, Speculates) Arrow Lake CPU tile will use TSMC 3nm.

Shouldn't be too hard?

 

[Squid Surprise]

Speculating something, not offering any valid arguments for claim and saying it as a fact despite multiple sources saying else is not speculation, but trolling.

Once again, provide single source from 2024 that Speculates (yeah, Speculates) Arrow Lake CPU tile will use TSMC 3nm.

Shouldn't be too hard?

I already told you it WASN’T fact - there aren’t any to be had. You don’t have any facts either…
Stop repeating the same thing over and over…

 

[HardReset]

I already told you it WASN’T fact - there aren’t any to be had. You don’t have any facts either…
Stop repeating the same thing over and over…

Perhaps you should finally undetstand difference between following:

1. Trolling
2. Making claim with source
3. Making claim with argument

In this case.

1. Saying Arrow Lake CPU will use TSMC 3nm without source or valid argument is trolling.

2. Saying Arrow Lake will contain some TSMC 3nm parts with source goes to "claim with source" category.

3. Speculating that AMD no longer have node advantage because (arguments here) is speculation with arguments.

2 and 3 are OK. 1 is not. Especially since there are Ultra Strong counter argument. Intel has made their own CPUs for decades. Now Intel has also launched IFS. Intel could say "we don't have capacity" etc and use TSMC for GPU tile. But for CPU tile? If even Intel doesn't believe on Intel CPU process, who else will? Intel admits their CPU process sucks, use TSMC instead. Good luck for stock price and IFS.

 

[erickmendes]

Ok let me set this for you guys.

1: "Not every speculation is BS"
2: "Every BS lacks arguments"

When proposition 1 states that "Not every speculation is BS" and proposition 2 states that "Every BS lacks arguments," it implies a distinction between speculation and BS, while also characterizing BS as lacking arguments.

1. "Not every speculation is BS" implies that there are speculations that do not fall under the category of BS.
2. "Every BS lacks arguments" implies that all instances of BS are devoid of arguments.

These statements suggest that there are speculations that are not considered BS, and that BS itself is characterized by a lack of supporting arguments.

In formal logic terms, this indicates a relationship where speculation and BS are not equivalent, with BS being defined by its lack of arguments.

You are welcome.



Now continue please.