Monday, December 26th 2022
Intel "Raptor Lake Refresh" Meant to Fill in for Scrapped "Meteor Lake" Desktop?
Intel's 2023 roadmap for the desktop processor segment sees the company flesh out its 13th Gen Core "Raptor Lake" desktop family with 65 W (locked) SKUs, and the new i9-13900KS flagship; followed by a new lineup of processors under the "Raptor Lake Refresh" family, due for Q3-2023, with no mentions of a desktop "Meteor Lake" processor in the year. It turns out that "Raptor Lake Refresh" is being designed to fill in for these (i.e. there won't be any "Meteor Lake" desktop chips). This, according to OneRaichu, a reliable source with Intel leaks.
"Meteor Lake" is Intel's first client processor to fully incorporate the company's IDM 2.0 product development strategy of disintegrating the processor into multiple chiplets built on various foundry nodes based on design needs; and combining them onto a single package with a high-performance interconnect. "Meteor Lake" has just one problem and that is CPU core-counts, with rumors pointing to 6P+16E (6 performance cores + 16 efficiency cores) being the maximum core-count possible, something Intel probably feels won't be competitive in the desktop segment against AMD, which will probably have a lineup of "Zen 4" X3D processors out by Q3-2023, with up to 16 P-cores. The company will, however, give "Meteor Lake" a sizable launch in the various mobile segments."Raptor Lake Refresh" remains shrouded in mystery, particularly what Intel does with packaging it—whether it retains LGA1700 or uses the next LGA1851 package; or whether it is a speed-bump, or like "Coffee Lake Refresh," Intel could even increases the core-counts. Assuming Intel doesn't change the silicon from the present 8P+16E, the "Refresh" series could see incremental core-count uplifts among each Core brand extension (eg: Core i5 going from 6P+8E to 6P+16E); besides clock speed increases. Should Intel take the path of changing the socket to LGA1851, the company might change the branding to 14th Gen Core, release a new chipset, with the socket probably offering improved I/O, such as CPU-attached PCIe Gen 5 NVMe (currently Gen 4). These LGA1851 motherboards will come with preparation for next-generation "Arrow Lake" processors due in 2024.
Sources:
OneRaichu (Twitter), HotHardware
"Meteor Lake" is Intel's first client processor to fully incorporate the company's IDM 2.0 product development strategy of disintegrating the processor into multiple chiplets built on various foundry nodes based on design needs; and combining them onto a single package with a high-performance interconnect. "Meteor Lake" has just one problem and that is CPU core-counts, with rumors pointing to 6P+16E (6 performance cores + 16 efficiency cores) being the maximum core-count possible, something Intel probably feels won't be competitive in the desktop segment against AMD, which will probably have a lineup of "Zen 4" X3D processors out by Q3-2023, with up to 16 P-cores. The company will, however, give "Meteor Lake" a sizable launch in the various mobile segments."Raptor Lake Refresh" remains shrouded in mystery, particularly what Intel does with packaging it—whether it retains LGA1700 or uses the next LGA1851 package; or whether it is a speed-bump, or like "Coffee Lake Refresh," Intel could even increases the core-counts. Assuming Intel doesn't change the silicon from the present 8P+16E, the "Refresh" series could see incremental core-count uplifts among each Core brand extension (eg: Core i5 going from 6P+8E to 6P+16E); besides clock speed increases. Should Intel take the path of changing the socket to LGA1851, the company might change the branding to 14th Gen Core, release a new chipset, with the socket probably offering improved I/O, such as CPU-attached PCIe Gen 5 NVMe (currently Gen 4). These LGA1851 motherboards will come with preparation for next-generation "Arrow Lake" processors due in 2024.
61 Comments on Intel "Raptor Lake Refresh" Meant to Fill in for Scrapped "Meteor Lake" Desktop?
Until then, I do understand the usage of those smiles.
That would be ironic.
If I would have said "yeah because people care so much about power consumption"
that could have been sarcasm.
But I said "Idk if you live on planet earth, but power consumption has been on people's mind a lot for a while now"
That would just be wrong if I did not mean it.
anywho, here ya go, thought Id make it about Greece because that is where you are from:
www.ekathimerini.com/economy/1198807/power-consumption-declines/
There. I can't make it simpler than that. Even in Greece.
Its not for nothing AMD makes a big deal out of their 50% performance per watt improvement and that they dont just give their gpu's a 600 watt ceiling allowing them to handily beat big N just at a MUCH higher power consumption.
I hope we've reached the point where power consumption starts decreasing because it's reaching the point where buying a graphics card has you considering how many amps your circuit supports, summer temperatures, and whether your case and power supply can handle the beast. People who want every last drop of performance can OC their cards as they wish.
Intel, can we PLEASE stop with the Lakes already? At this point, even the horse is screaming "I'm dead already! I'm dead already! Stop it!" :D :roll:
Some people fail to realise that they won't necessarily get more performance, depending on their cooling setup, if they pump too much power through the CPU that the GPU temps rise. That's an area where the 12/13th gen has issues when pushed to its limit.
And in gaming just put a 100W PL cap.
Seems like most of the "so called lakes" are mostly creeks or streams or even puddles in popularity.
2024 - Arrow Lake. LGA 1851
2025 - Panther Lake. LGA 1851
2026 - Nova Lake. LGA 1851
And that's assuming nothing else gets cancelled or postponed.
It will not look good on presentation and might be a side grade as do 11th gen was but after that the road is clear.
Just like AMD having now a less than successful gen and intel is booming, the wheel will turn in 1 year or so.
According to the "List of Intel manufacturing sites", while there are very few factories capable of producing Intel 4, there are construction plans for Intel 20A production plants, which will be as numerous as Intel 7.
Intel 7 (5 sites): Fab 12, 22, 28, 32, 42
Intel 4 (1 sites): Fab 34
Intel 20A (4 sites): Fab 27, 52, 62, 42(switched from Intel 7)
The site (real estate) has already been acquired, and we are certain that a significant amount of money has been invested. It would be obvious that the Intel 4 node is a middleman and there is no plan for mass production of products in all segments. To cover all segments with Intel 4, at least two more Fabs would have to be built by 2021 to be ready in time, and Intel does not have the necessary number of EUV lithography for that.
AMD makes a big deal about power consumption the last 7+ years, but it is backfiring in their face the last 2-3 years. People rush to buy power hungry stuff to have the best possible performance, no matter the power consumption. AMD already changed route with Zen 4, letting it turbo as high as possible, while also increasing TDP to 170W, to achieve best possible performance. They will do the same with RDNA 4 or even with RDNA 3, if they come out with a refresh that turbos probably over 3GHz.
AMD is in a desperate position with Nvidia increasing the lead, while Intel is coming from behind to steal OEM sales in the GPU market (from both AMD and Nvidia). At the same time, power hungry Intel CPUs, that are advertised as 16 and 24 core models, while having more E cores than P cores, are taking back market share in the retail market.
AMD is holding up thanks to servers and consoles. If Intel comes out with a good server CPU or even manages in 2-3 years to have something good for consoles to steal that market from AMD, well, many will get what they wised for. Monopolies, stagnation and ridiculous high prices in about everything. People will buy those power hungry models over more efficient models based on a very understandable logic
"I'll get the product that offers the best performance, undervolt it and if I ever need that extra performance, I will run it at defaults or even overclock it".
They will not get the more efficient but slower part, because they wouldn't probably have the option to ... unlock that extra performance when they need to. Things changed compared to the past. Companies try to win benchmarks and put price tags on new products based on their optimal performance. If the company A sells the product B with a frequency of 1GHz to keep it efficient, while that product can reach 1.2GHz frequency, it's leaving 20% performance on the table. That means it will be looking 20% slower on benchmark charts and the company will have to price it accordingly, as a 20% slower product. Instead companies try to overclock their new products to their maximum, to achieve maximum performance, to advertise them as fast, to put price tags accordingly.
Let me say it with differently. In the past someone was buying a product that could overclock 30%. That meant "FREE performance". Free is dead today. You buy the product at the highest possible price and then you have the option to LOSE performance, if you wish for efficiency.
@OneRaichu
1/x
About Meteor lake.
MTL focus on how to improve the efficiency of the instruction execution, it will not widen the microarchitecture crazy like Alder lake.
May 18, 2022
@Arjunownit0
People who upgrades to zen 4 /raptor lake will be like 30%♂️. Cuz next gen desktop parts (zen5 /Arrow lake ) is a bigger upgrade. And Meteor lake is only mobile. There is ni desktop parts on meteor lake, thats why they say Arrowlake is 40% over Raptorlake.
July 14, 2022
Mindfactory sales numbers, AMD outsells Intel 2 : 1
But the majority of AMD sales are Ryzen 5000, so perhaps people buying previous generation processors won’t be bragging about them in build request threads?
And of course many people dismiss Mindfactory numbers since they are a German store, so maybe not applicable to their part of the world, or representative of majority…
It's much much better than all the other manufactures who use three or four digit numbers and then when they get past 900 or 9000 start over with a completely different numbering scheme. Ask a layperson which is better - a Radeon HD 7970 or a Radeon RX 7900 XTX. They would probably say the 7970 because it has a higher number even though we know it is a decade older and the numbering scheme rolled over. The worst was AMD in the 2010s who just give all their GPU products a unique name (Vega, Fury, etc.) that provided no context on how new it was.