• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Intel Readies Two 22-Core "Nova Lake-S" SKUs with 108 MB bLLC Cache for Gaming

erek

erek

Fully [H]
2FA
Joined
Dec 19, 2005
Messages
17,483
See: "TSMC N2 Pricing
https://www.techspot.com/news/105008-tsmc-could-charge-over-30000-each-2nm-wafer.html TLDR..."

"As far as the exact date, the time from a tapeout to final product is months away. Right now, the taped-out silicon tile is being powered on in Intel's labs and tested, running various test cases that stress out the silicon for multiple use cases and check for correctness of operation. Typically, power on takes a few weeks to a month to achieve, and final high-volume manufacturing will commence only a few months later. From that point, another two to three months are needed for manufacturing and shipping the product, meaning that Q3 of 2026 is the most likely target for Nova Lake-S. As a reminder, the CPU will combine 52 cores (16 P-cores, 32 E-cores, and four LPE-cores) paired with 8,800 MT/s memory controller and Xe3 "Celestial" for graphics rendering and Xe4 "Druid" for media and display duties, making it definitely an interesting product, as well as a difficult manufacturing target due to the heterogenous complexity."

1752254463759.png


https://www.techpowerup.com/338867/intel-nova-lake-s-tapes-out-on-tsmc-n2-node
 
A bit of a deception, still using TSMC for second half of 2026 products...
 
The article says a mix of TSMC N2 and Intel 18A is expected for the compute tiles. It could be a hedge against 18A problems, but with all the trade spats going on lately they might do this even if 18A was "perfect". They'll get a lower tariff rate from some countries fabbing chips in Taiwan. Certainly from China. China treats stuff like CPUs as coming from the country where they were fabbed, so Intel could have TSMC fab the tiles in Taiwan, send them to Arizona for packaging, then ship them to China and bring them in duty free since China treats Taiwan as being part of China as far as import duties go. I can't imagine it would be worth paying TSMC to save 10% on import duties, but who knows what's going to happen with tariffs.

As long as 18A works and they can make most of them in-house it's also kind of a good thing to do from a fab services marketing perspective. It shows they can make the same thing with their own process and TSMC's. It's not as good as someone else dual sourcing chips, but it's better than nothing. There are a lot of companies and governments that would like to dual source chips. That's how AMD got its x86 license. Intel wanted a contract with the Department of Defense, and the DoD insisted on having 2 suppliers.

If they are dual sourcing I'd expect most of the Nova Lake chips available in the US to have Intel 18A compute tiles.

It's also possible they're using different processes for different models. There are some rumors floating around that 18A will be less power efficient but clock higher than TSMC 2N. So TSMC might get the thin & light laptop CPUs while 18A gets used for desktop, server and high performance laptops.
 

"Intel Readies Two 22-Core "Nova Lake-S" SKUs with 108 MB bLLC Cache for Gaming​

by AleksandarK Today, 12:46 Discuss (10 Comments)
According to well-known Intel leaker Jaykihn, Intel is preparing two "Nova Lake-S" CPU models with a 22-core configuration, enhanced by the big Last Level Cache (bLLC). These processors are part of the Core Ultra 5 400S family, and Intel is developing two mid-range configurations, each with 22 cores. Both models will feature 6 "Coyote Cove" P-Cores, 12 "Arctic Wolf" E-Cores, and 4 LPE-Cores. This design includes a single compute die aimed at mid-range gaming rigs. Alongside the 22 cores, Intel plans to add an extra 108 MB of bLLC cache to boost gaming performance, similar to AMD's approach with its Ryzen X3D gaming CPUs, which currently lead in gaming performance. Intel aims to match raw CPU compute power with newer CPU cores and use bLLC to improve latency-sensitive tasks like gaming. The only difference between the two new models is that one is an unlocked 125 W TDP version, part of the "K" overclockable family, while the other is a 65 W base TDP model.

Recently, it has been observed that Intel's upcoming Z9x0 platform for "Nova Lake-S" will support up to 52 cores in a single LGA-1954 socket. Some early rumors suggest a PL2 power draw of 474 W, which is a significant increase over previous designs. This is understandable, considering that the design brings workstation-like core counts into the consumer sector for the first time. Regarding the socket, Intel's LGA-1954 has been pictured with a 2-lever independent loading mechanism (ILM), similar to what Intel's HEDT sector uses. This is typically reserved for larger CPU packages, and with the pin count now nearing 2,000 socket pins, it is a necessary choice for Intel. Importantly, the first "Nova Lake-S" CPU SKUs are scheduled for release in just a few months, likely leading to more rumors as we approach the final launch."
 
erek said:

"Intel Readies Two 22-Core "Nova Lake-S" SKUs with 108 MB bLLC Cache for Gaming​

by AleksandarK Today, 12:46 Discuss (10 Comments)
According to well-known Intel leaker Jaykihn, Intel is preparing two "Nova Lake-S" CPU models with a 22-core configuration, enhanced by the big Last Level Cache (bLLC). These processors are part of the Core Ultra 5 400S family, and Intel is developing two mid-range configurations, each with 22 cores. Both models will feature 6 "Coyote Cove" P-Cores, 12 "Arctic Wolf" E-Cores, and 4 LPE-Cores. This design includes a single compute die aimed at mid-range gaming rigs. Alongside the 22 cores, Intel plans to add an extra 108 MB of bLLC cache to boost gaming performance, similar to AMD's approach with its Ryzen X3D gaming CPUs, which currently lead in gaming performance. Intel aims to match raw CPU compute power with newer CPU cores and use bLLC to improve latency-sensitive tasks like gaming. The only difference between the two new models is that one is an unlocked 125 W TDP version, part of the "K" overclockable family, while the other is a 65 W base TDP model.

Recently, it has been observed that Intel's upcoming Z9x0 platform for "Nova Lake-S" will support up to 52 cores in a single LGA-1954 socket. Some early rumors suggest a PL2 power draw of 474 W, which is a significant increase over previous designs. This is understandable, considering that the design brings workstation-like core counts into the consumer sector for the first time. Regarding the socket, Intel's LGA-1954 has been pictured with a 2-lever independent loading mechanism (ILM), similar to what Intel's HEDT sector uses. This is typically reserved for larger CPU packages, and with the pin count now nearing 2,000 socket pins, it is a necessary choice for Intel. Importantly, the first "Nova Lake-S" CPU SKUs are scheduled for release in just a few months, likely leading to more rumors as we approach the final launch."
Click to expand...
This is the kind of shit I love seeing, I mean I would love additional memory channels, but let's face it, Windows (non-server editions) and DX12 aren't really optimized for more than 2, and I doubt Microsoft's ability to optimize the platforms for more than 2 channels in a reasonable time frame if Intel and AMD actually did put out consumer parts with more than 2...

So adding extra cache to cut down on the memory latency is about the best workaround there is.

But we all should take a moment to admire the work Intel has been doing with power delivery.... 474W... really, the fact they can put that kind of juice through their silicon and not have it ignite is an engineering feat onto itself...
 
erek said:

"Intel Readies Two 22-Core "Nova Lake-S" SKUs with 108 MB bLLC Cache for Gaming​

by AleksandarK Today, 12:46 Discuss (10 Comments)

Recently, it has been observed that Intel's upcoming Z9x0 platform for "Nova Lake-S" will support up to 52 cores in a single LGA-1954 socket. Some early rumors suggest a PL2 power draw of 474 W, which is a significant increase over previous designs. This is understandable, considering that the design brings workstation-like core counts into the consumer sector for the first time. Regarding the socket, Intel's LGA-1954 has been pictured with a 2-lever independent loading mechanism (ILM), similar to what Intel's HEDT sector uses. This is typically reserved for larger CPU packages, and with the pin count now nearing 2,000 socket pins, it is a necessary choice for Intel. Importantly, the first "Nova Lake-S" CPU SKUs are scheduled for release in just a few months, likely leading to more rumors as we approach the final launch."
Click to expand...
Can AMD not have a competitive response. New socket? Chipset? 2027? 2028? later?
 
philb2 said:
Can AMD not have a competitive response. New socket? Chipset? 2027? 2028? later?
Click to expand...
zen6 early result seem great and they will have mobile chips up to 256 cores server one line up with it
 
I don't know about the 5800X3D being the GOAT. It was a good pre-cursor to the 7800X3D that overtook the gaming CPU crown from Intel. Still overpriced when it shouldn't be. It doesn't perform better in gaming than a stock 12900K or 13600K that sells for way less. Forget about a 13600K that can be overclocked to 5.8GHz - 5.4GHz at 1.35V. I'd consider the 7800X3D more the GOAT than the 5800X3D for gaming, but that got it's butt handed to it by the 9800X3D.
 
$350 5800x3d on ddr4-3600 matching the $600-12900k DDR5-6000 setups was quite something at the time would say not worse than the $450 7800x3d on similar ddr5 beating by a bit the $550 13900k, the price difference on the 5800x3d with the whole platform taken into account was much bigger.
 
Last edited:
Randall Stephens said:
AMD doesn’t need a response, they already brought back the GOAT. People don’t want ddr5 gaming platforms, they want to use what they have for ram already so that they have budget for GPU and Storage.
Click to expand...
Do we believe Nova Lake S makes zen 6 non competitive or what am I missing? Intel adding memory bandwidth, cores and cache vs AMD adding memory bandwidth (dual IMC), cache and cores and adding sea-of-wires. No idea how things will shake out, but I don't see this as a Core 2 moment for Intel.
 
Why so many E cores?
Thought these things handled the background tasks.
Is the load of background tasks under windows increasing that much
 
Jtal said:
Why so many E cores?
Thought these things handled the background tasks.
Is the load of background tasks under windows increasing that much
Click to expand...
I think they are working on deleting all e cores in the future but we are not there yet.
 
Jtal said:
Why so many E cores?
Thought these things handled the background tasks.
Is the load of background tasks under windows increasing that much
Click to expand...
Not really for background task, it is mostly to add a lot of core per mm of die. Outside fancy AVX-512 level of newer instruction they can do everything a p-core can do. background task was marketing talking point, they have nothing specially good for those and originaly they were not even more efficiant than p-core.

The strategy for desktop/workstation to have a lot of e-core once you reach say 16 pcore do make a lot of sense, not many workload on those platform that can use more than 16 cores will not prefer 3/4 time more e-core than p-core after that amount, server is a different story.

Epyon said:
I think they are working on deleting all e cores in the future but we are not there yet.
Click to expand...
We could say working on deleting all p cores... the new unified core will be more a descandant of the e-core and made by the e-core team than the other way around ?

It is the e-core team winning the battle here, being much better and modern.
 
DooKey said:
Show me some gaming benches and then we'll talk.
Click to expand...
Carefull, you will get useless 720P benches with an RTX 5090.

No one will bench like people game:
4K with maximum settings, because that will not get clicks, because the spread will be even.
Last site to do proper gaming benching was Tom's:
1783227753607.png


But that dosn't look a "thrilling" as 1080p benches (Who The Fuck games like that?!):
1783227849828.png


TL;DR
Avarage users are too stupid and we all suffer from it.
 
littlegreenmenfromhome said:
Carefull, you will get useless 720P benches with an RTX 5090.

No one will bench like people game:
4K with maximum settings, because that will not get clicks, because the spread will be even.
Last site to do proper gaming benching was Tom's:
View attachment 813332

But that dosn't look a "thrilling" as 1080p benches (Who The Fuck games like that?!):
View attachment 813333

TL;DR
Avarage users are too stupid and we all suffer from it.
Click to expand...

You don’t benchmark a CPU at ultra-high resolution if you’re trying to compare CPUs. A chart full of lines that look the same because every CPU is bottlenecks by the GPU anyway doesn’t tell you anything about how the CPUs themselves compare.
 
littlegreenmenfromhome said:
Last site to do proper gaming benching was Tom's:
Click to expand...
techpowerup usually do 4k bench (with some popular RT title)....
https://www.techpowerup.com/review/amd-ryzen-9-9950x3d2-dual-edition/20.html

Hardware unboxed went in dept on 9800x3d and 4k gaming:

View: https://www.youtube.com/watch?v=5GIvrMWzr9k

As for your example you gave above, people that go into reviews and buy very expensive system, not sure if they play at 45 fps 1% low / under 60 fps average either ;) lot of that type of people will setting down to reach at least 80/90 fps even on single player game.
 
HockeyJon said:
You don’t benchmark a CPU at ultra-high resolution if you’re trying to compare CPUs. A chart full of lines that look the same because every CPU is bottlenecks by the GPU anyway doesn’t tell you anything about how the CPUs themselves compare.
Click to expand...
You always test all resolutions. 1080P resolution to compare CPU performance and potential bottlenecks at 1080P, as well as 2K and 4K to compare GPU performance where CPU bottlenecks become a moot point except for 1% and 0.1% lows. I wish they tested CPU and GPU combinations, but smaller reviewers usually test those use cases.

I don't see anyone with an RTX 5090 using a Ryzen 3 CPU. But anyone with a Ryzen 3 CPU might wonder if an RTX 5060 Ti 16GB might be bottlenecked at 2K.
 
El_Capitan said:
You always test all resolutions. 1080P resolution to compare CPU performance and potential bottlenecks at 1080P, as well as 2K and 4K to compare GPU performance where CPU bottlenecks become a moot point except for 1% and 0.1% lows. I wish they tested CPU and GPU combinations, but smaller reviewers usually test those use cases.

I don't see anyone with an RTX 5090 using a Ryzen 3 CPU. But anyone with a Ryzen 3 CPU might wonder if an RTX 5060 Ti 16GB might be bottlenecked at 2K.
Click to expand...

If you want to compare CPU performance, then testing it in an environment where the GPU is the limiting factor doesn’t generate a valid comparison. The suggesting that no one does 4K benchmarks because the data is boring is not correct. A lot of sites don’t do them because it doesn’t provide a valid comparison of CPU performance because the CPU isn’t being taxed at full when it’s waiting for the GPU to keep up. This isn’t because it generates a “less thrilling” comparison, and it isn’t because most users, or reviewers, are stupid.

In your latter example, you’re looking at a GPU bottleneck, not a CPU bottleneck. That’s a different testing procedure. If a 5060 Ti is a bottleneck already, then the CPU you have doesn’t matter much, which is the same reason why the 4K data is often not presented in CPU testing (TechPowerUp being an exception). It takes time to generate this data, and everyone already knows there is little difference between CPUs at 4K in the vast majority of situations. Most reviewers don’t want to waste their time showing everyone that 4K gaming doesn’t benefit much from a CPU upgrade unless you’re doing a 4A game like Civ, but that’s because that type of game is more CPU dependent in general. Everyone already knows that.
 
  • Like
Reactions: Jtal
like this
HockeyJon said:
If you want to compare CPU performance, then testing it in an environment where the GPU is the limiting factor doesn’t generate a valid comparison. The suggesting that no one does 4K benchmarks because the data is boring is not correct. A lot of sites don’t do them because it doesn’t provide a valid comparison of CPU performance because the CPU isn’t being taxed at full when it’s waiting for the GPU to keep up. This isn’t because it generates a “less thrilling” comparison, and it isn’t because most users, or reviewers, are stupid.

In your latter example, you’re looking at a GPU bottleneck, not a CPU bottleneck. That’s a different testing procedure. If a 5060 Ti is a bottleneck already, then the CPU you have doesn’t matter much, which is the same reason why the 4K data is often not presented in CPU testing (TechPowerUp being an exception). It takes time to generate this data, and everyone already knows there is little difference between CPUs at 4K in the vast majority of situations. Most reviewers don’t want to waste their time showing everyone that 4K gaming doesn’t benefit much from a CPU upgrade unless you’re doing a 4A game like Civ, but that’s because that type of game is more CPU dependent in general. Everyone already knows that.
Click to expand...
You should bench as you game, otherwise your benches are useless.
 
littlegreenmenfromhome said:
You should bench as you game, otherwise your benches are useless.
Click to expand...

The objective of a review is to compare the performance of the part. 4K results don't do this for CPUs because the GPU is the limiting factor. It's a pointless comparison if you're trying to establish which part is better overall. That's the objective of the review. It's not done at 720p because people are stupid or to generate clicks, it's done to remove bottlenecks so you actually understand the differences. To remove that bottleneck, you need an overpowered GPU to make the CPU the limiting factor. It is not "useless information", it's actually the best basis of comparison and how the review is supposed to be done because it isolates the CPU as the factor you're measuring.
 
The question of what to measure is indeed interesting. IMHO, there are reasonable debates on what should be measured and how. What I think is important is we are explicit about what is being measured and the limitations such measurements have in terms of performance inferences, which I think is an under analyzed aspect of the bench marking community.
 
littlegreenmenfromhome said:
You should bench as you game, otherwise your benches are useless.
Click to expand...
I game at 4K and I agree. With that said it appears that the GPUs are still the bottleneck for the most part. CPU performance overall matters if you do more than game. I'm a gamer so I go with the most powerful GPU/CPU available for the most part.
 
1.1.2.3.5... said:
The question of what to measure is indeed interesting. IMHO, there are reasonable debates on what should be measured and how. What I think is important is we are explicit about what is being measured and the limitations such measurements have in terms of performance inferences, which I think is an under analyzed aspect of the bench marking community.
Click to expand...

I understand the point being made, but it tells you absolutely nothing about which processor is superior. The focus remains on eliminating bottlenecks so you can test the actual part. Otherwise, the conclusion will be "all processors are the same" if you're just looking at 4K, which is almost correct as far as 4K is concerned since the GPU is what's limiting it anyway (there are exceptions, but even then, it's seldom enough to matter).

In any case, it's practically a moot point, since the idea that a 4K gamer is looking to cut corners is a bit laughable. The up-front investment is large enough that you're not looking to save cash on the CPU, let's be honest. If you're gaming at 4K, you're looking at a top spec, even though you know the difference is minimal. Saving $100-$200 isn't moving the needle on that setup.
 
HockeyJon said:
I understand the point being made, but it tells you absolutely nothing about which processor is superior.
Click to expand...
Many non-enthusiasts like us don't care which processor is superior. They ask what they should buy to play x games with y settings, and z budget. If they have a $1000 budget for just a CPU/MB and GPU, getting them a 9800X3D already puts them back $479. That leaves them with what, a cheap MB for $100, and an RTX 5060 Ti for $330? Or, find an RX 9070 XT for $650, a cheap MB for $100, leaving them with $250 for a CPU. I found a used 270K Plus and used MB for about $350.

I think everyone here would agree, that a 270K Plus with an RX 9070XT beats a 9800X3D with an RTX 5060 Ti at 4K, 2K, and even 1080P.
 
El_Capitan said:
Many non-enthusiasts like us don't care which processor is superior. They ask what they should buy to play x games with y settings, and z budget. If they have a $1000 budget for just a CPU/MB and GPU, getting them a 9800X3D already puts them back $479. That leaves them with what, a cheap MB for $100, and an RTX 5060 Ti for $330? Or, find an RX 9070 XT for $650, a cheap MB for $100, leaving them with $250 for a CPU. I found a used 270K Plus and used MB for about $350.

I think everyone here would agree, that a 270K Plus with an RX 9070XT beats a 9800X3D with an RTX 5060 Ti at 4K, 2K, and even 1080P.
Click to expand...

Those guys aren't reading CPU reviews on hardware websites, either. They're probably calling Dell and having one made for them. They also probably aren't demanding a 4K benchmark.

Like I said, I understand why people want to see things done at the resolution they're playing at, but the original comment I was responding to was saying CPUs aren't benched at 4K by most reviewers because people are stupid. The reality is, most reviewers don't want to spend the time on it because it doesn't tell you anything in the end. The review will conclude with "we tested 20 CPUs, and 18 of them are identical, so just buy whatever's cheap I guess". This isn't useful information, nor is it accurate, nor do the people gaming at 4K not understand how this works, nor are they operating in a limited budget where any of this matters.
 
HockeyJon said:
Those guys aren't reading CPU reviews on hardware websites, either. They're probably calling Dell and having one made for them. They also probably aren't demanding a 4K benchmark.

Like I said, I understand why people want to see things done at the resolution they're playing at, but the original comment I was responding to was saying CPUs aren't benched at 4K by most reviewers because people are stupid. The reality is, most reviewers don't want to spend the time on it because it doesn't tell you anything in the end. The review will conclude with "we tested 20 CPUs, and 18 of them are identical, so just buy whatever's cheap I guess". This isn't useful information, nor is it accurate, nor do the people gaming at 4K not understand how this works, nor are they operating in a limited budget where any of this matters.
Click to expand...
But for us that do, we appreciate the review sites that do include 2K and 4K like TechPowerUp does. It tells us what we all should already know. However, every comment section is littered with, "just get a 9800X3D or 5800X3D and whatever GPU you can afford". Then they come back and ask why their 2K/4K performance is bad. Because the same individuals just point to every review site and say, "look, you have the best gaming CPU now, all you need now is an RTX 5090".
 
One big thing about testing CPU at 720p in the past, was that the GPU limitation would often change greatly during its useful life time.

People that bought an intel 4790 would sometime upgrade GPU twice before the cpu, making it quite valuable to look at low res benchmark to have some idea of the futur relative CPU performance in more demanding scenarios, because their gpu will be in 4-5 years twice as fast than the one they have right now, maybe a 4790 looked useless at my native resolution with a Nvidia GTX 680, but once upgraded to a 1080 with more recents games the 4790 > over a core I3 alternative of the time became really big.

GPU cycle slowing down (and CPU picking up again) changed that math a bit, often have similar upgrade cycle now in time and performance change, making it a bit less interesting.

And from time to time, a title-engine able to use cpu more than before come out, say spider man Remaster with high raytracing enabled, CPU had to work a lot to feed the fastest GPU on that one even at 4k, https://www.techspot.com/article/2520-spiderman-cpu-benchmark/, making the gap at 4k between a still really good at the time 10600k/3800x and the top one 5800x3d/12900k exist all of the sudden. Or baldur gates 3 that stay CPU sensitive even at 4k/ultra high details, without being a an ultra high FPS affair.
 
Last edited:
El_Capitan said:
But for us that do, we appreciate the review sites that do include 2K and 4K like TechPowerUp does. It tells us what we all should already know. However, every comment section is littered with, "just get a 9800X3D or 5800X3D and whatever GPU you can afford". Then they come back and ask why their 2K/4K performance is bad. Because the same individuals just point to every review site and say, "look, you have the best gaming CPU now, all you need now is an RTX 5090".
Click to expand...

I mean that's just terrible advice if that's what people are saying, because everyone knows higher resolutions are more GPU limited scenarios.

In any case, I don't disagree about the convenience of it, but read the comment I was responding to. The proper comparison is 720p, and it's not done because people are stupid. 4K testing doesn't tell you a whole lot about relative CPU performance, but 720p does.
 
littlegreenmenfromhome said:
Carefull, you will get useless 720P benches with an RTX 5090.

No one will bench like people game:
4K with maximum settings, because that will not get clicks, because the spread will be even.
Last site to do proper gaming benching was Tom's:
View attachment 813332

But that dosn't look a "thrilling" as 1080p benches (Who The Fuck games like that?!):
View attachment 813333

TL;DR
Avarage users are too stupid and we all suffer from it.
Click to expand...
It’s a holdover from the days we used to get a new Gpu release twice a year. The raw cpu power mattered because you were going to remove a bottleneck when you upgraded your gpu. These days, I agree with you, not so much a CPU issue except sim games it seems.
 
Randall Stephens said:
It’s a holdover from the days we used to get a new Gpu release twice a year. The raw cpu power mattered because you were going to remove a bottleneck when you upgraded your gpu. These days, I agree with you, not so much a CPU issue except sim games it seems.
Click to expand...


I feel we still keep our CPUs (especially AMD CPUs) far longer than a graphics card.

I upgrade plenty of machines running 3600 and 5600 CPUs to newer GPUs.
 
HockeyJon said:
I understand the point being made, but it tells you absolutely nothing about which processor is superior. The focus remains on eliminating bottlenecks so you can test the actual part. Otherwise, the conclusion will be "all processors are the same" if you're just looking at 4K, which is almost correct as far as 4K is concerned since the GPU is what's limiting it anyway (there are exceptions, but even then, it's seldom enough to matter).

In any case, it's practically a moot point, since the idea that a 4K gamer is looking to cut corners is a bit laughable. The up-front investment is large enough that you're not looking to save cash on the CPU, let's be honest. If you're gaming at 4K, you're looking at a top spec, even though you know the difference is minimal. Saving $100-$200 isn't moving the needle on that setup.
Click to expand...
Is that really true though? If all you're looking at is the average frame rate, sure. But there's a lot more nuance you can parse across the data. Lows, distribution of lows. Shader compilation time, sim games, etc...And, I would add running local models into the mix as well. I am not debating on your core point given how most benchmarks don't usually give you those details, but I am thinking of more holistic testing approaches and within those I think there is likely some worth while analysis of games running at 4k max settings. Maybe not, but I am not saying that is the only benchmark that should be included either.
 
1.1.2.3.5... said:
Is that really true though? If all you're looking at is the average frame rate, sure. But there's a lot more nuance you can parse across the data. Lows, distribution of lows. Shader compilation time, sim games, etc...And, I would add running local models into the mix as well. I am not debating on your core point given how most benchmarks don't usually give you those details, but I am thinking of more holistic testing approaches and within those I think there is likely some worth while analysis of games running at 4k max settings. Maybe not, but I am not saying that is the only benchmark that should be included either.
Click to expand...
I don't disagree with you one bit, we could do better, this just where most people seem to care, max frames, low lag, blah blah competitive online shooter BS...
 
1.1.2.3.5... said:
Is that really true though? If all you're looking at is the average frame rate, sure. But there's a lot more nuance you can parse across the data. Lows, distribution of lows. Shader compilation time, sim games, etc...And, I would add running local models into the mix as well. I am not debating on your core point given how most benchmarks don't usually give you those details, but I am thinking of more holistic testing approaches and within those I think there is likely some worth while analysis of games running at 4k max settings. Maybe not, but I am not saying that is the only benchmark that should be included either.
Click to expand...

That can be largely extrapolated from the 720p results though. The general trend would hold.

Like I said, I'm not saying there's zero merit to it at all, but look at the comment I was responding to. A lot of people seem to insist that lower resolution testing isn't valid since it isn't "real world use case" or whatever. They're missing that it's necessary in order to be able to establish the proper ranking for the CPUs being tested. It's not because "people are dumb".
 
HockeyJon said:
That can be largely extrapolated from the 720p results though. The general trend would hold.

Like I said, I'm not saying there's zero merit to it at all, but look at the comment I was responding to. A lot of people seem to insist that lower resolution testing isn't valid since it isn't "real world use case" or whatever. They're missing that it's necessary in order to be able to establish the proper ranking for the CPUs being tested. It's not because "people are dumb".
Click to expand...
No it cannot, the 1080p tells you nothing about 4K performance:

1080p would try to tell you there is big difference:
1783829298880.png


But 4K numbers tells you you were not testing it like you play it and 1080p is useless in that regard:

1783829360355.png


Tell the meaningfull difference between these CPU at 4K:
Ryzen 9 3950X
Ryzen 9 5950X
Ryzen 7 5800X Modified
Core i9-10900K
Ryzen 9 5600X
Ryzen 7 3800XT Modified
Ryzen 7 3800XT
Ryzen 9 5900X
Core i7-10700K
Ryzen 7 3900XT
Ryzen 7 5600X Modified
Ryzen 9 5800X
Core i7-10600K
Core i7-9700K
Core i7-9900K
Ryzen 3 3300X
Ryzen 5 3600XT

(Eg. like try and compare the Ryzen 3 3300X to the Ryzen 9 5950X....HINT-HINT....)

Because 1080p says there is a BIG difference...real world benchmarks rules, 720p/1080p are useless for gaming related stuff.
 
Last edited:
You must log in or register to reply here.
Back
Top