News Intel's Rocket Lake Blasts Off With Fewer Cores, Higher Pricing

[hotaru251]

imagine how high you have to be to charge more for less and not being the leader in single or multithread performance....

 

[Deleted member 2838871]

The future is unknown, but this gen is the biggest embarrassment to intel in its history.

I guess that's why their took their former engineer employee and made him CEO. If AMD can turn it around... Intel can too.

imagine how high you have to be to charge more for less and not being the leader in single or multithread performance....

They know there's a bunch of suckers out there like me who will buy a 10900k because all the flagship AMD chips aren't available. 😆 It's all good either way... I can live with my top 1% benchmarks.

 

[InvalidError]

Intel SHOULD have more leeway to lower pricing given their 100th year on 14nm.

A good chunk of that 'leeway' goes away when the chip is 33% larger than its predecessor - that's a yield double-whammy of at least 33% fewer dies per wafer, each with 33% higher chance of having a defect.

imagine how high you have to be to charge more for less and not being the leader in single or multithread performance....

At the higher-end, maybe. At the non-enthusiast mid-range though, Intel is set to dominate since AMD does not have anything that can compete with the 11400F-11600F: the 3600X costs more for less while the 5600X is about as fast at a 50+% higher price. AMD will need a significant price shuffle there.

 

[TerryLaze]

What is killing AMD here is that they made the safe bet in early 2020 to allocate the majority of their wafers to console APUs. They could not have anticipated the huge increase in demand as a result of the human malware situation. Consoles always sell out for months after release, so even though they are low-margin products, they are a safe investment. For example, TSMC can produce 4 Vermeer CCDs for every Xbox SoC, so the difference in potential capacity is huge.

On top of that, Intel has had shortages for a long time and increased 14nm capacity significantly over the last 3 years. At the same time, more and more of Intel's sales are being shifted over to 10nm. This has probably led to a surplus in 14nm capacity as they are still in the process of transitioning their 14nm equipment to 10nm.

I think they set that in stone when they made the first deal for the ps4 xbone, I very much doubt that sony and ms would leave that to the winds.

imagine how high you have to be to charge more for less and not being the leader in single or multithread performance....

Literally AMD in march 2021.

View: https://www.youtube.com/watch?v=oj8SuJ2Mb6A

 

[Questors]

"...especially as AMD seems to be finally correcting its supply shortages. "

They are?

 

[InvalidError]

"...especially as AMD seems to be finally correcting its supply shortages. "

They are?

Yes, for all of 10 more seconds on launch-day

 

[cryoburner]

AMD hiked prices on Ryzen 5000 because they were proud that they finally beat Intel

I'm pretty sure they hiked prices because they knew they were going to be limited by their 7nm manufacturing capacity this generation, with their CPUs, GPUs and console APUs all being manufactured on the same process. They undoubtedly made contracts with both Microsoft and Sony years in advance to supply millions of console chips each year, and each of those is several times the size of a Ryzen chiplet. Add to that their Zen3 chips being more or less all-around faster than Intel's 10th-gen lineup, and it should have been obvious that demand would be higher than usual as well. So they knew that even with pricing that would be unattractive to most, there would still be enough demand among enthusiasts to sell as many as they could produce, at least until Rocket Lake launched. So, I can see why they didn't initially launch any value models as they normally do, and even hiked prices on the "premium" ones.

With Rocket Lake coming out now though, demand for Zen3 among those less price-conscious enthusiasts is most likely dropping off, which is why we are now starting to see the 5600X and 5800X readily-available at major retailers. Even if Rocket Lake doesn't manage to match or exceed Ryzen in all areas, it should mostly close the performance and feature gap, and is priced more attractively, at least at the i7-level and below. So, I would expect AMD to launch some more value-oriented models before too long, as demand for the enthusiast models wanes. Though it's possible that they might hold out a little longer due to the increased demand for graphics cards now.

16/20 PCIe lanes is a joke. I have 10 PCs and they are all 28, 44 or 48 PCIe lanes.

Keep in mind that these are PCIe 4.0 lanes, each able to transfer twice as much as a 3.0 lane. And of course, the motherboard chipset is providing an additional 24 PCIe 3.0 lanes. Unless one has some specific use case requiring lots of high-bandwidth expansion cards, this probably isn't going to be much of a concern.

A good chunk of that 'leeway' goes away when the chip is 33% larger than its predecessor - that's a yield double-whammy of at least 33% fewer dies per wafer, each with 33% higher chance of having a defect.

Technically, if a chip were 33% larger, that would work out to around 25% fewer per wafer, at least before getting into edge waste. Just saying. : P

 

[wtfmang]

seems like i5 11500 is a better deal than the ancient ryzen 3400G.

i wonder if a UHD 750 can beat the old Vega.

Alder Lake is coming with 48EU right?

 

[cryoburner]

seems like i5 11500 is a better deal than the ancient ryzen 3400G.

i wonder if a UHD 750 can beat the old Vega.

Alder Lake is coming with 48EU right?

It's definitely not going to beat Vega 11, or even Vega 8, at least for things like gaming. Intel is advertising UHD 750 as being "up to 50% faster" than their current UHD 630, whereas Vega 11 is in many cases over 3 times as fast as UHD 630. So it should still only offer somewhere around half the graphics performance of AMD's existing Ryzen desktop APUs. It's possible that the Intel processors could pull off some wins in very CPU-limited games running at low resolutions and settings, but the graphics hardware won't be on the same level, and a 3400G or 2400G will likely win out in practically all graphically-demanding titles, even if they are behind on the CPU side of things.

If someone were intending to eventually upgrade to a dedicated card for gaming, an 11th-gen i5 would be more capable with that upgraded setup, but while on integrated graphics they are still at a big disadvantage compared to AMD's APUs, and will undoubtedly struggle to run newer AAA games even at 720p.

As for the EUs, UHD 750 has 32, while the 11400 only gets UHD 730, with 24.

 

[Conahl]

Intel isn't loosing any market, mind or any other type of share by releasing the 11th gen because they have other products for those other markets and for the desktop market the 11th gen is just as good as anything else and better.

oh ? you sure about that ?
https://www.tomshardware.com/news/ryzen-dominates-puget-systems-sales
if they are losing sales there, then chances are, they are losing sales in other areas and stores/ suppliers as well. NO one i know is even looking at intel now, i mentioned the write up on AT about rocket lake, and suggested he take a look, his reply ? why would i ? i dont care about intel now. they have nothing that interests me.
by looking at the 2 articles that AT has posted, at this point, rocket lake looks like it is a dud.

The only pool of customers intel is loosing is cheap skate semi/pros that try to build a workstation with desktop components and that's because intel isn't targeting that group of people.

yes cause they conceded the HEDT to AMD, cause they cant compete in that market, at all, now.

 

[Why_Me]

Meanwhile Newegg and Scan UK have both dropped prices on the Rocket Lake cpu's they have listed while AMD has raised prices on a large portion of their Ryzen cpu's including the 3600 which just went from $200 to $220. If you don't run cpu intensive task there's literally zero reason to purchase an AMD desktop cpu atm.

 

[TerryLaze]

oh ? you sure about that ?
https://www.tomshardware.com/news/ryzen-dominates-puget-systems-sales
if they are losing sales there, then chances are, they are losing sales in other areas and stores/ suppliers as well. NO one i know is even looking at intel now, i mentioned the write up on AT about rocket lake, and suggested he take a look, his reply ? why would i ? i dont care about intel now. they have nothing that interests me.
by looking at the 2 articles that AT has posted, at this point, rocket lake looks like it is a dud.

If theses CPUs would be left unsold then intel would lose sales and you would have a point but just because puget doesn't sell them doesn't mean that they are left unsold.
Sale numbers from intel are super clear that intel is selling more then before and that means that they are not losing market share.

yes cause they conceded the HEDT to AMD, cause they cant compete in that market, at all, now.

It's a super small market which makes producing special CPUs for it not worth it for intel if they can just sell standard CPUs to a huge market for a big margin.

 

[spongiemaster]

It appears the Gear setting is available in the BIOS and any CPU can be set to Gear 1. For some reason Intel just decided to set that to Gear 2 by default for any non-I9 chip. With that known, the I9 looks even less attractive vs the I7-11700K.

Intel i7-11700K Memory Benchmarks: Gear 1 vs. Gear 2 on 3200MHz - YouTube

 

[TerryLaze]

It appears the Gear setting is available in the BIOS and any CPU can be set to Gear 1. For some reason Intel just decided to set that to Gear 2 by default for any non-I9 chip. With that known, the I9 looks even less attractive vs the I7-11700K.

Intel i7-11700K Memory Benchmarks: Gear 1 vs. Gear 2 on 3200MHz - YouTube

Only the i9 parts have a mem controller that passed intel's quality check for gear 1 with 3200.
So for all other CPUs setting it to gear 1 to run 3200 mem is considered overclocking by intel.
It might be hugely overly protective and you might never run into issues doing it but intel having to cover their asses have to do it that way.
it's just like XMP, it voids the warranty.

 

[spongiemaster]

Only the i9 parts have a mem controller that passed intel's quality check for gear 1 with 3200.
So for all other CPUs setting it to gear 1 to run 3200 mem is considered overclocking by intel.
It might be hugely overly protective and you might never run into issues doing it but intel having to cover their asses have to do it that way.
it's just like XMP, it voids the warranty.

The warranty thing doesn't matter. When was the last time someone killed a CPU running faster memory? No enthusiast is running 2666 memory. Everyone uses XMP profiles today.

Still waiting on you to give the PL4 value for Rocket Lake.

 

[TerryLaze]

The warranty thing doesn't matter. When was the last time someone killed a CPU running faster memory? No enthusiast is running 2666 memory. Everyone uses XMP profiles today.

Still waiting on you to give the PL4 value for Rocket Lake.

It doesn't matter if it doesn't matter to you, I'm just saying why the difference exists.

PL4 is going to be extremely close to what anand found as the peak power draw I had already said that, they found around 290W.

 

[VforV]

For someone that claims to be a tech fan you sure have no idea what's going on, these CPUs aren't targeted at server or pro-sumers and gamers buy mid range CPUs at best and pair them with the highest GPU they can get away with or afford.
Intel isn't loosing any market, mind or any other type of share by releasing the 11th gen because they have other products for those other markets and for the desktop market the 11th gen is just as good as anything else and better.

The only pool of customers intel is loosing is cheap skate semi/pros that try to build a workstation with desktop components and that's because intel isn't targeting that group of people.

For someone who likes to teach others your sure don't know how to read, nor are you up to date about changes in the market.

I was talking about intel "worry", not about 11th gen being server CPUs. Worry as in general, about AMD resurrection and better performance in server, pro-sumers and desktop/laptop too. You can spin it all you like, all you're doing is twisting facts.

Also minshare is a real thing, again you ignore and make fun of, but intel did lose and still is losing on that front.

If I see a good AMD product I praise it, if I see a bad one I bash it. If I see a good intel product I also praise it, if it's bad, I call it what it is. I have no allegiance other than the best product price/performance wins my money.

Rocket Lake is a patethic gen from Intel and that's all that matters. It's a fact, not an opinion proved by numbers and will be proven again on launch.

I'm done talking to stuck up fainbois like you, you're a unreasonable and unfixable. The END.

 

[TerryLaze]

For someone who likes to teach others your sure don't know how to read, nor are you up to date about changes in the market.

I was talking about intel "worry", not about 11th gen being server CPUs. Worry as in general, about AMD resurrection and better performance in server, pro-sumers and desktop/laptop too. You can spin it all you like, all you're doing is twisting facts.

For the last five years people have been saying that AMD is taking *share but intel has ben increasing their sales, ETMLIF why they should worry.
Unless you mean it in the same way that someone should worry about not forgetting to eat every few hours when playing video games.
The market is big enough for intel to sell all they can and for AMD to do the same.

The only thing intel should worry about is AMD not being able to continue making good CPUs and putting intel into the same bad position intel was in for as long as AMD only had the FX chips. Intel doubled their net income with ZEN in the picture, from year one, the last thing they want is to go back to making the pre ZEN amounts.

 

[everettfsargent]

We now know that an i9-11900K will last for at least eight minutes at full OC burn ...

View: https://youtu.be/XGRE_7yz_kM?t=3108

 

[everettfsargent]

It doesn't matter if it doesn't matter to you, I'm just saying why the difference exists.

PL4 is going to be extremely close to what anand found as the peak power draw I had already said that, they found around 290W.

"4.2.1.1 Package Power Control
The package power control settings of PL1, PL2, PL3, PL4 and Tau allow the designer
to configure Intel® Turbo Boost Technology 2.0 to match the platform power delivery
and package thermal solution limitations.
• Power Limit 1 (PL1): A threshold for average power that will not exceed -
recommend to set to equal TDP power. PL1 should not be set higher than thermal
solution cooling limits.
• Power Limit 2 (PL2): A threshold that if exceeded, the PL2 rapid power limiting
algorithms will attempt to limit the spike above PL2.
• Power Limit 3 (PL3): A threshold that if exceeded, the PL3 rapid power limiting
algorithms will attempt to limit the duty cycle of spikes above PL3 by reactively
limiting frequency. This is an optional setting
• Power Limit 4 (PL4): A limit that will not be exceeded, the PL4 power limiting
algorithms will preemptively limit frequency to prevent spikes above PL4.
• Turbo Time Parameter (Tau): An averaging constant used for PL1 exponential
weighted moving average (EWMA) power calculation.
Notes:
1. Implementation of Intel® Turbo Boost Technology 2.0 only requires configuring
PL1, PL1 Tau and PL2.
2. PL3 and PL4 are disabled by default."
https://www.intel.com/content/dam/w...n-core-families-datasheet-vol-1-datasheet.pdf

https://cdrdv2.intel.com/v1/dl/getContent/631121

https://www.intel.com/content/www/us/en/products/docs/processors/core/core-technical-resources.html

AFAIK, exact same language applies to both 10th and 11th generations.

PL3 and PL4 are disabled by default.

The only limit appears to be Tjunction (or Tj). More than gladly stand to be corrected though (with some official words/quotes directly from Intel).

 

[TerryLaze]

https://wikidiff.com/disabled/off

disabled is made incapable of use or action while off is inoperative, disabled

Disabled means that you can't change the value by default, you can't use or take action on it, not that it doesn't work.

But you already showed the answer to your question yourself.
PL3 is an optional setting so you can find a way to fudge with it, PL4 is not, you can't change that, no matter what.

• Power Limit 3 (PL3): A threshold that if exceeded, the PL3 rapid power limiting
algorithms will attempt to limit the duty cycle of spikes above PL3 by reactively
limiting frequency. This is an optional setting
• Power Limit 4 (PL4): A limit that will not be exceeded, the PL4 power limiting
algorithms will preemptively limit frequency to prevent spikes above PL4.

 

[everettfsargent]

https://wikidiff.com/disabled/off

Disabled means that you can't change the value by default, you can't use or take action on it, not that it doesn't work.

But you already showed the answer to your question yourself.
PL3 is an optional setting so you can find a way to fudge with it, PL4 is not, you can't change that, no matter what.

So, if disabled then there is no PL4 limit, unless it were to be enabled. When the i9-11900K reviews come out then we will know for sure. But, so far, the 9th and 10th generation all exceeded at least 300W. It will be interesting to see if anyone can suss out the PL3 and PL4 limits (optional or imposed by Intel).

 

[TerryLaze]

So, if disabled then there is no PL4 limit, unless it were to be enabled. D'oh!

Disabled means you can't change it not that it's not there. You can cry about it all you want.

 

[everettfsargent]

Disabled means you can't change it not that it's not there. You can cry about it all you want.

See edited comment, and no, there will be no crying no matter what the reviews show.

 

[spongiemaster]

See edited comment, and no, there will be no crying no matter what the reviews show.

You're right. There's nothing ambiguous about what you posted from Intel. PL1, PL2 and PL3 values are required to be set for Intel's Boost to work, PL3 and PL4 are optional, and turned off by default. Rather then just admit he's wrong on this one and that Intel CPU's do not come with a hard limit from Intel, he's trying to redefine the word disabled.