News Leaked MSI slide mentions Arrow Lake Refresh — reignites hopes that LGA 1851 will last more than a single-generation

[Admin]

A leaked slide from MSI shows that Intel had plans for an Arrow Lake Refresh at some point, though it appears those plans have been scrapped for now.

Leaked MSI slide mentions Arrow Lake Refresh — reignites hopes that LGA 1851 will last more than a single-generation : Read more

 

[usertests]

I think there was a newer MLID leak on this, but in general it should be easy for Intel to do some kind of refresh even if they cancelled an 8+32 monster.

With "tiles", they can even update the IP blocks separately from each other, give it a better iGPU, NPU, whatever. Something better than increasing clocks by 100-200 MHz.

 

[EzzyB]

I think Intel is banking on it's 1.8A process being available and that's the rumor we got (it IS all just rumors at this point.)

With no date on that slide it's kind of nothing but a curiosity. We don't know if it was created before or after a rumor that Intel had cancelled the refresh. So, 🤷‍♂️

 

[Eximo]

If nothing else, it would be a good test bed even if it is a limited release. I could see them repeating the Broadwell scenario.

 

[logainofhades]

Intel needs to stop with the constant socket changes. One of the reasons AMD is so popular, even when performance is roughly even between the two, is upgrade path.

 

[EzzyB]

Intel needs to stop with the constant socket changes. One of the reasons AMD is so popular, even when performance is roughly even between the two, is upgrade path.

You know this came up in another forum. So I looked back on my unusually long history with personal computers (nearly all of it's history actually.)

I could only come up with replacing the 8088 on an IBM PC with an NEC version (circa 1986) as the only time I ever had a computer and replaced the processor with another in the same socket.

It's nearly impossible to do because pretty much everything advances, not just the CPU. On the other hand I've generally bought CPU's that were the top-of-the line at the time. So I can see someone who bought something like an Intel I-3 replacing it with an I-7 or I-9 a couple of years later.

It's just never been an issue for me, you build a new PC you get a new MB, drives, graphics, etc. because they all have advanced since the last one. I AM looking over that the old Lian Li 1000B battlewagon that's on it's 4th PC though. 😁

 

[logainofhades]

You know this came up in another forum. So I looked back on my unusually long history with personal computers (nearly all of it's history actually.)

I could only come up with replacing the 8088 on an IBM PC with an NEC version (circa 1986) as the only time I ever had a computer and replaced the processor with another in the same socket.

It's nearly impossible to do because pretty much everything advances, not just the CPU. On the other hand I've generally bought CPU's that were the top-of-the line at the time. So I can see someone who bought something like an Intel I-3 replacing it with an I-7 or I-9 a couple of years later.

It's just never been an issue for me, you build a new PC you get a new MB, drives, graphics, etc. because they all have advanced since the last one. I AM looking over that the old Lian Li 1000B battlewagon that's on it's 4th PC though. 😁

AM4 broke the mold on replacing everything. 1st-3rd gen Ryzen owners are still dropping in 5700x3d's into existing systems to get more life out of them, and getting basically the same gaming performance as a non x3d Zen4 chip. I went from a 3700x to a 5800x in my AMD rig myself, because what I play is so CPU dependent.

 

[EzzyB]

AM4 broke the mold on replacing everything. 1st-3rd gen Ryzen owners are still dropping in 5700x3d's into existing systems to get more life out of them, and getting basically the same gaming performance as a non x3d Zen4 chip. I went from a 3700x to a 5800x in my AMD rig myself, because what I play is so CPU dependent.

From what I've seen in the last 5 years or so, the only thing significantly CPU affected is 1080 gaming. And that strikes me as kind of ridiculous. Getting 168 FPS instead of 162 FPS is.... uh, what? I mean it's measurable, but is it even noticeable to the average human being?

I think, as a general rule, certainly there can be exceptions, CPU performance on gaming is somewhat over-valued these days. It almost isn't measurable at resolutions higher than 1080. Or maybe I'm just stuck in the old days when 40FPS was what we all strived for.

None of this is rhetorical BTW. My 9700K/1660 Super and 75hz monitor isn't exactly state-of-the art, but I built what I could during the COVID lock-in. (Definitely a new build coming in the new year.)

 

[logainofhades]

From what I've seen in the last 5 years or so, the only thing significantly CPU affected is 1080 gaming. And that strikes me as kind of ridiculous. Getting 168 FPS instead of 162 FPS is.... uh, what? I mean it's measurable, but is it even noticeable to the average human being?

I think, as a general rule, certainly there can be exceptions, CPU performance on gaming is somewhat over-valued these days. It almost isn't measurable at resolutions higher than 1080. Or maybe I'm just stuck in the old days when 40FPS was what we all strived for.

None of this is rhetorical BTW. My 9700K/1660 Super and 75hz monitor isn't exactly state-of-the art, but I built what I could during the COVID lock-in. (Definitely a new build coming in the new year.)

That is no longer the case.

View: https://www.youtube.com/watch?v=98RR0FVQeqs

 

[EzzyB]

That is no longer the case.

View: https://www.youtube.com/watch?v=98RR0FVQeqs

Thanks, I'll check it out!

 

[thestryker]

AM4 broke the mold on replacing everything. 1st-3rd gen Ryzen owners are still dropping in 5700x3d's into existing systems to get more life out of them, and getting basically the same gaming performance as a non x3d Zen4 chip. I went from a 3700x to a 5800x in my AMD rig myself, because what I play is so CPU dependent.

AM4 was an anomaly due to how far behind AMD. was starting from. I think that's part of the reason people have been disappointed by the jump from Zen 4 to Zen 5 when the gains are mostly a typical generational gain.

Thanks, I'll check it out!

Written version if you prefer: https://www.techspot.com/article/2837-cpu-performance-4k-gaming/

 

[thestryker]

Intel never really comments on socket longevity, but I think this one is one they likely should have. It's definitely possible there will be more CPUs on LGA 1851, but I certainly wouldn't count on it.

We know MTL-S was supposed to be the first CPU family on the socket and that was canceled.

Intel did keep the same physical socket from LGA 1156-1200 so if whatever was to come next is a situation similar to that it should be possible to release another CPU on 1851.

 

[dalek1234]

Sounds like somebody (Intel?) is trying to convince us that 1851 socket will have two CPU generations (when in actuality it will have only one), but without actually stating that themselves. Arrow Lake refresh was cancelled. This rumor/leak is bogus.

If Intel planned two CPU generations for this 1851 socket, why would they not come out and state that already? Such info would be welcomed by consumers.

 

[nogaard777]

AM4 was an anomaly due to how far behind AMD. was starting from. I think that's part of the reason people have been disappointed by the jump from Zen 4 to Zen 5 when the gains are mostly a typical generational gain.

People love to put AM4s longevity on a pedestal but in reality it was a nightmare that magically none of the fanboys remember.

First off, only AMD fans or budget users bought 300 series because AMD was still WELL behind at the time. You bought AMD because you couldn't afford Intel. Not because you magically knew the future that you would get 3 gens. The vast majority of users didn't even pay attention to Ryzen until Zen2, at which time AMD was still well behind.

No the only time people actually wanted Ryzen instead of just being what they could afford was Zen3. Now when you look at the huge gap between Zen1 and Zen3 then yeah, it's big. Although not just because Zen3 was decent, but because Zen1 and 2 sucked. And let's not forget you AMD faithful very nearly got screwed because AMD TRIED to not support your 300 and 400 series boards until the backlash made them admit incompatibility was a lie all along, and even then made 400 users wait 6 months for BIOS, and 300 owners wait an insulting 18 MONTHS.

That's not being the "good guy." That's Apple level tantrum compliance the 73rd time the EU sues them. NO BIOS takes 6 months to write, let alone 18. That's AMD still trying to force you into buying a 500 series board that would get ZERO more generations. That's a straight up dick move and the fact their fanboys actually applaud them for that sh*t is hilarious. And let's not forget how many generations the all-benevolent AMD gave TRX40 buyers, and that was a THOUSAND DOLLAR motherboard.

While it's good AMD finally did the right thing by it's users, AM4 was only amazing if you were poor, then later on weren't poor so you could afford a $450 X3D and $800(!) 5950x. You also need the memory of a goldfish.

 

[logainofhades]

Ryzen 1xxx was not that far behind. It was on par with Haswell, with regards to IPC. Intel with their meager gen over gen gains didn't give them a massive lead. Hence why so many held onto their 2600k's for so long. I was running a 6700k myself back then, and other than better multitasking, it wasn't that much of a leap over my 3570k. A gen 1 Ryzen 7 thrashed Intel in multithreaded tasks, which was a big selling point. They aged better than the Intel chips from that time too. Hence why in more modern gaming titles, a Ryzen 5 1600 can beat it's core i5 7600k counterpart.