AMD’s Lisa Su: ‘We will have a high-end Navi’

Status
Not open for further replies.

erek

[H]F Junkie
Joined
Dec 19, 2005
Messages
10,875
An Ultra exclusive venue only for the Elite were invited!

"We have spoken to board partners about Big Navi and they are very enthusiastic about 2020 in general. Based on what they heard from AMD, next-gen RDNA (which could either be RDNA+ or RDNA2) is more power-efficient than RDNA1. They do not have any specific information to share, such as die sizes or compute unit counts, but they are sure it will be on par if not better than the current high-end RTX series.

Their biggest concern is the arrival of the 7nm RTX 30 series, as they simply do not know what to expect. The sooner Big Navi drops, the better.

No specific date was given, but they expect Big Navi to be shown either in June (at Computex) or July (a year after RX 5700). One would guess we are expecting an RX 6000 series graphics card at that point."


https://videocardz.com/newz/amds-lisa-su-we-will-have-a-high-end-navi
 
Let me just the link the article that your article links to give a bit more validation: https://www.anandtech.com/show/1534...-lisa-su?utm_source=twitter&utm_medium=social

I will say that it is slightly disappointing to have to wait longer for Big Navi than we already have. However I guess it does extend the life of my Radeon VII and also validated Apple's choice on going with (modified) Vega 20 is the Mac Pro as the launch for AMD is just too far out.
Here's hoping that as was noted, that by the time it launches it will still be relevant. That's more for AMD's sake than for anyone else's. AMD still has a bit of catch-up to do in terms of GPU cycling.
 
  • Like
Reactions: erek
like this
I dont think anyone doubts that it's coming, but I think people would like to hear a target date and what the performance target will be.
 
This is getting tiresome. While competing on the CPU level is important to computing in general, from a gaming standpoint we've needed high-end GPU competition for more than half a decade and we haven't had it, either compete at the Ti level or don't already.
 
This is getting tiresome. While competing on the CPU level is important to computing in general, from a gaming standpoint we've needed high-end GPU competition for more than half a decade and we haven't had it, either compete at the Ti level or don't already.

Agreed, its always "We have this in the pipe" with AMD, and always there is something better when it finally drops. Shit or get off the pot AMD.
 
This is getting tiresome. While competing on the CPU level is important to computing in general, from a gaming standpoint we've needed high-end GPU competition for more than half a decade and we haven't had it, either compete at the Ti level or don't already.

Well the answer for you then is "don't". I think if the announcements and stuff gets to you, just ignore the news and when it comes time to buy a GPU, just look at what's out and buy the fastest card you can afford.
AMD hasn't been competing at the top for a while, but they definitely have been competing at the low, medium, and high arenas. Just not "ultra". $1000 GPU's are less than 1% of the consumer market (2080Ti's and 2080 Supers COMBINED are less than 1%). Can't say I blame them for not pushing into an incredibly expensive but not particularly lucrative market.
Still, to be clear, I don't disagree with you: I would like more competition to drive quality up and prices down.
 
I'm sure they will have a big Navi by end of summer most likely but the problem is Ampere will likely release a month or two later and just kill it and AMD will be behind a generation again.
 
I was more reacting to the announcement of more announcements to come but we're all generally on the same page here, we want to see what they (finally) bring to the table. I'd like a Ti competitive part under $1000......keep the gen1 and gen2 hardware raytracing solutions, thanks, gimme dem' VR and 4K frames instead.
 
At the very least it'll lower 2080 ti prices. Who knows when ampere will be out in volume and at what price point
 
I'm sure they will have a big Navi by end of summer most likely but the problem is Ampere will likely release a month or two later and just kill it and AMD will be behind a generation again.


Also likely show stopping launch day drivers that inhibit a proper review.
 
Well the answer for you then is "don't". I think if the announcements and stuff gets to you, just ignore the news and when it comes time to buy a GPU, just look at what's out and buy the fastest card you can afford.
AMD hasn't been competing at the top for a while, but they definitely have been competing at the low, medium, and high arenas. Just not "ultra". $1000 GPU's are less than 1% of the consumer market (2080Ti's and 2080 Supers COMBINED are less than 1%). Can't say I blame them for not pushing into an incredibly expensive but not particularly lucrative market.
Still, to be clear, I don't disagree with you: I would like more competition to drive quality up and prices down.

While I agree with the sentiment and Have Zero interest in ever buying a $1000+ GPU, simply combining two lower ranking GPUs to make your point really isn't so great, and it seems kind of arbitrary. Regular RTX 2080 is above 1% alone, and of course more expensive GPUs are going to typically sell in lower volume.

Also look where AMD is for any category in that survey, it's highest ranked card is the RX 580, in 16th place, not far ahead of that previously mentioned RTX 2080.
 
Big Navi is not going to drop until the next gen consoles are almost out the door. I think late summer early fall is realistic. Think that was always where it was going to end up.

Really until NV gets their 7nm out the door AMD is not really in any rush. The 5700s are selling well... and that no doubt is the sweet spot for both sales and margins. The 5500 has been a disappointment based on pricing... the 5600 if it proves to be 90% a 5700 will sell well as well. If its more 10% more then 5500 it will also be pointless. Guess we'll see when benchmarks for that hit. I believe AMD is pretty happy with where their GPUs are for the next 6 months though... 5700 selling well, 5600 hopefully will have a place in the market and sell well, and I expect the 5500 is priced as it is mostly cause they still have lots of 580/590s to move - Hopefully in a couple months when that old polaris stock drops so will the 5500.

I don't blame AMD for being fine with that... they will sell 5700s as fast as they can make them in the sweet spot. The 2070 super and up price points are a small market with lower overall margins. Even if they have a Big Navi they could ship tomorrow they probably don't want to cannibalize the lucrative 5700 market for another few months at least. There wise to wait and see what NVs 7nm looks like... by the summer no doubt they will have heard enough from their partners and other grapevine sources if they should be worried or not.

Personally I doubt NVs 7nm card is going to be a giant leap in performance.... I expect it will be more about smaller dies, better yields and more margin for NV. It is disapointing that the next 6 months are likely going to be boring the the GPU space... hopefully the end of 2020 does bring a big NV improvement and a Crazy good big Navi, the last few years have been mostly boring.
 
While I agree with the sentiment and Have Zero interest in ever buying a $1000+ GPU, simply combining two lower ranking GPUs to make your point really isn't so great, and it seems kind of arbitrary. Regular RTX 2080 is above 1% alone, and of course more expensive GPUs are going to typically sell in lower volume.
You missed the point of my post then. I was quoting a member (HeadRusch) who said (paraphrasing), "compete at 2080ti levels or don't already". And then I pointed out why AMD isn't in a rush to compete for 1% of the market. So, it wasn't arbitrary at all, it was a direct response.
Also look where AMD is for any category in that survey, it's highest ranked card is the RX 580, in 16th place, not far ahead of that previously mentioned RTX 2080.
What you're pointing out is that the 580 is more profitable than a Ti level card would be for them. And I agree with that. That was more or less my point to begin with.
AMD's placement in overall GPU sales obviously is also a problem, but that's a mindshare issue rather than a product issue. The 5700 and 5700XT are excellent value for money, but people currently would rather buy a card with RTX features (though they may never use them this gen due to the penalties involved) rather than get better value for dollar. So it is what it is.
 
Last edited:
Agreed, its always "We have this in the pipe" with AMD, and always there is something better when it finally drops. Shit or get off the pot AMD.
It isn't about having a top competitor at this point in the GPU market with AMD, it is about gaining market share via price wars at the entry-level (against Intel) to mid-level (against NVIDIA) and even semi-custom-level (consoles).
AMD knows what they are doing, and even if they never reach the top-end again (great for marketing but less so for overall sales) as long as they turn a profit from what they are currently doing and competing with, that is a clear win.
 
  • Like
Reactions: Mega6
like this
It isn't about having a top competitor at this point in the GPU market with AMD, it is about gaining market share via price wars at the entry-level (against Intel) to mid-level (against NVIDIA) and even semi-custom-level (consoles).
AMD knows what they are doing, and even if they never reach the top-end again (great for marketing but less so for overall sales) as long as they turn a profit from what they are currently doing and competing with, that is a clear win.

I don't agree with the second best strategy, Nvidia doesn't, that's why they're number 1.
 
I don't agree with the second best strategy, Nvidia doesn't, that's why they're number 1.
AMD knows who they are and sure its #2. So what? As Red said - as long as they are making money, they are good. No one can offer a value solution for a majority of the market like AMD. #1 bragging rights doesn't mean shit if no one can afford to buy your hardware.
 
I don't agree with the second best strategy, Nvidia doesn't, that's why they're number 1.
Intel is probably #1 on overall GPU sales numbers and real-world profits, since their iGPUs are included/embedded in nearly all of their CPUs from embedded/SoC to mid-range.
NVIDIA is at the top in terms of performance, but that doesn't mean they are at the top in terms of sales and revenue.

This is why, when AMD won the bid for the current-generation consoles (Xbone & PS4), it was such a win due to the sheer number of sales of their semi-custom APUs.
Even if their profit margins are thin and performance isn't the best, the sales of those consoles are in the millions, and that means millions to billions in overall revenue for AMD.

How that could not possibly be a great business strategy, is beyond me.
 
Last edited:
AMD knows who they are and sure its #2. So what? As Red said - as long as they are making money, they are good. No one can offer a value solution for a majority of the market like AMD. #1 bragging rights doesn't mean shit if no one can afford to buy your hardware.

The green eyed monster is selling hardware like hotcakes making hand over fist and still nobody can afford to buy their hardware....
 
Intel is probably #1 on overall GPU sales numbers and real-world profits, since their iGPUs are included/embedded in nearly all of their CPUs from embedded/SoC to mid-range.
NVIDIA is at the top in terms of performance, but that doesn't mean they are at the top in terms of sales and revenue.

This is why when AMD won the bid for the current-generation consoles (Xbone & PS4), it was such a win due to the sheer number of sales of their semi-custom APUs.
Even if their profit margins are thin and performance isn't the best, the sales of those consoles are in the millions, and that means millions to billions in overall revenue for AMD.

How that could not possibly be a great business strategy, is beyond me.
Seen stats in the past showing AMD has about 50% of total GPU market (excluding iGPUs) due to console design wins.
 
It isn't about having a top competitor at this point in the GPU market with AMD, it is about gaining market share via price wars at the entry-level (against Intel) to mid-level (against NVIDIA) and even semi-custom-level (consoles).
AMD knows what they are doing, and even if they never reach the top-end again (great for marketing but less so for overall sales) as long as they turn a profit from what they are currently doing and competing with, that is a clear win.

So, let me get this straight... AMD says "We will have a high end Navi" and people go "god I hope so" and your response is "AMD doesnt care about the top competitor".

Ok.
 
So, let me get this straight... AMD says "We will have a high end Navi" and people go "god I hope so" and your response is "AMD doesnt care about the top competitor".

Ok.

YAGpXPd.png



...damn, they did just say that.
Well, my thoughts are correct circa Q4 2019, just not so much in Q1 2020... :whistle:
 
Seen stats in the past showing AMD has about 50% of total GPU market (excluding iGPUs) due to console design wins.
Why stop at 50%? Let's take that one step further and not only exclude IGPs, but also exclude the PC dGPU market, voila - AMD has *100% of the GPU market!
 
Why stop at 50%? Let's take that one step further and not only exclude IGPs, but also exclude the PC dGPU market, voila - AMD has *100% of the GPU market!
Because an iGPU isn't always required for a system to operate, but in the systems used in that stat, the GPUs are a mandatory part of the system..
But I get where you are coming from.
 
Status
Not open for further replies.
Back
Top