You're completely missing the point. Normal for Intel is keeping prices high even when they don't have the performance compared to the competition to keep prices high. It's extremely rare for Intel to drop prices much less slash them like this. The point I'm making is they can do this because Intel will usually have something coming out which will "justify" the higher prices because the new product will again be worth the high price. The fact that Intel is slashing prices is a strong indicator that Intel will have nothing out soon to compete and justify the higher prices.
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Cascade Lake-X Pricing and Model List
- Thread starter TurboGLH
- Start date
You're completely missing the point. Normal for Intel is keeping prices high even when they don't have the performance compared to the competition to keep prices high. It's extremely rare for Intel to drop prices much less slash them like this. The point I'm making is they can do this because Intel will usually have something coming out which will "justify" the higher prices because the new product will again be worth the high price. The fact that Intel is slashing prices is a strong indicator that Intel will have nothing out soon to compete and justify the higher prices.
cybereality
[H]F Junkie
- Joined
- Mar 22, 2008
- Messages
- 8,789
SmokeRngs Yeah, I get it. What I'm saying is that when a competitor come with a game-changing product (beating on performance and price), all thoughts of "business as usual" are thrown out the door.
Seriously though, Intel has been killing it in the HEDT line if they can do this just from some competition.
GoodBoy
2[H]4U
- Joined
- Nov 29, 2004
- Messages
- 2,768
The 10 and 12 core chips look like the sweet spot. The lower base freq but higher boost is kinda odd tho.
Price is great. Thanks AMD!
Might be some cool shit out by the time I upgrade again (probably late 2021 or 2022)
Price is great. Thanks AMD!
Might be some cool shit out by the time I upgrade again (probably late 2021 or 2022)
RanceJustice
Supreme [H]ardness
- Joined
- Jun 9, 2003
- Messages
- 6,629
Its nice to see some competition, but I'd still rather support AMD with Threadripper. All I can hope is that Threadripper Zen2 (or whatever they call it) will have some equivalent to these Intel "TB 3.0" setups where you can clock a few cores quite highly. Even the highest end of Intel's really can only hold an all-core turbo comparable to what we've seen with Ryzen and older TR (ie 4.2 - 4.3ish) but they can clock 2 cores up to 4.8 or whatnot and 2 more to 4.7 , which would be nice for Threadripper 3000 to do as well. After all, they have to know that there are situations where not every core is going to be used to the maximum (ie gaming) so I hope AMD has a plan.
And delivered in April?
Agreed the only thing that can really stop AMDs from dominating this round will be availability. I think with the 3900x although I'm sure they would love the fabs to turn out more... the real issue was just the insane level of demand. I have seen 3900s pop up at my local part shop basically every couple days since launch... its just they come in and there gone by the next morning. TR demand should be a little slower being a smaller market... and I believe a lot of first Gen TR customers have probably already grabbed 3900s or are going to opt for the 3950s.
I agree Intel isn't going to have a real answer till 7nm. I am honestly looking forward to it, even though I am an admitted AMD booster. If Intel does end up with a 7nm 3D stacked chiplet design, 2021 could be a very interesting year... perhaps the most interesting and fun year in CPUs since the 90s.
End of 2019 should be fun... game is AMDs to win or loose right now it seems.
trixdout
n00b
- Joined
- Feb 1, 2012
- Messages
- 26
So do I go return my i9-9900k / z390 master, and go with one of the i9-10xxx chips ???? Was building my new rig around the 9900K, now I'm second guessing.
Really surprising that prices are coming down so low. Something is up. Either Intel has nothing groundbreaking in the pipes or they need time to work out kinks in something new.
Really surprising that prices are coming down so low. Something is up. Either Intel has nothing groundbreaking in the pipes or they need time to work out kinks in something new.
What's up is that AMD has highly competitive chips. Intel doesn't charge what they do for their CPUs because that's what they cost and they have to, they charge it because they can because AMD didn't compete for so long.
Something similar happened back in the original Athlon days. The Pentium 3 was out and was the big dog, running at up to 500Mhz. Then the Athlon came out offering basically the same performance per clock, speeds up to 1GHz, and a lower price. Suddenly P3s became available up to 933MHz and their prices jumped down.
Same reason why cable companies drop prices and raise speeds in areas where the phone company or someone else puts fibre to the home: Competition works.
Do you need the extra cores, and/or are you willing to pay the extra price?
Do you actually NEED extra cores, or just want them?
If you are asking, then it seems most likely it's just a "want", and the answer is No. Stick with the 9900K which will be faster in general usage and gaming.
Want and need are synonyms in [H]land. So pick your flavor of [H]: raw speed or enough cores for the neighborhood.
Armenius
Extremely [H]
- Joined
- Jan 28, 2014
- Messages
- 42,114
If the primary usage for your PC is gaming then getting a HEDT processor would be a downgrade. Some scenarios for when you would want a HEDT processor are for things like rendering video and 3D animation, running a server, crunching a lot of data, maybe streaming and gaming on the same machine.
DejaWiz
Fully [H]
- Joined
- Apr 15, 2005
- Messages
- 21,825
I think Ryzen 3000 is still a better value for enthusiast/gaming use, when comparing closest cores/threads/clocks:
Ryzen 9 3900X
7nm
12C/24T
3.8/4.6 GHz
DDR4-3200
105W TDP
$580
i9-10920X
14nm
12C/24T
3.5/4.8 GHz
DDR4-2933
165W TDP
$689
************************************************
Ryzen 9 3950X
7nm
16C/32T
3.5/4.7 GHZ
DDR4-3200
105W TDP
$749
i9-10940X
14nm
14C/28T
3.3/4.8 GHz
DDR4-2933
165W TDP
$784
i9-10980XE
14nm
18C/36T
3.0/4.8 GHz
DDR4-2933
165W TDP
$979
Ryzen 9 3900X
7nm
12C/24T
3.8/4.6 GHz
DDR4-3200
105W TDP
$580
i9-10920X
14nm
12C/24T
3.5/4.8 GHz
DDR4-2933
165W TDP
$689
************************************************
Ryzen 9 3950X
7nm
16C/32T
3.5/4.7 GHZ
DDR4-3200
105W TDP
$749
i9-10940X
14nm
14C/28T
3.3/4.8 GHz
DDR4-2933
165W TDP
$784
i9-10980XE
14nm
18C/36T
3.0/4.8 GHz
DDR4-2933
165W TDP
$979
trixdout
n00b
- Joined
- Feb 1, 2012
- Messages
- 26
I don't think I need the 8 cores. I was going to go with i7-9800X originally, but microcenter had a bundle deal for the 9900K w/ Aorus Master. I just want to build a rig and slowly increase the performance of the CPU until its maxed out as years go by. So the extra cores would help for longevity in my case.
I looked for overclocking results on the i9-9900x. I didn't find much but it looks like 5GHz all core is obtainable. Obviously there's no way to tell yet if the 10900x will be able to do it. I'd guess that it's not worth going from the 9900K to the 10900X except in very specific scenarios (I saw a post somewhere (Phoronix?) by someone who claimed they needed 10 actual cores for an unstated reason).
GoodBoy
2[H]4U
- Joined
- Nov 29, 2004
- Messages
- 2,768
What did you pay for the 9900K vs what the new cpu is expected to be at?
Chances are, you should keep what you have, unless a few hundred more isn't really a big deal. The price drop/deal you got is likely to just move them out to make way for the new products. You got in at the perfect time.
We don't know for sure how well these new ones will OC, but it might not be as good an OC'er as the 9900 you have, which will probably hit 5.0 easily, and do 5.1 or 5.2 on water. We can't really know for sure on the new stuff till its out. But a best case scenario would be that the new stuff OC's just as good. So then you are back to comparing the core count/price equation. If you can afford the extra cost now, take a bit of a gamble it will OC just as well, then go ahead and return it and wait for the new stuff.
And really, either way you are going to be getting a good system.
Chances are, you should keep what you have, unless a few hundred more isn't really a big deal. The price drop/deal you got is likely to just move them out to make way for the new products. You got in at the perfect time.
We don't know for sure how well these new ones will OC, but it might not be as good an OC'er as the 9900 you have, which will probably hit 5.0 easily, and do 5.1 or 5.2 on water. We can't really know for sure on the new stuff till its out. But a best case scenario would be that the new stuff OC's just as good. So then you are back to comparing the core count/price equation. If you can afford the extra cost now, take a bit of a gamble it will OC just as well, then go ahead and return it and wait for the new stuff.
And really, either way you are going to be getting a good system.
As a reminder, the 9900K is an 8C16T part. The 9900X has 10 cores and costs $250 more at MIcrocenter and the motherboards are usually a bit more expensive, too. (Thanks, Intel, for having two parts with a different core count but the same name except for one letter!)
Master_shake_
Fully [H]
- Joined
- Apr 9, 2012
- Messages
- 17,794
damn those TDP numbers on panic lake are crazy.
I just looked up the stats on Silicon Lottery (https://siliconlottery.com/pages/statistics) if anyone cares.
They show that 30% of 9900Ks can hit 5GHz (91% hit 4.9).
Unfortunately they don't have numbers for the 9900X--only the 9800X and 9920X. Both the 9800X and 9900X are 100core parts, but IIRC are different dies, one being LCC, the other HCC. As such the 9920X numbers are probably more relevant to how a 9900X would do. But they show that only 3% of 9800X chips can hit 5GHz (62% can get 4.7) and the 9920 is noticeably lower: 18% can hit 4.6, 57% reaching 4.5. I wouldn't expect much better from the 9900X.
I would think that the 9900K would be a nominally better choice for most people (depending on how you feel about a ~400MHz all-core difference), assuming you *need* 8+ cores.
They show that 30% of 9900Ks can hit 5GHz (91% hit 4.9).
Unfortunately they don't have numbers for the 9900X--only the 9800X and 9920X. Both the 9800X and 9900X are 100core parts, but IIRC are different dies, one being LCC, the other HCC. As such the 9920X numbers are probably more relevant to how a 9900X would do. But they show that only 3% of 9800X chips can hit 5GHz (62% can get 4.7) and the 9920 is noticeably lower: 18% can hit 4.6, 57% reaching 4.5. I wouldn't expect much better from the 9900X.
I would think that the 9900K would be a nominally better choice for most people (depending on how you feel about a ~400MHz all-core difference), assuming you *need* 8+ cores.
trixdout
n00b
- Joined
- Feb 1, 2012
- Messages
- 26
I paid $449.99 for the 9900K vs the expected price of $589.99 for the 10900X.
So the $150 is not a big deal (itll just delay the build that much longer).LOL! And don't forget those numbers are at *base* clocks, not the all-core turbo speeds. According to Anandtech, the 9980XE pulls 191W at "full load". The 7920x, the closest that page had to the 9900X, was 141W on a 140WTDP chip--with the TDP on the 10900X being 165, well, it sounds to a layman like they're doing the same thing as AMD: pushing the chip past the point of maximum efficiency on the power curve to get the faster speeds.
I paid $449.99 for the 9900K vs the expected price of $589.99 for the 10900X.
It's not a huge jump, but the more I think about it the more I think that--to me--it's not going to be worth it, as I don't see how I need the extra two cores enough to justify the price. I'd like the higher clock speeds vs my current 3600X, but I'd probably stick with a (presumed) 10700K or 10900K rather than the HEDT chips. The only question is how--outside of gaming--the (again, presumed) 8C8T would do compared to the 6C12T I have now. (I have one MMO I play that--in world boss fights--is severely frame-limited based on single-core CPU speeds, and I wonder if I'd get enough extra FPS from an OC'd 9700/9900/10700/10900K to be worth replacing my current CPU + motherboard.)
trixdout
n00b
- Joined
- Feb 1, 2012
- Messages
- 26
So I made the better choice unknowingly.
I would take SL stats with a grain of salt. Their business model is to sell bin chips. It seem it is a lot more common to hit 5Ghz based of what I seen on forums.
Comixbooks
Fully [H]
- Joined
- Jun 7, 2008
- Messages
- 22,019
Fire sale on those chips too hot for me since I only air cool.
Dayaks
[H]F Junkie
- Joined
- Feb 22, 2012
- Messages
- 9,774
It’s all in where they want to set it. If you try to run the 3900x anywhere near it’s boost, all core, you’ll get in the couple hundred watts area.
It gets crazy, fast, with all those cores.
Huh, I must have missed that at the time. That makes the 9820 kind of stupid.
Edit: Or the other 10-core part, your choice.
Last edited:
Fair enough. I was just looking for some quick numbers.
I'd love to see a more accurate estimate of how good those chips are likely to OC, but I don't know where else you'd get them.
GoodBoy
2[H]4U
- Joined
- Nov 29, 2004
- Messages
- 2,768
The new chips are on a newer process right? They might oc pretty well, as their prior stuff has all oc'd pretty well.
**** Fab 'Process' side discussion ****
The 7nm/10nm/12nm++ etc is all a bit misleading. If you take the transistor count and divide it by the mm², you get transistors per square mm, or a standardized picture of the transistor density for each of the processes. Comparing those numbers across several of these 'processes' myself, I noticed the nvidia GPU's built on 16nm and 12nm were basically the same density.
And to top it all off, these numbers given as the "process size" are not at all in relation to "transistor size", which you would expect since transistor density is what the process size is purported to represent (maybe it isn't but that's what everyone has assumed). The ones I divided out it was a 14.5 to 15, to 1 ratio of process size to nm length on one edge of a transistor. (If my math was right).
One of the last densities I had calculated was 25.5 million transistors per mm², square root is 5050, so a square with:
5050 x 5050 transistors
1mm = 1,000,000 nm
so 1,000,000/5050 = 198nm
And this was one of the more recent, densest chips... either the 7nm AMD, maybe a 12nm GPU or CPU. The transistors are 198nm x 198nm... A rough generalization, maybe they are rectangles, but the point still stands. 198x198 = 39204 nm2 area for a single transistor. If they are rectangle its 100nm x 392nm.. so whatever the "process" is claimed to be, it's not at all the "transistor size".
I believe it is supposed to represent the smallest detail able to be created that make up the transistor as a whole.
Just pointing out that the 7nm vs 12nm isn't really telling us anything useful, in regards to density anyway.
The way the new chips are running (how hard navi is to oc, clocks speeds attained on Ryzen, both of which are 7nm parts) I am inclined to think the Intel process is the better process, as those chips reach higher speeds, even with it being 14nm+++ or whatever it is called.
**** Fab 'Process' side discussion ****
The 7nm/10nm/12nm++ etc is all a bit misleading. If you take the transistor count and divide it by the mm², you get transistors per square mm, or a standardized picture of the transistor density for each of the processes. Comparing those numbers across several of these 'processes' myself, I noticed the nvidia GPU's built on 16nm and 12nm were basically the same density.
And to top it all off, these numbers given as the "process size" are not at all in relation to "transistor size", which you would expect since transistor density is what the process size is purported to represent (maybe it isn't but that's what everyone has assumed). The ones I divided out it was a 14.5 to 15, to 1 ratio of process size to nm length on one edge of a transistor. (If my math was right).
One of the last densities I had calculated was 25.5 million transistors per mm², square root is 5050, so a square with:
5050 x 5050 transistors
1mm = 1,000,000 nm
so 1,000,000/5050 = 198nm
And this was one of the more recent, densest chips... either the 7nm AMD, maybe a 12nm GPU or CPU. The transistors are 198nm x 198nm... A rough generalization, maybe they are rectangles, but the point still stands. 198x198 = 39204 nm2 area for a single transistor. If they are rectangle its 100nm x 392nm.. so whatever the "process" is claimed to be, it's not at all the "transistor size".
I believe it is supposed to represent the smallest detail able to be created that make up the transistor as a whole.
Just pointing out that the 7nm vs 12nm isn't really telling us anything useful, in regards to density anyway.
The way the new chips are running (how hard navi is to oc, clocks speeds attained on Ryzen, both of which are 7nm parts) I am inclined to think the Intel process is the better process, as those chips reach higher speeds, even with it being 14nm+++ or whatever it is called.
What is stupid about the 10 and 12 core parts?
Nothing. I said that having 2 10-core HCC parts with different specs seems silly. When one was LCC and the other HCC, since it was a different layout, it kinda made sense that L3 cache, speed, etc., might be different. When they're both built on the same physical die, there's no sane reason to have one with one speed/cache amount/etc and the other with *different* settings.
Sadly process nodes have become marketing terms which means that companies have been increasingly bullshitting about them. A new number doesn't necessarily mean a process shrink as you might expect, just a fab marketing a new process and different fabs use different terms. It is annoying. Basically just don't worry about what the process is called and look at how the chip performs.
HAL_404
[H]ard|Gawd
- Joined
- Dec 16, 2018
- Messages
- 1,240
was there a price drop? I couldn't tell ... there were TOO MANY ADS on the page to be able to read the content
GoodBoy
2[H]4U
- Joined
- Nov 29, 2004
- Messages
- 2,768
PROTIP: Don't ever bother reading anything on Forbes.
Ha! But the actual answer is these are about half the price of the 9000 series equivalent chips (10980 $979, 9980 $1979). 10 cores for $590 is pretty good, especially if it can get close to a 5GHz all-core.
Well for gaming, AMD's mainstream socket is likely going to be a superior option. I think Ryzen this time competes well even with Intel's HEDT. However if you need over 128GB RAM and all the PCI-e lanes then you'll have to go TR.
TR3 should fair better with gaming this time around as it does not use NUMA, which the Windows scheduler has struggled with. So a game mode will probably not be necessary like past TR chips. I would expect similar boost frequencies as in Ryzen, probably will need some serious cooling to maintain higher clocks given how many cores these are expected to have though.
Yeah, we get that the nn size can be a gimmick, but few buy on that measure alone. AMD 7nm still does not clock as high as 14nm++, but often smaller nodes do not clock as high anyhow given the same architecture. Still, AMD's offering will be much more efficient given the same clocks.
Here is an article on a new 12nm node that looks to be true to its namesake:
https://www.google.com/amp/s/amp.to...ndries-12lp-plus-node-process-12nm,40478.html