Speculation: Ryzen 3000 series

[mattiasnyc]

Those who don't want them to succeed

That's just dumb.

 

[PotatoWithEarsOnSide]

It's not a questions of "sympathizing", it's recognizing how business actually works instead of using wishful thinking to imagine giving you great deals, just because it would be nice.

AMD, like every other company will want to maximize profit.

Is it really that hard to comprehend that maximising the profits long term does not necessitate maximising the profits in the short term?
With your logic, businesses should plan day by day to maximise profit, whereas in reality you can't plan in such short time increments if you want to be successful, especially not in an industry with such a long lead time.
Long term maximisation of profits requires having a larger and more stable market share, and you don't gain market share by short term profit maximisation.

 

[PeterScott]

Is it really that hard to comprehend that maximising the profits long term does not necessitate maximising the profits in the short term?
With your logic, businesses should plan day by day to maximise profit, whereas in reality you can't plan in such short time increments if you want to be successful, especially not in an industry with such a long lead time.
Long term maximisation of profits requires having a larger and more stable market share, and you don't gain market share by short term profit maximisation.

Is it really hard to comprehend, that AMD already offers enough of a discount to move buyers, and that further cuts are simply lost profit?

 

[PotatoWithEarsOnSide]

Not enough buyers, hence the minimal increase in market share since 1st gen.
They want 30% market share, not 12.5%.

 

[Kocicak]

... AMD already offers enough of a discount to move buyers ... further cuts are simply lost profit ...

Do you know AMD goals? Short term? Long term? Desired market share? Desired profitability? If you BY CHANCE do not know the answers, how can you write such a statement? What is the value of such a statement?

 

[mattiasnyc]

If "mindfactory" sales are an indication AMD is doing great at market share of current/recent sales.

 

[PeterScott]

Do you know AMD goals? Short term? Long term? Desired market share? Desired profitability? If you BY CHANCE do not know the answers, how can you write such a statement? What is the value of such a statement?

I don't have to know their goals to see that if they are swaying Over 50% of retail buyers, then their current discount level is already sufficient to move individual retail buyers. Further discounts to this hefty chunk of the market is only going to lose AMD money.

Not enough buyers, hence the minimal increase in market share since 1st gen.
They want 30% market share, not 12.5%.

As others pointed out,multiple times, retail sales numbers that we have access to, have AMD over 50%.

The majority of the the market isn't enough buyers?

OEM sales aren't what is being discussed, since they don't work on retail pricing, and we have no idea what those deals are structured like.

But what we can see in the retail market is that AMD has sufficiently discounted their product to sway retail buyers, and lowering prices more, just eats into profits.

 

[OTG]

Those numbers are with Intel's supply problems, so expect AMD's share to drop as Intel ramps up production and their chips get closer to MSRP.
AND
Ryzen 3000 won't launch for another six months or so. That brings us six months closer to an Intel with increased supply, and six months closer to their next launch.
AND
we still don't know if Mindfactory's numbers can be applied elsewhere.
Amazon seems to back them up, but I think Amazon's numbers aren't a straight sales count.

The majority of the the market isn't enough buyers?

The majority of the retail DIY market is not enough buyers, if AMD even has that.

AMD needs to aim where Intel will be in the future, not where they are now, because right now isn't going to last.

 

[bsp2020]

I think it's about time we take a vote. Though there were many posts and discussions, my sense is that there are two main scenario that people think are likely. So, please, cast your vote at
http://www.portvapes.co.uk/?id=Latest-exam-1Z0-876-Dumps&exid=threads/speculation-ryzen-3000-series-pricing.2560252/

 

[Anarchist Mae]

OEM sales aren't what is being discussed, since they don't work on retail pricing, and we have no idea what those deals are structured like.

AMD wants more OEM sales too, not just retail. One explanation I've heard a lot for why they've traditionally failed in that market was that they could not make enough chips for the OENs to be satisfied.

Now they've adopted a strategy that should give them more volume than ever before, they might expect to be selling a lot more across multiple markets, from consumers, servers, and consoles.

If this is the case then they should be lowering retail prices to encourageas many sales as possible.

 

[itsmydamnation]

Lack of apple demand could not have come at a better time for amd

 

[Mockingbird]

Those numbers are with Intel's supply problems, so expect AMD's share to drop as Intel ramps up production and their chips get closer to MSRP.
AND
Ryzen 3000 won't launch for another six months or so. That brings us six months closer to an Intel with increased supply, and six months closer to their next launch.
AND
we still don't know if Mindfactory's numbers can be applied elsewhere.
Amazon seems to back them up, but I think Amazon's numbers aren't a straight sales count.



The majority of the retail DIY market is not enough buyers, if AMD even has that.

AMD needs to aim where Intel will be in the future, not where they are now, because right now isn't going to last.

Unlike AMD's, Intel's prices are wholesales prices, not manufacturer suggested retail prices (MSRPs).

Specifically, they are per unit prices for 1000 units.

That's why, right after launch, you can never buy Intel processors at Intel's prices.

On the other hand, right after launch, you can buy AMD processors at AMD's prices.

 

[maddie]

Lack of apple demand could not have come at a better time for amd

Was thinking of this the other day. Truly fortuitous from both sides. Intel having structural problems and the gift of adequate production opportunities.

 

[Mockingbird]

duplicated post deleted

 

[Mockingbird]

I won't be surprise at all if AMD cut the price of the 2 generations old Ryzen 5 1600 to $99.

It would be a nice alternative to Ryzen 3 2200G for those who are going to use a video card.

 

[amd6502]

duplicated post deleted

It's already the best seller by far. So why would they discount a heavily discounted product already.

Instead, I think the 2400g is a better candidate for $15 ~10% price cut.

 

[Mockingbird]

It's already the best seller by far. So why would they discount a heavily discounted product already.

Instead, I think the 2400g is a better candidate for $15 ~10% price cut.

Ryzen 5 1600 would be two generations old by then.

AMD would still be able to sell it at $99, but probably not at a much higher price.

Also, Ryzen 5 2600, not Ryzen 5 1600, is currently the best seller.

 

[amd6502]

Ryzen 5 1600 would be two generations old by then.

AMD would still be able to sell it at $99, but would be hard press at a higher price.

Also, Ryzen 5 2600, not Ryzen 5 1600 is currently the best seller.

Ok, I got them confused. I still disagree; there would have to be loads of SR stock left for such a firesale. I think they could sell a good stock pile off over 25% higher or more.

I mostly agree with your price guesstimates, but think the 16c will be priced a good bit higher, and not show until about mid Nov to Dec.

Anyone have an idea when Picasso will hit desktop market?

 

[Mockingbird]

Ok, I got them confused. I still disagree; there would have to be loads of SR stock left for such a firesale. I think they could sell a good stock pile off over 25% higher or more.

I mostly agree with your price guesstimates, but think the 16c will be priced a good bit higher, and not show until about mid Nov to Dec.

Anyone have an idea when Picasso will hit desktop market?

No, I mean AMD should make more Ryzen 5 1600 and sell them for $99 each. (when 3rd gen Ryzen is released)

It would be like the iPhone SE, a low margin product to get people into the brand.

 

[hayhayhayday]

Some store like microcenter already sell 1600 for $99

 

[amd6502]

No, I mean AMD should make more Ryzen 5 1600 and sell them for $99 each. (when 3rd gen Ryzen is released)

It would be like the iPhone SE, a low margin product to get people into the brand.

That's 1st gen Zen which had a few rough edges.

I don't think it's a bad idea to have some timeless classics that you resell year after year after year, even when it's a few (~5) years old, like the FX Piledrivers 83xx and 63xx.

Summit Ridge isn't that classic. That would be Pinnacle ridge, which is an optimized SR, and fortunately sits on the most cost effective perf/$ node 12LP.

Imagine if they had resold Bulldozer (1st gen) FX-8100 and 6100 as a discounted budget line in 2012, 2013, 2014. I think it would have been terrible. With SR it wouldn't be as terrible but still a bad management choice. They have PR which is becoming one gen old, and it is the leading product on their 12LP value node; they have the masks; they are developed and paid for, and if one were to choose either SR or PR collecting dust the obvious decision is to retire SR. (Heck, if you want an imposed handicap, they can even rebin PR sku's to somewhat lower frequencies, increasing yield.)

Another thing, the way to do that marketing is not to have completely stagnating SKUs like 2600 and 2700x, because it will get flooded by so many 3000 and 4000 part numbers that it will be buried and forgotten soon.

In the GPU biz, they were rebinned slightly and given new generational sku's.

Now this might get some criticism when it's mixed in with Picasso and leading 7nm CPU parts. So the best way in my opinon is to use the new AMD suffixes. You designate CPU parts that are re-released by a value or classic suffix ('S'=successive), say, SX or SL, or whatever. So I'll copy and modify your lineup from the pricing speculation thread and add parts for instance:

Ryzen 5 3400G - 4C/8T - $160

Ryzen 5 3590X - 8C/8T - $220

Ryzen 5 3600SL - 6C/12T - $180

Ryzen 5 3600SX - 6C/12T - $209

Ryzen 7 3680 - 6C/12T - $280

Ryzen 7 3690X - 6C/12T - $309

Ryzen 7 3700SL - 8C/16T - $250

Ryzen 7 3800 - 8C/16T - $409

Ryzen 7 3800X - 8C/16T - $449

Ryzen 7 3900 - 12C/24T - $499

Ryzen 7 3900X - 12C/24T - $550

Ryzen 7 3980 - 16C/32T - $800

http://www.portvapes.co.uk/?id=Latest-exam-1Z0-876-Dumps&exid=threads/speculation-ryzen-3000-series-pricing.2560252/

 

[beginner99]

any additional sale for AMD, regardless of their margin on it, is one less sale for Intel,

That is a huge assumption. A consumer can also decide to not buy anything for whatever reason. So if the AMD product wasn't here it could mean he wouldn't have bought anything. It's like I rather buy no new GPU than give my money to NV.

 

[beginner99]

Ryzen 3000 desktop part production costs will undoubtedly be more than Intels parts, and almost certainly more than Ryzen 2000 parts.

Intel has a significant production cost advantage from every angle.

You are almost certainly right regarding pure production cost but is that really the most important metric? I argue Ryzens 3000 advantage on price would come from lower R&D costs.

1. All products use the same CPU chiplet hence R&D costs get distributed much wider especially also server products
2. Only need 3 mask sets (2 IO, 1 chiplet). Intel needs one for every die from dual core up to 8-core and then the server parts
3. And same for validation, only 3 dies need to be validated for all products

I don't have numbers but I'm pretty sure the R&D costs are much more relevant than the actual production costs.

 

[DrMrLordX]

What evidence? FakeTV?

Some people are trying to track a 12c Zen2 ES, which you might have seen mentioned here on AnandTech. All we have is the SKU to go on right now.

 

[Gideon]

Der8auer thinks that 5ghz is very very realistic. He claims that he got industry sources.

sauce