9900k paper launch

sanitarium16

Limp Gawd
Joined
Dec 5, 2007
Messages
328
Shades of 8700k again? Latest rumor has the chip paper launching August 14 with availability in October. Lets hope they make enough chips this time!!!
 
they have to try and rain on AMD's TR2 launch some how but yeah my guess would be to expect another 8700k launch shit show given how blatantly things have "accidentally" been leaked over the last week.
 
How would that even remotely compete with TR2? Completely different crowds... Also, if it's just another 8700k with 2 extra cores I'll start to really worry about Intel. They've been releasing a lot of meh from Z270/x299 on.
 
How would that even remotely compete with TR2? Completely different crowds... Also, if it's just another 8700k with 2 extra cores I'll start to really worry about Intel. They've been releasing a lot of meh from Z270/x299 on.
they had zero competition for 10 years. Every couple years AMD said something else would save them and completely failed. nobody saw this coming. It takes a while to counter and all they have to do is continue dominating mobile and hold on enough. AMD still hasn't caught up on per core performance but its getting really close now.
 
they have to try and rain on AMD's TR2 launch some how but yeah my guess would be to expect another 8700k launch shit show given how blatantly things have "accidentally" been leaked over the last week.
TR2 is going to own the HEDT market but how big is that market really? Its the server market that matters and thats where things will start getting interesting. performance per watt matters to a lot of markets. I think intel is still doing just fine in that side.
 
How would that even remotely compete with TR2? Completely different crowds... Also, if it's just another 8700k with 2 extra cores I'll start to really worry about Intel. They've been releasing a lot of meh from Z270/x299 on.

What else could they release but an 8700K with 2 extra cores? The process they've been depending on to advance their business is over 2 years late at this point.
 
TR2 is going to own the HEDT market but how big is that market really? Its the server market that matters and thats where things will start getting interesting. performance per watt matters to a lot of markets. I think intel is still doing just fine in that side.


This, I think AMD has a decent proposition for eeking out a few percentage points from Intel in the server market but to make proper inroads they need to have the low voltage and Xeon D equivalents.

If they dropped some decent 35-65w processors their share price would go crazy.
 
This, I think AMD has a decent proposition for eeking out a few percentage points from Intel in the server market but to make proper inroads they need to have the low voltage and Xeon D equivalents.

If they dropped some decent 35-65w processors their share price would go crazy.
I'm hoping AMD makes some moves but forums like this are packed with people saying whats intel doing? mostly just doing what they do; print money. Intel will continue to have at least a perceived stability advantage too. The ryzen rollout was very bumpy. This battle is still very early.
 
It's always like that where people don't understand the enterprise context of things. We look at the world through a niche (of enthusiasts) of a niche (largely pc gamers). So Intel not doing much seems odd, but then you have see that the worldwide server market is conservatively 8 times Nvdia's revenue (white box is a bit of a guess). Or that Amazon literally wheel in 50,000 dual CPU servers when they open up a new AZ. Or that each server admin at FB looks after more than 25,000 servers. Or Google has a couple of million boxes.

The enthusiast market is important, and earns revenue that boggles most peoples mind, but it also ain't shit.

AMD have a chance to start making some inroads, but it's a slow road as you say. Look how long it took Opteron to get share and that had a marked advantage, and it was at a time when hardware investment from individual companies was high because of the virtualisation wave. Now we're in a world where CPU capacity is seldom a limit and energy efficiency is the prime motivator of any large purchaser.
 
if it doesnt perform better in games then the 8700 then i hope to get a used seller on here that wants to upgrade
 
TR2 is going to own the HEDT market but how big is that market really? Its the server market that matters and thats where things will start getting interesting. performance per watt matters to a lot of markets. I think intel is still doing just fine in that side.
Theyre getting beaten at power efficiency too... Server and desktop especially hedt.
 
25000 servers per admin is holy wtf i dont even just wow

thats a lot of threads

Whatever you think of their product or management, Facebook's engineering is absolutely world class. They're also pretty good about letting it out the door as well, Cassandra, React etc, then obviously there's the hardware stuff with the Open Compute Platform.

They put a lot of pressure on the scalability of Chef over the years.

Googles admin to server ratio is way higher btw, but these days that's pretty much all just Skynet.
 
The first page of the report being noticeably thicker than the rest.
 
Whatever you think of their product or management, Facebook's engineering is absolutely world class. They're also pretty good about letting it out the door as well, Cassandra, React etc, then obviously there's the hardware stuff with the Open Compute Platform.

They put a lot of pressure on the scalability of Chef over the years.

Googles admin to server ratio is way higher btw, but these days that's pretty much all just Skynet.
even the push they did on PHP/caching quite a while back. The engineering at facebook is probably the best in the world which is why all of the crap going around right now is so shocking. Nothing happens without them letting it happen. the compute power that these big corporations have is just insane. The 2670 8 core chip that hit like $50 a few years back was due to I think it was facebook upgrading and dumping/flooding the market. I'm watching for the next wave like this but I think they've been tapering the upgrades better.
 
What is the real possibility that this chip will be ready for purchase in August or September? I'm getting ready to do a major upgrade after *cough* seven years.
 
It's always like that where people don't understand the enterprise context of things. We look at the world through a niche (of enthusiasts) of a niche (largely pc gamers). So Intel not doing much seems odd, but then you have see that the worldwide server market is conservatively 8 times Nvdia's revenue (white box is a bit of a guess). Or that Amazon literally wheel in 50,000 dual CPU servers when they open up a new AZ. Or that each server admin at FB looks after more than 25,000 servers. Or Google has a couple of million boxes.

The enthusiast market is important, and earns revenue that boggles most peoples mind, but it also ain't shit.

AMD have a chance to start making some inroads, but it's a slow road as you say. Look how long it took Opteron to get share and that had a marked advantage, and it was at a time when hardware investment from individual companies was high because of the virtualisation wave. Now we're in a world where CPU capacity is seldom a limit and energy efficiency is the prime motivator of any large purchaser.

The last 10 years have been VERY good for datacenters. But it is always the same story. A very small niche is far more vocal than their market-share entitles them to be. And they for some reason always seem to be clueless about validation.
 
TR2 is going to own the HEDT market but how big is that market really? Its the server market that matters and thats where things will start getting interesting. performance per watt matters to a lot of markets. I think intel is still doing just fine in that side.

have you seen Dr8auers overclocked dual Epyc 7000 series? It blew away a quad platinum Xeon machine that cost $60,000 and the Dual AMD was only around $14000.

Intel is not power efficient when it takes 4 of theirs to eq... actually end up less than 2 of AMDs.

Caveat emptor ... Of course that was overclocked.
 
have you seen Dr8auers overclocked dual Epyc 7000 series? It blew away a quad platinum Xeon machine that cost $60,000 and the Dual AMD was only around $14000.

Intel is not power efficient when it takes 4 of theirs to eq... actually end up less than 2 of AMDs.

Caveat emptor ... Of course that was overclocked.

Judging by how you only look at single metric...I am sure you don't work in data centers/enterprise. It is a lot more complex than the consumer space. It will be at least 3-4 years before Epyc 7000 is validated enough to even consider using it.
AMD's track-record in enterprise is less than stellar...I don't think I have a single co-worker in my department that even remotely considers AMD a viable option at this point in time FYI.

Validation, stability, support, TCO, training, operations ect. the list is long..and will not be influence by simple benchmarks.
Again, consumer-mentality does not work in data centers/enterprise...kinda like science, a single study does nothing to consensus.
 
have you seen Dr8auers overclocked dual Epyc 7000 series? It blew away a quad platinum Xeon machine that cost $60,000 and the Dual AMD was only around $14000.

Intel is not power efficient when it takes 4 of theirs to eq... actually end up less than 2 of AMDs.

Caveat emptor ... Of course that was overclocked.
Yes, if one also overclocked dual-socket Xeon 8180s then it would be close. At 3.2 GHz, they hit 7400.



Though nobody who has the need and budget for these type of machines will overclock them.
 
Judging by how you only look at single metric...I am sure you don't work in data centers/enterprise. It is a lot more complex than the consumer space. It will be at least 3-4 years before Epyc 7000 is validated enough to even consider using it.
AMD's track-record in enterprise is less than stellar...I don't think I have a single co-worker in my department that even remotely considers AMD a viable option at this point in time FYI.

Validation, stability, support, TCO, training, operations ect. the list is long..and will not be influence by simple benchmarks.
Again, consumer-mentality does not work in data centers/enterprise...kinda like science, a single study does nothing to consensus.

Only years 8n the data center my friends

Sigh. I'm so tired of this site trying to be enterprise all the damned time. When I log on here I expect this to be a consumer enthusiast site, not a co location facility review of standards and practices and validation and government security scrutiny and blah blah.

I have a dense amount of professional time in data centers and I quit the career because it's not fun at all in the least to me. I did Cisco Route Switch work day in and day out for a very long time but hit burn out. I come to hard because Cinebench matters not if your server meets company Xs or Ys application security standards. That shit is so not H. Maybe in it's own subforum but Der8auer is an enthusiast not a data center equipment validator.
 
Last edited:
Paper Launch! :D :) Let me know if you can actually buy this processor on August 14th because otherwise, it is a classic paper launch by any definition. Intel is scrambling to compete and do not have a clue what they are doing at the moment. Guess they cannot just pay off OEM's anymore and actually have to learn how to compete for real again. :)
 
Only years 8n the data center my friends

Sigh. I'm so tired of this site trying to be enterprise all the damned time. When I log on here I expect this to be a consumer enthusiast site, not a co location facility review of standards and practices and validation and government security scrutiny and blah blah.

I have a dense amount of professional time in data centers and I quit the career because it's not fun at all in the least to me. I did Cisco Route Switch work day in and day out for a very long time but hit burn out. I come to hard because Cinebench matters not if your server meets company Xs or Ys application security standards. That shit is so not H. Maybe in it's own subforum but Der8auer is an enthusiast not a data center equipment validator.

Well, it is not the site but those who are deflecting and commenting. I will not personally own anything but AMD in my desktop but, a paper launch is exactly what this appears it will be. After all, I wonder what sort of Cinebench score and piece of paper will get?
 
Didn't AMD say EPYC 2 based on zen 2 on 7nm is already sampling? They are firing on all cylinders and I won't be surprised if they light fire under intels ass in server arena with eypc 2. With intel 10nm not being ready until end of 2019, but I suspect it will be 2020 they are just doing their best to reassuring shareholders. AMD will have a process lead by a year or year and half. I do think AMD will be able to penetrate the big market.
 
I'm not touching another CPU until the known security flaws are declared + proven fixed and performance exceeds current gen without patches.
They dont deserve money when there is no compensation for cutting corners on the CPUs I bought. When to still not be fully safe I need to apply an update that harms performance. They arent even genuine about how much.
And as soon as the fixed processors are released the demand and value of anything preceding will drop.
Its still a bad state of affairs.
 
Back
Top