• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

AMD is Saving the Ryzen 7 2800X for a Rainy Day

How would this require new silicon? You realize they designed the chips for sever market Thread ripper (16 cores) 1st, and adopted it for the desktop (8), they can stack more cores at any time with very little effort long as the power output on the boards will handle it. So it could be more cores and or higher clocks. Nobody outside AMD knows for sure... I guess it depends on intels response.

Mind explaining to everyone where AMD will get the die space to Siamese another die in?
 
Because what I've said some post ago - they can't add CORE, but CCX with 4 cores, and if they add 1, they might as well add 2 (and that's a Threadripper).

Too much redesign for what is expected to be a reasonably small speed improved.
 
Because what I've said some post ago - they can't add CORE, but CCX with 4 cores, and if they add 1, they might as well add 2 (and that's a Threadripper).

Too much redesign for what is expected to be a reasonably small speed improved.

It wouldn't fit with the current sized PCB.
 
naah,belive, amd 2800z is only they dreams...well they might have it paper,but nothing else.


all know that 2700x 8-core ryzen just and just can fight against intel 8700k 6-core cpu and gaming 8700K win. its fact.

cinbench ,handbrake or thouse useless test 8-core win but not overdrive way.so if amd have 10-core ryzen they release it now,they need it.
amd know 100% that 9700K is coming.

is it weird that amd compete 8-core ryzen for 6-core intel 8700k,sure its win just like cinbech are that kind as i say useless theoretic syntethics program,but
when we start test real world as gaming and 3damrk test, 8700K win!

exmaple 3dmark test, default 8700k win ryzen 2700x even 2700x is oc'd 4300mhz! its tell all how week ryzen is,thats why it is lausy with memory score,bad inside tech.

is it any1 see test where 8700k is oc'd 4600mhz and then compete against ryzen 2700x?? 4300 mhz? no!? , dunno why,bcoz 8700K win all test!

so,take 8700K oc'd it 4600-4700 (can use 24/7), take DDR4 4000mhz 2x8 mem parts and then asus x hero mobo... 2700x loose!!

8700k is better cpu than ryzen 2700x,if you play games and running 3dmarks,its fact!

if you running cinbench,handbrake or so, buy ryzen 2700x, BUT ONLY bcoz is 8-core cpu.

when intels ALSO 8-core cpu release soon ryzen crushed, belive it. even amd release ryzen 2800x! just wait and see.

price is not issue.8700k cost almost same as ryzen. ryzen 2700x is not good cpu,too much trouble eith mobo and memory issues.

erhh, and again, 2800x ryzen not release never. wanna bet.
 
Aargh! So the weak and incompetent management at AMD is still alive and well, and finding new ways to mess up I see! When AMD have the goods to properly stomp on the competition, they pull back and play the nice guys. This is really not good business practice for a company that was on the brink of oblivion just a couple of years ago...

AMD, will your next move be to delay the launch of the next new Zen architecture, just because Intel's CPUs are running the same performance as your existing products? Here is a clue... If you give Intel another year of development, they might just overtake your "new" architecture before you even get a chance to release it, and you'll be left with millions of second rate CPUs that you're going to have to sell at a loss just to get rid of them.

Intel slowly killing you for the last 10 years, and the almost fatal pounding your GPU division has taken by nVidia should have been a lesson for you in how not to run your company. :banghead:
 
Last edited:
Considering that 2700X can't go much higher than 4.2Ghz with all cores, probably 2800X will be a limited version, at a price that wouldn't be making much sense, compared to the 2700X, but at the same time will be helping AMD to have a model in the top spots in every review published by tech sites.
 
AMD is likely setting aside their top rated silicon for a Zen+ Threadripper release later this year (2920X and 2950X).... anything left over might turn into a 2800X

my $0.02 FWIW
 
AMD is likely setting aside their top rated silicon for a Zen+ Threadripper release later this year (2920X and 2950X).... anything left over might turn into a 2800X

my $0.02 FWIW

Are there any upcoming anniversaries for amd? Can't really see any other reason why would they release a bit more clocked version of R7 2700x. The chip is quite maxed out already, which is quite obvious looking at minuscule OC headroom.
 
I think that they held back the 2800X to use as a response to the new Intel Coffee Lake S, 8-Core Processor that will be here shortly. (2nd half of 2018)
Also, Threadripper gets a refresh in Q-3 2018. (this is what interests me the most)

Things are heating up and we are the winners of this new CPU war.
 
Last edited:
I would bet money that's it. None of the reviewers are gettibg 4.5+ out of the chips what makes you think amd has a big of 4.4/4.5ghz chips.
wait what? did i actually say that, 4.4+? Noooooo.
 
2 years back if i had told anyone that we would see such a day where AMD is waiting for Intel's response, i would have been mocked and laughed at. GG AMD, WP, its still not EZ.
 
naah,belive, amd 2800z is only they dreams...well they might have it paper,but nothing else.


all know that 2700x 8-core ryzen just and just can fight against intel 8700k 6-core cpu and gaming 8700K win. its fact.

cinbench ,handbrake or thouse useless test 8-core win but not overdrive way.so if amd have 10-core ryzen they release it now,they need it.
amd know 100% that 9700K is coming.

is it weird that amd compete 8-core ryzen for 6-core intel 8700k,sure its win just like cinbech are that kind as i say useless theoretic syntethics program,but
when we start test real world as gaming and 3damrk test, 8700K win!

exmaple 3dmark test, default 8700k win ryzen 2700x even 2700x is oc'd 4300mhz! its tell all how week ryzen is,thats why it is lausy with memory score,bad inside tech.

is it any1 see test where 8700k is oc'd 4600mhz and then compete against ryzen 2700x?? 4300 mhz? no!? , dunno why,bcoz 8700K win all test!

so,take 8700K oc'd it 4600-4700 (can use 24/7), take DDR4 4000mhz 2x8 mem parts and then asus x hero mobo... 2700x loose!!

8700k is better cpu than ryzen 2700x,if you play games and running 3dmarks,its fact!

if you running cinbench,handbrake or so, buy ryzen 2700x, BUT ONLY bcoz is 8-core cpu.

when intels ALSO 8-core cpu release soon ryzen crushed, belive it. even amd release ryzen 2800x! just wait and see.

price is not issue.8700k cost almost same as ryzen. ryzen 2700x is not good cpu,too much trouble eith mobo and memory issues.

erhh, and again, 2800x ryzen not release never. wanna bet.
You mention AMD only wins in synthetics and not real world but 3DMark is a synthetic and not a real world test

Push come to shove AMD is right at intels ass in gaming

For people who do encoding, rendering, or use any app that can use multiple threads AMDs takes the cake. Price/Performance is in AMDs ballpark

And just an FYI intel already has a 8 Core CPU in the form of HEDT and it doesn’t crush AMDS offerings
 
wait what? did i actually say that, 4.4+? Noooooo.

So what do you think clockspeed will be when the current model does 4.3ghz on XFR? I don't believe even AMD has a dumb enough marketing team to try and sell someone a 2800X that is identical in every way to the 2700X. The only step you could take past 4.3 would be to go to 4.4+ that is kind of how it works.
 
Not having upgraded since 3770K, I guess 2800X would be a viable purchase for me. Few more months worth of waiting would be worth some performance boost. And with whatever upgrade I might do being expected to last me at least three more years, any price difference between 2700X and theoretical 2800X would be negligible.
There aren't that many people like myself though, I guess.
 
Oh right, I wonder why they are leaving it so long? Low 12nm yields?

My worry is this will mean a very staggered Zen 2 release and before you know it we will have the same situation as we do with Intel where the high end parts are a generation or two behind the entry / mainstream.
The yields are probably similar to 14nm but binning is still a factor. They need to stockpile the best chips. It's exactly the same they did the first time around. If they keep up the cadence it won't be too bad. It'll be less than 6 months between mainstream and HEDT. That's pretty good all things considered. Intel relies on a completely different chip to fill their HEDT platform so naturally it takes longer but I feel like Intel could speed time to market or perhaps more accurately realign their product launches because their HEDT schedule has slipped considerably.
 
2950X would be Threadripper+ to replace the 1950X

And I have to admit that I'm most interested in the new Threadrippers.
I've heard that they'll be along in Q3 of this year.
 
And I have to admit that I'm most interested in the new Threadrippers.
I've heard that they'll be along in Q3 of this year.

Good to know, That's when I can anticipate the TR2 3950 in 2019.
 
this whole core war seems stupid to me, gamers lose because as we have seen higher core count chips dont clock as high which means less fps, d3d12 was supposed to fix that but it doesnt seem to have made a difference, both companies are better off sticking to 8 core on the consumer platform and just building stronger architectures instead of adding more cores
 
How would this require new silicon? You realize they designed the chips for sever market Thread ripper (16 cores) 1st, and adopted it for the desktop (8), they can stack more cores at any time with very little effort long as the power output on the boards will handle it. So it could be more cores and or higher clocks. Nobody outside AMD knows for sure... I guess it depends on intels response.

I have confidence that in this instance, w1zzard knows a lot more than you think he does.
 
I have confidence that in this instance, w1zzard knows a lot more than you think he does.

I don't think people in here have quite figured math out. AMD isn't using common core to get the core counts. 4+4 does not equal 10 and there is not room on the die to fit three modules to then disable multiple cores to make some off the wall 10 core BS. I mean they cannot even fit 8 cores with the polaris (sorry "vega") IGP in this package.
 
Back
Top