AMD Zen 2 (Ryzen 3000) - *** NO COMPETITOR HINTING ***

[MyBrainz]

In a bit of unexpected news, AMD has posted a link on its investor website to a webcast for its 50th Anniversary Celebration that will take place tomorrow, May 1, 2019, at 1:00 pm PT.

1pm Pacific Time (PT) = 9pm GMT

 

[Wommers]

And the mindshare is still heavily against AMD, so £500 for a Ryzen - regardless of how much it'll smash Intel - just won't sell.

The 1800X launched at £500. I'd say it's a certainty they will use this for the 16c part.

 

[EsaT]

They have in the last two generations, IIRC Epyc trailed consumer releases by 2-3 months.

All testing and verifying associated with enterprise products certainly needs time.
Though this time it's not completely new architecture and desktop CPU release has possibly been delayed by getting new chipsets ready.
So timing might not be entirely about engineering aspects.

Earning details for the quarter thats just gone:

https://www.theregister.co.uk/2019/04/30/amd_q1_2019_future_guidance/

Considering current products nearing state of becoming old and major new product launches coming that's as good as can be expected.
(for company who needs to actually compete instead of being able to just rape&rob buyers at will)

 

[~>Dg<~]

the markets places are already there for each amd chip. thats why all the old ones are being gotten rid of so the new ones take the pricing sector places.

That was back when Intel's top of the line HEDT CPU was the 10 core 6950X at £1600, the 8 core 6900 was £1000. so for the same performance the 1800X was half the price.

Too many people are forgetting this.

many people also forget that the next month intel launched the 7900x. which was even faster. so its sort of a amd trait. they launch a new product at the same time of the death of the last intel chips. this makes amd chips look better and then the next intel arrives. launch at a time where you may have advantage. when your product appeals most.

 

[TrixX]

many people also forget that the next month intel launched the 7900x. which was even faster. so its sort of a amd trait. they launch a new product at the same time of the death of the last intel chips. this makes amd chips look better and then the next intel arrives. launch at a time where you may have advantage. when your product appeals most.

Except this time we know the next Intel's are gonna be crap by comparison. 10c Comet Lake on 14nm+++++++++++++++++ is gonna have a hard time being relevant. Ice Lake is just a a reflection in the ice at the moment, nothing concrete on arch or capabilities and there's not a lot else on the horizon for them. That's assuming Ice Lake is on track to release to desktop and not just languish in the no-mans land of low power laptop CPU's with it's 10nm 4yr late process.

 

[humbug]

AMD said Rome was Q3.

Yup, she didn't actually say anything about Ryzen 3000.

Su said AMD has achieved key milestones to deliver its next-generation 7nm Epyc "Rome" processor – which claims 4x better floating point performance and 2x better compute performance per socket – to production in Q2 and availability in Q3. The company's 7nm Navi gaming GPUs are on track for Q3 introduction, and will be cheaper than Radeon 7, she said.

This is also exciting for AMD's revenue, they have already secured Googles Game Streaming service, they will be using AMD GPU's. and EPYC CPU's will be used for Amazon's Data centers.

In the company's conference call for investors, Su highlighted AMD's tie-ups with cloud giants Google – for its upcoming Stadia service – and Amazon – for EC2 instances based on AMD Epyc 7000 chips – to underscore the company's positive outlook.

With the possibility of competing Microsoft and Amazon streaming service also using AMD.

Amazon and Microsoft could be among AMD's future buyers of streaming tech, given that both are said to be planning game streaming platforms to compete with Stadia.

If MS and Amazon also get onboard with Game Streaming using AMD's tech their revenue will grow massively.

 

[humbug]

Del

 

[~>Dg<~]

Except this time we know the next Intel's are gonna be crap by comparison. 10c Comet Lake on 14nm+++++++++++++++++ is gonna have a hard time being relevant. Ice Lake is just a a reflection in the ice at the moment, nothing concrete on arch or capabilities and there's not a lot else on the horizon for them. That's assuming Ice Lake is on track to release to desktop and not just languish in the no-mans land of low power laptop CPU's with it's 10nm 4yr late process.

we know just as much about the new intel chips as we do the amd ones. all speculation. amd are playing catch up with these chips to intels old chips. so the new ones are obviously going to be faster than whats out now. so if amds new chips are just on par with old intel tech with more cores then obviously they will be behind again with the new intel ones. thats just logic. also why amd launches there cpus at the end of intels old chips.

 

[TrixX]

we know just as much about the new intel chips as we do the amd ones. all speculation. amd are playing catch up with these chips to intels old chips. so the new ones are obviously going to be faster than whats out now. so if amds new chips are just on par with old intel tech with more cores then obviously they will be behind again with the new intel ones. thats just logic. also why amd launches there cpus at the end of intels old chips.

Please can you supply some evidence to show how a 10c Comet Lake CPU will be able to beat a 9900K for anything. It's gonna need more power. It's gonna be hotter. It's going to put mobo's already stressed out VRM's with the 9900K through the ringer.

So please by all means illuminate me on how Intel are going to pull a rabbit out of the hat for Crapit Lake. You have such an interesting narrative each time you attempt to spin things in favour of Intel it's getting a little tiresome.

I'm all in favour of some good Intel competition but we won't see it with Comet Lake. Likely not to see it with Ice Lake either and likely not until whatever 7nm process they move to. The earliest that's happening is 2021 at best 2022 at worst. We'll be on Zen 4 by then so the comparison to the 3xxx is utterly useless. Just like most of your commentary in this thread.

 

[beany_bot]

Please can you supply some evidence to show how a 10c Comet Lake CPU will be able to beat a 9900K for anything. It's gonna need more power. It's gonna be hotter. It's going to put mobo's already stressed out VRM's with the 9900K through the ringer.

So please by all means illuminate me on how Intel are going to pull a rabbit out of the hat for Crapit Lake. You have such an interesting narrative each time you attempt to spin things in favour of Intel it's getting a little tiresome.

I'm all in favour of some good Intel competition but we won't see it with Comet Lake. Likely not to see it with Ice Lake either and likely not until whatever 7nm process they move to. The earliest that's happening is 2021 at best 2022 at worst. We'll be on Zen 4 by then so the comparison to the 3xxx is utterly useless. Just like most of your commentary in this thread.

He has 30 THOUSAND posts of trolling AMD users lol. It's mad.

 

[DragonQ]

also why amd launches there cpus at the end of intels old chips.

Why do you just make stuff up? Ryzen 1 launched a couple of months after Kaby Lake. Coffee Lake came out in October 2017 (although in very limited quantities and only top SKUs until around April 2018), whereas Ryzen 2 launched in April 2018. Coffee Lake Refresh launched in October 2018. You could say Ryzen 2 was 6 months after Coffee Lake but it was also 6 months before Coffee Lake Refresh, so I don't know how you can say AMD launches their CPUs at the end of Intel's cycle.

 

[Potatowithearsontheside]

I like how he says that the next gen is obviously faster than the last gen. I don't see this with 10th gen on 14nm+++++++++++++++++ Comet Lake, nor with 11th gen on 10nm Ice Lake (if it ever materialises).
Intel openly said that their first 10nm CPUs wouldn't be faster than their current 14nm CPUs, and that was before they threw a 9900K into the mix.

 

[d_brennen]

The "new" Intel chips coming in the next 12-18 months aren't "new" at all. Zen 2 is doing plenty of new things though.

 

[beany_bot]

Why do you just make stuff up? Ryzen 1 launched a couple of months after Kaby Lake. Coffee Lake came out in October 2017 (although in very limited quantities and only top SKUs until around April 2018), whereas Ryzen 2 launched in April 2018. Coffee Lake Refresh launched in October 2018. You could say Ryzen 2 was 6 months after Coffee Lake but it was also 6 months before Coffee Lake Refresh, so I don't know how you can say AMD launches their CPUs at the end of Intel's cycle.

 

[~>Dg<~]

He has 30 THOUSAND posts of trolling AMD users lol. It's mad.

why do people like you say this ? i had amd pcs for best part of twenty years now and still do. i even said im buying a 12 core cpu when these come out. do people just make stuff up as they go along ? out of curriosity how many years have you had amd hardware ? im just not brand biased but if amd is faster at what i need it for i buy it. its not out of blind loyalty . that is the difference to many here. yet if you buying intel because they do what you want faster and better your a amd troll ? such weak minded excuse of a reply .

 

[~>Dg<~]

Please can you supply some evidence to show how a 10c Comet Lake CPU will be able to beat a 9900K for anything. It's gonna need more power. It's gonna be hotter. It's going to put mobo's already stressed out VRM's with the 9900K through the ringer.

So please by all means illuminate me on how Intel are going to pull a rabbit out of the hat for Crapit Lake. You have such an interesting narrative each time you attempt to spin things in favour of Intel it's getting a little tiresome.

I'm all in favour of some good Intel competition but we won't see it with Comet Lake. Likely not to see it with Ice Lake either and likely not until whatever 7nm process they move to. The earliest that's happening is 2021 at best 2022 at worst. We'll be on Zen 4 by then so the comparison to the 3xxx is utterly useless. Just like most of your commentary in this thread.

you are just speculating. just like i am. intel are bringing comet lake out end of this year. so where you getting 21/22 out i dont know. do you really think intel will bring out a new slower chip ? there would be no point in that . it will be faster. take brand loyalty out of arguements and debates. try and look at the logic in the debate. i am not brand loyal. this is what so many cant grasp. not every one sticks to one brand and believes anything thats said. i believe what is and true and what is proved. not hyperbole.

 

[beany_bot]

you are just speculating. just like i am. intel are bringing comet lake out end of this year. so where you getting 21/22 out i dont know. do you really think intel will bring out a new slower chip ? there would be no point in that . it will be faster. take brand loyalty out of arguements and debates. try and look at the logic in the debate. i am not brand loyal. this is what so many cant grasp. not every one sticks to one brand and believes anything thats said. i believe what is and true and what is proved. not hyperbole.

You are. By definition. Literally the most brand loyal person I have seen on these forums.

 

[~>Dg<~]

You are. By definition. Literally the most brand loyal person I have seen on these forums.

your just assuming i am without knowing. that is the issue. dont be one of those people. take brand loyalty out of the equation and look at the logic and history of cpus.

 

[smilingcrow]

Please can you supply some evidence to show how a 10c Comet Lake CPU will be able to beat a 9900K for anything. It's gonna need more power. It's gonna be hotter. It's going to put mobo's already stressed out VRM's with the 9900K through the ringer.

One advantage of more cores is that often you can get the same performance at a lower clock speed which can mean lower voltage also.
So for a given power consumption you might get more performance which is important for architectures like this which are being pushed beyond what is typically considered desirable. i.e. Desperate measures.
Sure, it also depends on scaling and other factors but for heavily threaded workloads I wouldn't be surprised if it will offer more performance.
Definitely not enough to compete with a Zen2 16C but it will push it closer to Zen2 12C probably.
Hopefully that is the case as it might put more downward pressure on Zen2/12C which will please some.
But knowing Intel it will have a stupid price and/or be in low supply so might have little if any impact on AMD.

 

[Steampunk]

I think Intel will have problems not just because of their issues getting to 10/7nm, but because they have to redesign to mitigate against the various security issues we've seen recently (Meltdown, etc), which can have an impact on performance. At a time when AMD are executing well and ahead in process nodes and so on, Intel will be facing tough competition at a time when they are having lots of problems of their own making.

AMD products could be good at a time when Intel products could be poor, and that might be the case for the next few years.