Alleged Leaked Details on Intel Comet Lake-S Platform Require... You Guessed It... A New Platform

[TheoneandonlyMrK]

All of this is true, but, again, @juiseman was implying that because Intel fucked up the engineering of 10nm, that intel would therefore be incapable of implementing PCI-E 5.0 on time.

Would you, @theoneandonlymrk, please like to explain to me, exactly how juiseman's point is true, bearing in mind that:

1 - designing a controller to implement an existing standard is nowhere near as complicated as building a new semiconductor manufacturing process from scratch
2 - PCI-SIG doesn't release standards that aren't ready to be implemented, whereas Intel *did* announce a 10nm technology that was nowhere near implementation.

Im not here to back his word on it ,i get you and him, I was just adding my opinion.
Which was that Intel's reluctance to advance their architecture is clear to see and detrimental to their profit line.
But why, ask that question, they have not even upped lane count though they are changing the designs ,adding cores , security and even increasing the Gpu capabilities.
Amd did a lot of work to get to zen, The pciex 4 support work should have been mirrored by intel, probably lead by, but no ,they NEED the die space to get anywhere near competitive on core counts which likely has constrained their design, I think 2022 is a pipe dream though because they're going to be too busy and constrained to bring pciex5 to consumer's.

ImHo it will take Intel's new Arch on 7nm euv to make room for that and I don't see that coming before 2023-4.

Just my opinion though feel free to disagree.

 

[Dave65]

This is just SAD
It is like a new car dealer trying to sell a 14 year old car as new yelling hey, look what I can do

 

[Vayra86]

This is just SAD
It is like a new car dealer trying to sell a 14 year old car as new yelling hey, look what I can do

Maybe that's an idea. They have to push the nostalgia button! 'Remember Sandy Bridge'...

'Intel Vintage Series'
"For 10th gen, we went for a blast from the past, because hey, its all the same anyway!"

 

[TheinsanegamerN]

Their arrogance is leading to their fall

Personally, I LOVE it.

For one thing, AMD's stocks will climb higher as they steal more and more of intel's shares, pay off debts, and increase R+D spending (which is great for me, I can finally sell my stocks and build a whole new gaming PC).

If AMD CPUs get enough marketshare, the sheer revenue increase should also mean RTG gets the much needed shot int he arm it needs to compete with nvidia again. We may yet see another HD5000 series of GPUs.

And it makes intel compete again, which regardless of which CPU you have spells great news for consumers.

The only thing I hope is intel takes another 2-3 years to get their heads on straight, perhaps get close to a 50% market share, so we have a few nice years of intense competition to enjoy, without being so incompetent that AMD gets complacent and rests on their laurels on the CPU side.

 

[awesomesauce]

hopefully it will come with fixed security

nothing really new here, intel always need new chipset for new processor

 

[eidairaman1]

hopefully it will come with fixed security

nothing really new here, intel always need new chipset for new processor

It's just not a new chipset for new processor. It's a new socket and motherboard every 6 months to a year for them.

 

[Tomgang]

For those talking PCIe gen 4 vs. gen 5. Here is a question for you.

If PCIe gen 4.0 requires a fan, does that mean Intel will bring out the chiller next round for PCIe gen 5?

 

[Vya Domus]

So the rumors were true, no 10nm high-performance part for desktop anytime soon. I can no longer see how Intel is going to ever have the process lead again, the more they dedicate resources and products towards this old node the bigger the gap is going to get between them and AMD. Sunny Cove is literally their last bastion of performance, after that it's over, without a good node there is nothing more that they can squeeze. However, AMD will also be on another node by then, one that will provably work.

 

[Wavetrex]

PCIe 4.0 to Intel: Am I A Joke To You ?

 

[Deleted member 67555]

This has happened before and it will happen again (creepy Cylon Basestar pilot in a slime tub skin job voice)
13 years ago but instead of socket it was chipset..
Pay no attention to AMD's FX-60...You heard about Conroe, right? But before we give you that check out this $1000 Pentium D 965 EE...
Same thing...same

 

[Shou Miko]

No supports Intel is gonna have to spending more money on a new motherboard with a new CPU it's all in their tick tock system.

 

[S4BRE]

I hope the board partners (ASUS, MSI, et al) at least make some decent mATX boards for this platform... the last few iterations have been quite lackluster at best, or missing from entire markets altogether (like the Maximus Gene series not being officially available in the US). Then again there aren't too many good boards on the AMD side either; tough spot to be in as an mATX fan .

 

[Manoa]

waste of cpu's, and waste of mobo's for them, intel is a waste

 

[Shou Miko]

I hope the board partners (ASUS, MSI, et al) at least make some decent mATX boards for this platform... the last few iterations have been quite lackluster at best, or missing from entire markets altogether (like the Maximus Gene series not being officially available in the US). Then again there aren't too many good boards on the AMD side either; tough spot to be in as an mATX fan .

Don't get your hopes up even AsRock demoted their m-atx motherboards to their Pro series instead of the Extreme.

Their last really oc formula m-atx board was their Z170M OC Formula so if you are looking for a high-end m-atx there is only Asus' Gene board left if they choose to keep it bcs ram oc has moved to ITX.

 

[Alexandrus]

"LGA 1200 package (with 9 more pins that the current LGA 1151)."

Serisouly ? Are you THAT bad at math ? What a joke, sorry, but this is elementary school algebra.
Just to clarify, for those "journalists" that don't really get it, but 1200-1151=49, NOT 9.

 

[TheoneandonlyMrK]

"LGA 1200 package (with 9 more pins that the current LGA 1151)."

Serisouly ? Are you THAT bad at math ? What a joke, sorry, but this is elementary school algebra.
Just to clarify, for those "journalists" that don't really get it, but 1200-1151=49, NOT 9.

So there Are 49 pins extra in use for deffinite, proven?.

Oh I Definitely think the socket will change but are they seriously telling us that when they specced up the 3xx boards they didn't expect to have to put more cores in a year later, I think they did but don't mind those chipset sales, And yet again Need more power and ground pins, really.

 

[dyonoctis]

"LGA 1200 package (with 9 more pins that the current LGA 1151)."

Serisouly ? Are you THAT bad at math ? What a joke, sorry, but this is elementary school algebra.
Just to clarify, for those "journalists" that don't really get it, but 1200-1151=49, NOT 9.

It might just be a typo, you don't have to assassinate him like that. The article linked as the source does say that it's 49 more pins.

 

[Metroid]

Thankgod I went ryzen 3xxx, now I will upgrade my platform only when ddr5 is required.

 

[Aquinus]

AMD gives you four more PCIe lanes. Now that may not seem like much but that's four more PCIe lanes to be dedicated to your system's boot NVMe SSD thus not leaving you hoping that your GPU doesn't bottleneck the NVMe SSD.

You know what also should seem like a lot? The X570 chipset having 4 PCIe 4.0 lanes, or the equivalent of 8 lanes of PCIe 3.0. Even with NVMe cards running at PCIe 3.0 off the chipset, twice as much bandwidth to the chipset is a big win. That means a single good NVMe card isn't going to even come close to saturating the bus with AMD and you probably could comfortably run two good NVMe cards off the chipset, but it could be really close with Intel with just a single NVMe card.

Let's put this in perspective though. I'm using a SBe 3930k which launched in Q4 of 2011 and I'm running a Vega 64 and NVMe card at PCIe 3.0.

 

[voltage]

Oh no, we have to buy another mother board, oh no, its time to constantly complain... I say Big Deal... the 3 series has had its run, bring on the next. And while your at it Intel, hurry at something, and bring DDR5, and PCIE-5 out, soon. like in 2021 at the latest.

 

[eidairaman1]

Oh no, we have to buy another mother board, oh no, its time to constantly complain... I say Big Deal... the 3 series has had its run, bring on the next. And while your at it Intel, hurry at something, and bring DDR5, and PCIE-5 out, soon. like in 2021 at the latest.

Silverspoon syndrom?

I remember skt 7/super 7 being the platform to go with, easy cpu upgrades, not having to rip the board out every year...

AM4/TR4 have taken its place.

 

[lexluthermiester]

No PCIE 4.0?...hmmmm

Why does it matter? No device in existence currently requires the maximum bandwidth of PCIe 3, let alone PCIe 4. Your implied complaint is a shallow one.

 

[Totally]

Keller isn’t saying what he’s working on yet, and maybe he won’t for years.

Here we’re going to make the best possible server, the best possible client. We’re going to build great graphics. We’re participating heavily in the AI revolution. There’s a bunch of interesting problems there. We’re going to do some interesting stuff in that space.

Why rock star chip architect Jim Keller finally decided to work for Intel

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More In Silicon Valley, we take things like faster chips and Moore’s law for granted. It seems like progress is always inevitable, given the history of semiconductors and the...

venturebeat.com

Probably referring to the gpu on the cpu die. He's always been a CPU guy and only a CPU guy.

 

[Dave65]

Mah FPSess will explode!



Behind ... in games.

Yeah 5 fps really matter... Grow up!

 

[lexluthermiester]

Silverspoon syndrom?

I remember skt 7/super 7 being the platform to go with, easy cpu upgrades, not having to rip the board out every year...

AM4/TR4 have taken its place.

That's a fair assessment.