Discussion Comet Lake Intel's new Core i9-10900K runs at over 90C, even with liquid cooling TweakTown

[Makaveli]

Not really. We're only talking about high-performance desktops, mostly DIY.

That is the market I'm talking about most of us on this site not the general public.

 

[Rigg]

Agreed, though I think TECs may actually be a way to fix the problem through y-axis cooling. Not your bog standard eBay TEC, but actually a set of tiny TECs composed of only a few thermocouples each set up in tile configuration and mounted on top of each die. AMD CPUs at least have a large network of temp sensors on each die which allows them to pinpoint hotspots (and accurately report those temps to monitoring software). All you would have to do is cycle current towards the tile on top of the hotspot to aid in rapid dissipating of heat from that locality. It would shift the hotspot away from the die and towards the IHS. TECs are usually hideously inefficient, but I suspect the power/mm2 will be low enough at individual hotspots that activating a few thermocouples in the same area wouldn't chew up that much power. It's something Intel should probably look at since I think their R&D capabilities are potentially much stronger than AMD's (at least from a budget perspective). They don't need it for 14nm, but we haven't seen their 7nm process yet.

AMD has filed a patent for cooling a 3D stacked memory with thermo-electric coolers.

AMD Files a Patent for Cooling of 3D Stacked Memory

Scaling and manufacturing of ever shrinking semiconductor devices is becoming more challenging as smaller nodes are introduced. As we have approached 7 nanometers, economies of scale are becoming more influential than scales of manufacturing. For example, the development of the 7 nm node...

www.techpowerup.com

 

[piokos]

That is the market I'm talking about most of us on this site not the general public.

Sure. But you're decorating this with a general "Intel is in trouble".
DIY is not a priority niche for Intel - hasn't been for a while.

They're focusing on mobile designs - also giving them the early 10nm supply - simply because that's where investing makes more sense.
But now, with AMD finally launching 7N mobile chips, Intel would gladly give away all their DIY sales just to defend their mobile market dominance.
And they probably will. At least until they can afford to make desktop chips on the cutting edge node.

Keep in mind they're doing strategic and financial planning for the next 10 years or more. Losing the DIY segment for few years is obviously a blow, but not the end of the world. Especially if they can make more money elsewhere.

 

[Markfw]

Sure. But you're decorating this with a general "Intel is in trouble".
DIY is not a priority niche for Intel - hasn't been for a while.

They're focusing on mobile designs - also giving them the early 10nm supply - simply because that's where investing makes more sense.
But now, with AMD finally launching 7N mobile chips, Intel would gladly give away all their DIY sales just to defend their mobile market dominance.
And they probably will. At least until they can afford to make desktop chips on the cutting edge node.

Keep in mind they're doing strategic and financial planning for the next 10 years or more. Losing the DIY segment for few years is obviously a blow, but not the end of the world. Especially if they can make more money elsewhere.

I see it this way:
1) Desktop - in every segment but high FPS gaming, AMD has price/performance and performance/watt locked up (as far as technical dominance.)
2) HEDT - AMD wins this one in the same metrics as number 1.
3) Server - AMD wins this one in the same metrics as number 1.
4) Laptop - With Renoir, AMD competes with Intel 10 nm.

So the ONLY thing Intel has a clear lead in, is high FPS gaming, the 9900k. This chip simply ups that, except its even hotter and more power hungry, and 10 cores does not buy you anything in todays games as far as high FPS gaming, and since its harder to cool, and more power hungry, the 9900k might even beat it. (not withstanding upcoming reviews, so with a grain of salt). And BTW, high FPS gaming IS A NICHE

Where am I wrong ?

Now this does not mean that I think Intel is in trouble as a company, financially, it means that AMD will continue to eat away at the market dominance and if that continues too long, then it could hurt Intel. Right now its a wait and see. Marketing or not, you can;t continue "winning" sales based solely on marketing, at some point you have to be competitive.

 

[DrMrLordX]

@Rigg

Seems like someone's already on it then.

Sure. But you're decorating this with a general "Intel is in trouble".

If AMD can offer an R5 4600 @ $199 with an all-core turbo of 4.3 GHz and a 65W TDP, Intel really is in trouble.

 

[aigomorla]

Where am I wrong ?

The fact most OEM's in enterprise sector has secured contracts with Intel, and can offload 5,000 dollar cpu's like hotcakes in bulk quantities.
For every ryzen sold, OEM's have bought 1000 (i3 - i7)
For every EYPC cpu sold, there is probably 100 Xeon which went to OEM vendors.

So you are wrong in your assessment that intel is in trouble.

Here is intel's Sales:

Top Companies for | Find the Company - Vault

Here are your search results. We found a number of top companies offering the types of criteria you specified. Find the company that's right for you.

www.vault.com

Intel sells its products primarily to original equipment manufacturers (OEMs) and original design manufacturers (ODMs). In addition, Intel products are sold to makers of industrial and communications equipment.

Its customers also include those who buy PC components and other products through distributor, reseller, retail, and OEM channels. Intel's worldwide reseller sales channel consists of thousands of indirect customers, who are systems builders that purchase microprocessors and other products from distributors.

Intel's three largest customers account for nearly 40% of revenue. They are Dell Technologies, more than 15% of sales, and Lenovo Group and HP Inc., each with about 10%.

So mark unless we start seeing a mass exodus in OEM's going over to Ryzen / Threadripper / EYPC which i do not yet, even supermicro has just finally released a EYPC board, your statement is a bit over the top my friend.

 

[Markfw]

The fact most OEM's in enterprise sector has secured contracts with Intel, and can offload 5,000 dollar cpu's like hotcakes in bulk quantities.
For every ryzen sold, OEM's have bought 1000 (i3 - i7)
For every EYPC cpu sold, there is probably 100 Xeon which went to OEM vendors.

So you are wrong in your assessment that intel is in trouble.

Here is intel's Sales:

Top Companies for | Find the Company - Vault

Here are your search results. We found a number of top companies offering the types of criteria you specified. Find the company that's right for you.

www.vault.com

So mark unless we start seeing a mass exodus in OEM's going over to Ryzen / Threadripper / EYPC which i do not yet, even supermicro has just finally released a EYPC board, your statement is a bit over the top my friend.

I said nothing about Intel being in trouble. See my last sentence. I was saying technical dominance. And I said only if this keeps up, may it change that picture.

"Now this does not mean that I think Intel is in trouble as a company, financially, it means that AMD will continue to eat away at the market dominance and if that continues too long, then it could hurt Intel. Right now its a wait and see. Marketing or not, you can;t continue "winning" sales based solely on marketing, at some point you have to be competitive. "

Edit: and I bought a supermicro EPYC board a year ago.

Edit2 Here is a review from March 2018, 2 years ago: https://www.servethehome.com/supermicro-h11dsi-nt-review-dual-amd-epyc-motherboard/

 

[piokos]

Where am I wrong ?

Here:

it means that AMD will continue to eat away at the market dominance and if that continues too long, then it could hurt Intel. Right now its a wait and see. Marketing or not, you can;t continue "winning" sales based solely on marketing, at some point you have to be competitive.

This is the same issue that makes some people struggle to understand why Intel continues to grow despite AMD offering better value.

You compare CPUs based on product specs and price - like a reviewer, not like a financial analyst. You compare a representative of an Intel CPU and a representative of AMD CPU. You don't think about the economy behind it.
These companies aren't competing with infinite supply of CPUs.

Yes, AMD offers better value, wins reviews, has great press. But at this moment AMD can only make enough chips to cover ~5% of server and ~20% of consumer market.
And that's what Intel adjusts their offer for.
They can ask a price that's higher than AMD's - as long as clients are willing to pay. Because if Intel asked 2-3x more than AMD, some consumers would say: "nah, I can use my PC for a bit longer".
(Mind you: enterprises seldom have the luxury of being able to wait, so Intel actually asks 2-3x more for server chips.)

From Intel's perspective: selling 80% consumer CPUs at higher prices can yield (almost surely does) higher profit than selling 90% CPUs at prices that would match AMD's "value" and make Intel look better in reviews.

The most obvious proof of this is that Intel has (significantly) higher margins.

And of course, as AMD's manufacturing potential increases, Intel will have to become competitive on value and performance.
They plan to match AMD this year in mobile and next year in servers. Whether they succeed or not - we'll see soon enough.
High-end gaming desktops, being a low priority segment, has to wait until 2023 or something like that.

However, and that's another thing people forget, as AMD's market share goes up, they'll also raise prices.

 

[piokos]

I was saying technical dominance.

Again: you're comparing representatives of what these companies are selling right now.

To reach that "technical dominance" AMD is very likely spending more (relatively) on CPU R&D and pays for the most expensive node they could use.
So yes, they win in reviews. But they lose in financial statements.

Now, do you think that Intel's engineering team suddenly can't make a desktop CPU better than a revamped Skylake? Or utilize MCM? Or use mesh interconnects in all SoCs?
Do you really think they couldn't go for the same node AMD did?

And if you really want to know (and discuss) how Intel is doing "technologically", you should look at Lakefield - not at Comet Lake.

 

[Zucker2k]

If AMD can offer an R5 4600 @ $199 with an all-core turbo of 4.3 GHz and a 65W TDP, Intel really is in trouble.

Not too long ago, I suggested Intel will cut prices, IF THEY HAVE TO, in order to remain competitive. I nearly got my chewed off. Yet, here we are. AMD cannot win a price war against Intel. Intel won't like it, but they'll cut prices if they're forced to.

1) Desktop - in every segment but high FPS gaming, AMD has price/performance and performance/watt locked up (as far as technical dominance.)

Technical dominance has to be looked at from the perspective of Core vs Core, imho. As it stands, AMD has a node advantage, thanks to TSMC. Core vs Core, Intel is still in the lead according to this:

Spoiler: SPECint2006 Speed

Spoiler: SPECfp2006 (C/C==++) Speed

Spoiler: SPECint2007 Rate-1

Spoiler: SPECfp2007 Rate-1

The normalized for frequency metric is flawed. So, Intel wins overall. What a surprise!

4) Laptop - With Renoir, AMD competes with Intel 10 nm.

I agree

So the ONLY thing Intel has a clear lead in, is high FPS gaming, the 9900k. This chip simply ups that, except its even hotter and more power hungry, and 10 cores does not buy you anything in todays games as far as high FPS gaming, and since its harder to cool, and more power hungry, the 9900k might even beat it. (not withstanding upcoming reviews, so with a grain of salt). And BTW, high FPS gaming IS A NICHE

It is certainly not more niche than Cinebench and Blender. In fact, the people who mostly use rendering software professionally, are on HEDT and Workstation systems. And, when it comes to gaming, Intel has a whole squad of capable gaming chips; namely: 4790k, 6700k, 7700k, 8400, 8600k, 8700k, 8086k, 9600k, 9700kx, 9900kx; and now the 10th gen series.
I'll not concede any area of mainstream desktop usage where AMD is better than Intel. I've been using PCs for more than 20 years now and I don't recall ever having to render anything on my pc, except when running benchmarks. Gaming on the other hand, different story.

Where am I wrong ?

Intel dominates gaming across the board, not just high fps gaming. Also, are you familiar with Intel's boosting profile? The 10900k is going to be boosting more aggresively at every turn than the 9900k. This means, the 10900k will be boosting it's single core all the way to 8 cores higher than the 9900k boosts its cores. This is the same way between the 9900k and 8700k

Now this does not mean that I think Intel is in trouble as a company, financially, it means that AMD will continue to eat away at the market dominance and if that continues too long, then it could hurt Intel. Right now its a wait and see. Marketing or not, you can;t continue "winning" sales based solely on marketing, at some point you have to be competitive.

Intel is competitive. Yes, they have a power deficit now, but they are still very competitive considering they are still on a five year old architecture. And that, folks, can only mean bad news for future competition.

 

[Markfw]

Here:

This is the same issue that makes some people struggle to understand why Intel continues to grow despite AMD offering better value.

You compare CPUs based on product specs and price - like a reviewer, not like a financial analyst. You compare a representative of an Intel CPU and a representative of AMD CPU. You don't think about the economy behind it.
These companies aren't competing with infinite supply of CPUs.

Yes, AMD offers better value, wins reviews, has great press. But at this moment AMD can only make enough chips to cover ~5% of server and ~20% of consumer market.
And that's what Intel adjusts their offer for.
They can ask a price that's higher than AMD's - as long as clients are willing to pay. Because if Intel asked 2-3x more than AMD, some consumers would say: "nah, I can use my PC for a bit longer".
(Mind you: enterprises seldom have the luxury of being able to wait, so Intel actually asks 2-3x more for server chips.)

From Intel's perspective: selling 80% consumer CPUs at higher prices can yield (almost surely does) higher profit than selling 90% CPUs at prices that would match AMD's "value" and make Intel look better in reviews.

The most obvious proof of this is that Intel has (significantly) higher margins.

And of course, as AMD's manufacturing potential increases, Intel will have to become competitive on value and performance.
They plan to match AMD this year in mobile and next year in servers. Whether they succeed or not - we'll see soon enough.
High-end gaming desktops, being a low priority segment, has to wait until 2023 or something like that.

However, and that's another thing people forget, as AMD's market share goes up, they'll also raise prices.

I am not going into detail here, as what you and Zucker2k say does not really refute anything I have said. Aside from the fact that you think nothing will change. You think AMD will never get inroads to OEMS. You think AMD can't make enough CPU's, despite TSMC manufacturing most of the CPU. History tells us that things DO change. I already said it may or may not come to fruition, but, dispute this:

If a company continues to provide an inferior product, eventually it will fail (or lose serious market share and sales)

I think Intel will recover eventually, but the question is for me, how long until they do ? how much market share will AMD get back ? Will AMD continue to deliver serious improvements in the next several series ? Will Intel at some point be better than them ? (in the next 5 or 10 years) I don't have the answers to all these, but I don't think you can assume that nothing will change in the distribution of market share.

The only thing I do challenge is the 14 year old benchmarks that Zucker2k shows. I will not even try.... Lets see some benchmark in the last 3 years, thats RELEVANT.

 

[piokos]

I am not going into detail here, as what you and Zucker2k say does not really refute anything I have said. Aside from the fact that you think nothing will change. You think AMD will never get inroads to OEMS.

I absolutely never said that.
If anything, we're now getting back to how things should be. AMD, being the smaller company, should be launching more cutting-edge products - like they did 20 years ago.
Intel, being the larger company, should lag behind and try to keep up.
This doesn't mean the larger company has worse tech or worse employees. It simply is more inert because of the scale.

I don't see Zen as sudden explosion of AMD's genius. It's just a leap back to the proper relation that we missed, because for a decade AMD was a complete mess.

I just don't understand why you're so emotional about AMD. They're doing all right without your help. And it affects your judgement.
Most of the advantage AMD has right now stems from outsourcing manufacturing to TSMC. But suddenly you start running around and praising their "technical dominance".
Let's be adults.

You think AMD can't make enough CPU's, despite TSMC manufacturing most of the CPU.

I seriously don't understand this sentence. You'll have to explain...

If a company continues to provide an inferior product, eventually it will fail (or lose serious market share and sales)

Absolutely not.
Companies fail when they don't make money. Companies default when they don't make money and no one wants to lend them any.
There's no direct connection to quality of products.

As long as AMD can't get more TSMC 7N supply, Intel can continue to make power hungry CPUs with less cores. We still have to buy them.
And if Intel was willing to accept the margin levels AMD has, they could probably make any of their CPUs on 10nm. But why rush?
If you were an Intel shareholder, would you like a fat dividend or a positive review on a PC enthusiast portal?

 

[Thunder 57]

I just don't understand why you're so emotional about AMD. They're doing all right without your help. And it affects your judgement.
Most of the advantage AMD has right now stems from outsourcing manufacturing to TSMC. But suddenly you start running around and praising their "technical dominance".

Now that is some grade A nonsense. Let's say AMD moved its FX series to the best TSMC had to offer at the time, or if AMD continued developing FX and released it on 7nm TSMC today, it would be a complete joke. Don't try to pretend that AMD hasn't had some massive success in the CPU market. They deserve credit for that. Do you think we would see Intel releasing high core count CPU's at reasonable prices without the competition? Now featuring HT on all models! How generous.

 

[Markfw]

I seriously don't understand this sentence. You'll have to explain...

Absolutely not.
Companies fail when they don't make money. Companies default when they don't make money and no one wants to lend them any.
There's no direct connection to quality of products.

English must be a second language to you, so I am not going to try and explain further.

 

[piokos]

English must be a second language to you, so I am not going to try and explain further.

Well, it is. And to you?
You said:
"You think AMD can't make enough CPU's, despite TSMC manufacturing most of the CPU."
This makes no sense.

 

[piokos]

Now that is some grade A nonsense. Let's say AMD moved its FX series to the best TSMC had to offer at the time, or if AMD continued developing FX and released it on 7nm TSMC today, it would be a complete joke.

?
Why compare to FX? Why not to Palomino?

Take Zen+, add 10% IPC. How would that compete with current Intel 14nm?

It wouldn't be bad, but not exactly drastically leading (especially in laptops).

Don't try to pretend that AMD hasn't had some massive success in the CPU market.

They had a massive crisis for few years. They got back to the expected curve. I'm not going to praise them for that. Sorry.

They're making good CPUs. It's their business. They're making money (or should be). I'm not going to build shrines and write poems.

 

[dmens]

Technical dominance has to be looked at from the perspective of Core vs Core, imho. As it stands, AMD has a node advantage, thanks to TSMC. Core vs Core, Intel is still in the lead according to this:

So, Ryzen 3 roughly matched Coffelake ST throughput with a smaller physical core drawing significantly less power.

In this business that's called technical dominance.

 

[aigomorla]

This seriously reminds me of the days intel has prescott and amd had athlon.
Everyone was raining on prescott even calling it boycott or the ever warming space heater, and the Athlon X2 was so hot people thought AMD had a chance to over take intel.

Then out of nowhere a completely different node archtech was being experimented on and toss'd on laptop processors.
That laptop processor happend to be what was called Banias, which later evolved to be called Dothan.
Dothan was tucked under everyone's eyes because everyone was looking at Athlon X2, and the all mighty Opteron 180.
Then Dothan evolved to what was Penryn and completely smashed AMD so bad it wasn't even funny.
Penryn other name is the allmighty Core 2 Duo.

What i do not understand is Intel had this ellegant road map all the way to 4nm with new litho designs so taught to us by IDontCare.
What i do not also understand is how intel decided to scrap those roadmaps and say it was not profitable.
Was was not profitable? Was it because the FAB's were expensive or was it not possible because Intel saw the End of the Road for Penryn.

Until Intel starts selling there FAB's and decide not to FAB inhouse anymore, i think that is when we can probably say Intel has gone down a black hole.

Otherwise im willing to bet there is another Penryn hiding behind closed doors, they are waiting to pop out on us, but now is just not the right moment as its not profitable, nor is the public ready for it, and they are feeding us the end scraps and fat off penryn until this new chip comes out.

Because these new chips feel exactly like that... the leftover scraps and EOL improvements on a dying arch.

 

[piokos]

So, Ryzen 3 roughly matched Coffelake ST throughput with a smaller physical core drawing significantly less power.

In this business that's called technical dominance.

No. That could be called "product dominance" (if you really need to use this word).
Basically, you've just confirmed what he said: Intel and AMD make similarly good architectures and actual advantage (chip size and power draw) is provided by the node.

In other words:
Intel vs AMD: pretty much a draw
Intel vs TSMC: currently solid lead by the latter.

 

[itsmydamnation]

Otherwise im willing to bet there is another Penryn hiding behind closed doors, they are waiting to pop out on us, but now is just not the right moment as its not profitable, nor is the public ready for it.

We call this magical thinking, it happens once for one company so it must always happen right?

 

[itsmydamnation]

No. That could be called "product dominance" (if you really need to use this word).
Basically, you've just confirmed what he said: Intel and AMD make similarly good architectures and actual advantage (chip size and power draw) is provided by the node.

In other words:
Intel vs AMD: pretty much a draw
Intel vs TSMC: currently solid lead by the latter.

this wrong and you are wrong.

 

[aigomorla]

We call this magical thinking, it happens once for one company so it must always happen right?

lol... yeah but you have to ask where the hell did all that RnD funding go?
The RnD funding intel has done was astronomic.

 

[itsmydamnation]

lol... yeah but you have to ask where the hell did all that RnD funding go?
The RnD funding intel has done was astronomic.

I can name so many tech companies that by that logic should be massively successful right now but aren't.

SGI?
Nortel?
compaq ?
Nokia

i could just keep going.

 

[piokos]

Otherwise im willing to bet there is another Penryn hiding behind closed doors, they are waiting to pop out on us, but now is just not the right moment as its not profitable, nor is the public ready for it, and they are feeding us the end scraps and fat off penryn until this new chip comes out.

They don't feel the need to launch anything revolutionary. R&D is one thing. Companies like Intel fund development all the time. Sometimes it results in a product, sometimes not. It's always some know-how gain.

Launching a product is another story. It's hugely expensive and has to make sense financially. New product replaces the previous one. So it has to offer the same profit + cover the new lineup campaign (setting up manufacturing etc).

Intel shows a lot of new stuff. They just don't update all product segments. Skylake still makes money.

Think: Sunny Cove. They could make it on 14nm to give us some IPC gain last year. They decided to skip it and do that with Tiger Lake this year.
They're still selling everything and breaking records. But they saved a lot as well.

 

[darkswordsman17]

lol... yeah but you have to ask where the hell did all that RnD funding go?
The RnD funding intel has done was astronomic.

It went all over the place. I'm not being glib, Intel had so many projects, a bunch of which they just cancelled and never got return on. Larrabee, cellular modems (which I think they might be developing again after having canned their 5G one?), wearables (you know, like IoT chips to put in clothing so you can control LEDs and stuff - yes that really is something Intel did and I think spent like a billion dollars on), literally paying companies to try and use Atom (which didn't work), and I'm sure I'm forgetting a bunch of others. 3D Xpoint. Hybrid Memory Cube (I think that's the term, HMC which was going to compete with HBM). Heck, they I think sold most of their flash production to Micron.

There's no doubt that Intel has capability, and they have the talent. But it takes more than that. What if management decides to take a chance that doesn't pan out (like Bulldozer designs did to AMD, or some of the P4 decisions, or say Itanium and there's others in Intel's history)? They're also chasing a lot of markets that they've failed spectacularly multiple times in (GPUs) at the same time they're trying to integrate stuff that no one has ever. Plus they're planning for stuff further down the line. Jim Keller for instance viewed his role there as developing next gen interchip communication architecture (which I have a strong hunch has plans for fiberoptics or possibly even lasers; which speaking of, Intel spent tons of money developing Thunderbolt which was going to be an external fiber optic).