lexluthermiesterThat's one take. Another would be someone buying the best because they can afford it. It isn't important how practical or affordable it is, so long as it's the best. Kinda like the argument of a 4070 vs 4080 vs 4090. One of those is the smart buy. Another is less so but not completely outrageous. The last is crazy expensive, but is the best, full stop.

See the point there?

That's different because the 4090 offers measurable gains over the rest, it's not just a name and a shiny box.

AusWolfThat's different because the 4090 offers measurable gains over the rest, it's not just a name and a shiny box.

And that's true, but not really the point. Likewise the Ryzen 3900X and 3900XT had marginal difference[less than 3%], but the XT was marketed to the same kind of consumer. At the time the 3900XT was the best of the 12core CPU's and it appealed to that kind of consumer.

Here's another thought: These CPU's show that manufacturers are constantly striving for excellence. This is only ever a good thing for the rest of the buying public who can't[or don't want to] afford the top tier models.

lexluthermiesterAnd that's true, but not really the point. Likewise the Ryzen 3900X and 3900XT had marginal difference[less than 3%], but the XT was marketed to the same kind of consumer. At the time the 3900XT was the best of the 12core CPU's and it appealed to that kind of consumer.

And that's different because the 3900XT launched with the same MSRP as the 3900X did before. I also did get a 6750 XT for the same price as I would've a 6700 XT half a year earlier. Now, that's what I call striving for excellence. Upselling a product by let's say 10% that comes with a 1% clock speed advantage is not. It's just marketing.

lexluthermiesterHere's another thought: These CPU's show that manufacturers are constantly striving for excellence.

...profit. Corrected that for you. :)

lexluthermiesterHere's another thought: These CPU's show that manufacturers are constantly striving for excellence.

Turning a knob on a product one notch further, taking one more step into hot hell of inefficiency and energy waste, which everybody else tries to leave, is that really striving for excellence?

AusWolfAnd that's different because the 3900XT launched with the same MSRP as the 3900X did before. I also did get a 6750 XT for the same price as I would've a 6700 XT half a year earlier. Now, that's what I call striving for excellence. Upselling a product by let's say 10% that comes with a 1% clock speed advantage is not. It's just marketing.


...profit. Corrected that for you. :)

BoggledBeagleTurning a knob on a product one notch further, taking one more step into hot hell of inefficiency and energy waste, which everybody else tries to leave, is that really striving for excellence?

Guys, I'm not trying to be rude or combative with you here. You're missing the point and I can't put it in any simpler terms. I can't tell whether you're purposefully side-stepping the issue or it's just not registering to you, but it's really very simple.

At the end of the day, if you don't like it, don't buy it. Someone else will.

lexluthermiesterGuys, I'm not trying to be rude or combative with you here. You're missing the point and I can't put it in any simpler terms. I can't tell whether you're purposefully side-stepping the issue or it's just not registering to you, but it's really very simple.

At the end of the day, if you don't like it, don't buy it. Someone else will.

I might be missing some point, but my point is that the increase in performance ideally has to outweigh the increase in price, or at least go hand in hand in the worst case. If the increase in price outweighs the increase in performance, then it's nothing more than corporate greed. Combine it with a performance gain as small as it is here, between the K and KS models, and there is no logical reason left to buy the more expensive product. It's pure vanity.

AusWolfIt's pure vanity.

Yes. Hence was my prior supercar analogy. It’s a flex CPU. Sometimes the price is irrelevant, performance increase is irrelevant. What matters to some people is simply having the “best”. Even if it’s the “best” by 1-3% while being two times more money.

There is not much else to do here but :roll: :slap::kookoo::rolleyes:

OnasiYes. Hence was my prior supercar analogy. It’s a flex CPU. Sometimes the price is irrelevant, performance increase is irrelevant. What matters to some people is simply having the “best”. Even if it’s the “best” by 1-3% while being two times more money.

Except that you can't buy an "almost supercar" for a bit less money, but you can buy a 14900K, or non-K, and have the exact same experience.

Edit: A supercar costs a lot because it's a totally different design than your everyday runabout. It's got advanced suspension, a big engine, an aero optimised chassis, lots of tech, etc... but what does the 14900KS have? +100 MHz boost? C'mon! :rolleyes:

AusWolfExcept that you can't buy an "almost supercar" for a bit less money, but you can buy a 14900K, or non-K, and have the exact same experience.

Sure you can. For most practical purposes there would be very little difference between, say, an M2 or a Cayman and a full blown supercar like a 911 Turbo or a 720S. Both types of cars will be far more capable than 99.9% of their owners, even on the track. In real world, there are speed limits, so the difference will be null and void.

OnasiSure you can. For most practical purposes there would be very little difference between, say, an M2 or a Cayman and a full blown supercar like a 911 Turbo or a 720S. Both types of cars will be far more capable than 99.9% of their owners, even on the track. In real world, there are speed limits, so the difference will be null and void.

The real road argument is like saying that you buy anything faster than an i3 for web browsing. Of course it's pointless. On the track, though, you feel the difference between a Cayman and a 911 Turbo. You also feel the difference between an i5 and an i9 in heavy work. But there is no scenario whatsoever where you feel any difference between a 14900K and KS. None.

On the track? Depends on the driver. As I said, unless you are a semi-professional racer you will not extract significantly better lap-times from a Turbo versus a Cayman. Cayman is that good nowadays. But we are getting off-topic.
You seem to be under the impression that I am rationalizing or defending the KS. I do not. It’s a silly product. And yet my point is it will still absolutely sell. There are people who would buy one. Now, whether it’s a sustainable SKU is another question. I had a talk with @Dr. Dro about it and, as a KS owner himself, he reckons that with 14900KS Intel is pushing too far. He might be right on that. But spending time arguing about how little rational sense releases like this makes is irrational in itself.

OnasiOn the track? Depends on the driver. As I said, unless you are a semi-professional racer you will not extract significantly better lap-times from a Turbo versus a Cayman. Cayman is that good nowadays. But we are getting off-topic.
You seem to be under the impression that I am rationalizing or defending the KS. I do not. It’s a silly product. And yet my point is it will still absolutely sell. There are people who would buy one. Now, whether it’s a sustainable SKU is another question. I had a talk with @Dr. Dro about it and, as a KS owner himself, he reckons that with 14900KS Intel is pushing too far. He might be right on that. But spending time arguing about how little rational sense releases like this makes is irrational in itself.

My point is that with the right driver, under the right conditions, it is possible to extract more performance out of a 911 Turbo than a Cayman. The same cannot be said about the 14900KS. It's the exact same CPU as the K for every single user, under every condition, just at a higher price.

Will it sell? Oh, absolutely! People are stupid enough to believe in marketing placebo.

AusWolfMy point is that with the right driver, under the right conditions, it is possible to extract more performance out of a 911 Turbo than a Cayman. The same cannot be said about the 14900KS. It's the exact same CPU as the K for every single user, under every condition, just at a higher price.

Not… every condition. If we stretch the “pro driver on the track” analogy to its limits, then the 14900KS with its, ostensibly, almost perfect binning should be interesting to professional liquid nitrogen overclockers in terms of record potential.
Yeah, not a great selling point, admittedly, but it is what it is.

AusWolfI might be missing some point, but my point is that the increase in performance ideally has to outweigh the increase in price

Yeah, that's easily understood. Some people just don't care about that imbalance. They want the best and are willing to pay for it.

It's kinda like the Threadripper system I built for a guy over the holidays. Totally impractical, but in his eyes that was the best and what he wanted. He paid for the parts, I built it for him. He loves it and didn't bat an eye at the $8000+ price tag.
(BTW, THE most powerful system intended for personal use I've ever built to date!)

Different strokes for different folks is all I'm saying. Intel, AMD, Nvidia, etc, etc would be fools not to cater to that sector of people. After all look at Apple. Perfect example of over-priced tat that people see as "premium"...

wheresmycarlmao. Don't worry about me, unfortunately i got locked down with a one-chick deal for life. I can't leave the house without dropping a GPS ticker let alone dream about supercars and YOLO STDs.

Lucky! I thought I had it. I was so close to calling her Mrs.! But it wasn't meant to be :(

Hey, at least I have the KS chip so I got that going for me :D

AusWolfExcept that you can't buy an "almost supercar" for a bit less money, but you can buy a 14900K, or non-K, and have the exact same experience.

Edit: A supercar costs a lot because it's a totally different design than your everyday runabout. It's got advanced suspension, a big engine, an aero optimised chassis, lots of tech, etc... but what does the 14900KS have? +100 MHz boost? C'mon! :rolleyes:

I agree. The 13900KS did offer a nominal improvement over the 13900K, being binned a ton better on average, but this one just seems... excessive to me. It's really pointless, the clocks are so high it'll be a miracle if they're sustainable on a computer with standard cooling, binning or not.

Dr. Drothey're sustainable on a computer with standard cooling

They're not, VERY good liquid cooling is required for this model of CPU. Otherwise very swift thermal throttling will take place consistently.

EDIT: for full load max clocks..

Premium binning has always costed more and be worse on value for dollar in general. It's really nothing new. It's good enough chip, but the price premium for binning and higher overall power draw aren't viewed very favorably to the majority and hard to fault people on those particular criticisms.

There are scenario's where you can justify it if it's used for work to make them more money in the end, but for general usage the value for dollar isn't favorable and neither is it's stock efficiency. The latter is less important than the former though as you can defiantly improve it's general efficiency by forgoing a bit of peak stock performance expectations. Value is a key consideration and if particularly go it just makes it easy to forgo some performance in favor of efficiency and suddenly those talking points hold much less merit.

It's peak performance can and does matter as well at the same time for certain usage considerations especially so. For most people value for dollar is going to be the most important aspect and should be generally speaking. I got incredible value on 14700K so that was a huge factor in purchasing it overall and if anything the primary one rather than secondary one. It was great value for dollar for performance delivered and no one will convince me otherwise. Better performance for dollar certain does exist just not at that performance level for enough of the workloads I care about and longevity was factor I also took into consideration. I don't want to have to replace it too quickly because I didn't account for that and it starts to feel a bit underwhelming too quickly and it certainly won't in the case of the 14700K.

My biggest critique overall on 14900KS would be value for dollar looks rather poor overall, but we'll have to see what it actually retails at MSRP once the the initial dusts settles a bit. I don't expect it will be very favorable though given past history. The 14900K should certainly be higher value and similar enough performance while a 14600K or 14700K will be dramatically higher value for dollar it appears and more sensible for most buyers. It was never meant to be sensible though it's a halo product.

InVasManiPremium binning has always costed more and be worse on value for dollar in general. It's really nothing new. It's good enough chip, but the price premium for binning and higher overall power draw aren't viewed very favorably to the majority and hard to fault people on those particular criticisms.

There are scenario's where you can justify it if it's used for work to make them more money in the end, but for general usage the value for dollar isn't favorable and neither is it's stock efficiency. The latter is less important than the former though as you can defiantly improve it's general efficiency by forgoing a bit of peak stock performance expectations. Value is a key consideration and if particularly go it just makes it easy to forgo some performance in favor of efficiency and suddenly those talking points hold much less merit.

It's peak performance can and does matter as well at the same time for certain usage considerations especially so. For most people value for dollar is going to be the most important aspect and should be generally speaking. I got incredible value on 14700K so that was a huge factor in purchasing it overall and if anything the primary one rather than secondary one. It was great value for dollar for performance delivered and no one will convince me otherwise. Better performance for dollar certain does exist just not at that performance level for enough of the workloads I care about and longevity was factor I also took into consideration. I don't want to have to replace it too quickly because I didn't account for that and it starts to feel a bit underwhelming too quickly and it certainly won't in the case of the 14700K.

My biggest critique overall on 14900KS would be value for dollar looks rather poor overall, but we'll have to see what it actually retails at MSRP once the the initial dusts settles a bit. I don't expect it will be very favorable though given past history. The 14900K should certainly be higher value and similar enough performance while a 14600K or 14700K will be dramatically higher value for dollar it appears and more sensible for most buyers. It was never meant to be sensible though it's a halo product.

I agree, but what I question here isn't the value, it's the viability. It'd be a much better processor IMHO if Intel targeted 6 GHz on all 8 P-cores rather than go with the 6.2 scheme, but it's probably even harder than finding silicon that can do 6.2 on just 2 P-cores, I mean, 13900KS chips can in general overclock to the 14900KS's 5.9 turbo/6.2 targets with relative ease but... it also comes at the same cost of that tremendous power consumption, it's just way, way off curve. Anyhow, I reckon it's a storm in a teacup once again. People who buy these know what they're getting into, and it's not really a product for me, obvious reasons.

Even just targeting 4P cores at 6GHz probably would've been more sensible. At the same time maybe just 2P cores at 6GHz and a higher all E core ratio than the 14900K or like half of the E cores at a higher ratio. Pushing it to 6.2GHz on 2P cores is a bit extreme on requirements to get stable and probably less beneficial overall thus not exactly worth while. The 14700K is pushed far enough really even with it's clock speeds for the amount of additional cores and cache 14900K/KS brings. It's overkill on the power curve for certain and minimal on uplift when majority of uplift is more derived from the additional core count and cache itself in the end. As I said though halo product some people will pay asinine amounts higher for negligible difference in spite of it being entirely irrational. It's a bit like buying all game DLC just to have all of it which some people absolute do in certain instances even for just for a tiny bit of added cosmetic nonsense.

Just chipping in: these are screenshots I produced with my 13900K on 19th October 2022, monitored and unmonitored single thread runs in Cinebench:




And the chip was not hitting 100°C while running 1 thread of heavy load at 6200 MHz. That was almost 1,5 years ago, one could expect that they polished the manufacturing somehow and today they are able to produce even better chips.

InVasManiPremium binning has always costed more and be worse on value for dollar in general. It's really nothing new. It's good enough chip, but the price premium for binning and higher overall power draw aren't viewed very favorably to the majority and hard to fault people on those particular criticisms.

Halo products are not intended to make sense for the majority. I do not get that side of criticism at all - why would you or me care? If there is a 750€ product that has only couple hundred megahertz over a 600€ product then do not buy the more expensive one.

The real criticism about premium binned processors was always from overclockers, that it takes the silicon lottery out of equation. Significantly lowered chance to get a golden sample out of the normal SKU.

Super cars are for attention and to bag a better looking one-night stand. Speed can be had for much cheaper.

BoggledBeagleJust chipping in: these are screenshots I produced with my 13900K on 19th October 2022, monitored and unmonitored single thread runs in Cinebench:




And the chip was not hitting 100°C while running 1 thread of heavy load at 6200 MHz. That was almost 1,5 years ago, one could expect that they polished the manufacturing somehow and today they are able to produce even better chips.

So...here's the issue. The increase in performance of manufacturing does not insure a "better" product. What insures a better product is testing, and identifying the product which has exceeded the average in a positive way. If you want to make a point of it, I'd suggest you research a thing called a control chart. More control means less required testing and sorting, less negative product to be scrapped, and thus a more profitable product.

What you are hoping for is that they narrowed the control limits, then pushed them higher. You make a product that is better, and is cheaper. In reality, you are not looking at directly better manufacturing with a CPU. You are looking at the output of the CPU based on electro-mechanical properties...which could be linked directly to manufacturing controls or not. It's much simpler to simply produce more parts with less scrap, then bin the larger volume to get better performers that you can label as such without ever actually making a "better" chip...as it was within your controls to produce previously.



The TL;DR is better process<>better chips. These new chips are only as good as the profit margins...and halo products demanding a huge binning effort is fine when the 99% of chips not meeting the halo can be marked as a much higher moving part.

lexluthermiesterYeah, that's easily understood. Some people just don't care about that imbalance. They want the best and are willing to pay for it.

It's kinda like the Threadripper system I built for a guy over the holidays. Totally impractical, but in his eyes that was the best and what he wanted. He paid for the parts, I built it for him. He loves it and didn't bat an eye at the $8000+ price tag.
(BTW, THE most powerful system intended for personal use I've ever built to date!)

Different strokes for different folks is all I'm saying. Intel, AMD, Nvidia, etc, etc would be fools not to cater to that sector of people. After all look at Apple. Perfect example of over-priced tat that people see as "premium"...

Honestly for a lot of people, if spending $1000 on a KS or $2000 on a 4090 keeps them from taking on a different hobby then it is money well spent.

~10% of households in the US have income > 220K/year. PCs as a hobby, and going for the 'best' in that space, is actually quite cheap for them. Compare to say motorcycling, boating, RVing, hot rodding, and so on. Most of those have a minimum $20K price to enter, plus ongoing storage / insurance / maintenance costs.

Hopping up a PC is one of the cheapest hobbies you can have.