SK Hynix reports record profits despite DDR4 "shortages"

[asus killer]

1 hour ago, Taf the Ghost said:

It would cost AMD & Nvidia a lot of money to bring on more capacity late in a generation, and if the mining bubble bursts, they lose new sales to cheap 2nd hand GPUs. It's better for both major players to keep their production targets the same. It's bad for the end-user, but the economics are pretty straight forward.

Nvidia maybe. AMD no! The vega are just a couple months old and a new architecture.

[Taf the Ghost]

10 minutes ago, asus killer said:

Nvidia maybe. AMD no! The vega are just a couple months old and a new architecture.

HBM2 supply limits whatever AMD wants to do with Vega, right along with them selling more to Apple & Intel than they would in Mainstream.

[asus killer]

23 minutes ago, Taf the Ghost said:

HBM2 supply limits whatever AMD wants to do with Vega, right along with them selling more to Apple & Intel than they would in Mainstream.

That's actually what doesn't make sense in all this. If you make a component (the actual gpu) for a product that is sold out for months, that as a problem whit another component supply (hbm) and what do you do?

Launch another sku of that product. And worst they did this to supply a rival company. 

 

Weird

[cj09beira]

7 minutes ago, asus killer said:

That's actually what doesn't make sense in all this. If you make a component (the actual gpu) for a product that is sold out for months, that as a problem whit another component supply (hbm) and what do you do?

Launch another sku of that product. And worst they did this to supply a rival company. 

 

Weird

there is probably more supply of slower hbm 2 than the faster stuff

[Murasaki]

yeah thats nice im still on 8 gigs because i cant find my hyperx kit fucks sake

[thorhammerz]

5 hours ago, Eibe said:

I am not surprised either as it means everything they produce it sells immediately for record prices. 

I am just saying that it is not good for us average consumers

Your average consumer is likely dead last on the priority list when corporate/enterprise clients are willing (and able) to pay much, much, more. All for the small cost of system builders potentially buying fewer products overall, but still forking over the same (if not more) money overall.

 

Quote

If they managed to ramp up production, thus lowering prices, they would make even more profits, as long as demand does not die. But memory demand is unlikely to remain at these levels for long, so they might not have the incentives to increase their production facilities. 

 

(Obviously I am assuming elastic demand, truth is that probably firms have an inelastic demand curve at these prices in economics terms). 

 

Hence why any production ramp up will be slow, incremental, and unlikely to be substantial (Samsung's new China facilities that will begin production over the course of the next 2 years will only increase supply by perhaps 10-15%). The semi-conductor/memory industry has historically operated in a cyclical pattern - there's literally no point in increasing production now, and a lot of it when management can reasonably assume that clients, especially enterprise (screw the average consumers) are able to absorb price increases, and that the demand for the current sets of chips (e.g. DDR4) is not guaranteed to persist for the medium/long term.

 

TLDR: charge more for memory prices now, while you can still charge more for it!

[Misanthrope]

Yeah well stfu about shortages then you price fixing motherfuckers.

[Okjoek]

3 hours ago, mynameisjuan said:

Please dont bring bullshit conspiracy theories here.

It's true I don't have proof to back up my hypothesis, but mine is just as valid a theory as his was. Companies/states are known to do these things.

 

[mynameisjuan]

1 minute ago, Okjoek said:

It's true I don't have proof to back up my hypothesis,

Then its not true like at all. its pure nonsense. 

[Okjoek]

Just now, mynameisjuan said:

Then its not true like at all. its pure nonsense. 

You have done nothing to prove or disprove my hypothesis.

 

It's just like a country such as Saudi Arabia artificially controlling the supply of oil.

[mynameisjuan]

1 minute ago, Okjoek said:

You have done nothing to prove or disprove my hypothesis.

YOU LITERALLY FUCKING SAID YOU DONT HAVE PROOF TO YOUR OWN HYPOTHESIS 

[Okjoek]

Just now, mynameisjuan said:

YOU LITERALLY FUCKING SAID YOU DONT HAVE PROOF TO YOUR OWN HYPOTHESIS 

That doesn't make it any more or less a hypothesis just because I can't prove it to be true or not. 

[mynameisjuan]

Just now, Okjoek said:

That doesn't make it any more or less a hypothesis just because I can't prove it to be true or not. 

You have no fucking clue how arguments work do you. "I have no proof but its true"....im sorry but the DDR4 demand spike when phones switched from DDR3 and the factories that were damaged by floods last year alone blow your dumb hypothesis with ZERO support out the window. 

 

 

[NumLock21]

What a stupid report by Sk Hynix. Record high profits? More like a desperate attempt to get some good new out of the company.

And to blame Intel for the cause of dram price hikes? This dram price hike happened way before anyone knew about the Intel cpu flaws.

Pathetic.

[Okjoek]

1 minute ago, mynameisjuan said:

"I have no proof but its true"

I never wrote that.

6 minutes ago, mynameisjuan said:

 the DDR4 demand spike when phones switched from DDR3 and the factories that were damaged by floods last year alone blow your dumb hypothesis with ZERO support out the window. 

It is plausible that those things are still causing the price issues.

[mynameisjuan]

5 minutes ago, Okjoek said:

I never wrote that.

[rawrdaysgoby]

Stortages are actually good for companies actually they get to raise the prices of things while making less of the products because of Supply and demand, terrible for consumers 

[Okjoek]

2 minutes ago, mynameisjuan said:

That's not what you're claiming I said.

 

"I have no proof but its true"

 

I omitted the implied word "that" from the sentence.

 

It's true that I don't have proof to back up my hypothesis

 

[Nocte]

7 hours ago, Notional said:

Yes, yes it does. At least in this case. None of the nand/ram vendors wants to start a price war; something that only ever happens when supply > demand.

 

There isn't a demand for tripple the amount of ram. If they manufacture that and the profits lowers by 66+%, then they make less. In fact they might make a lot less, as they get massive investment costs, increased fixed costs, and might even endure increased variable costs.

 

Why do you think GPU vendors don't care that GPU prices are through the roof? They sell every single card they make without risk, and don't have to lower prices as a usual product lifecycle dictates.

There is demand for more RAM, did you even read the article?

 

I have already said that they are not going to increase prodiction as it would create too many risks. 

 

And I have already stated the assumption of elastic demand, and refuted it in the same sentence saying why it is like it is. 

 

Read all my posts please... I really do not need economics teachings. 

[Dabombinable]

4 hours ago, cj09beira said:

there is probably more supply of slower hbm 2 than the faster stuff

Which isn't helped by the faster/up to spec HBM2 being released far later than it was supposed to be. I swear HBM might end up being another case of RDRAM the way its going.

Edit: FYI, RDRAM is the extremely costly stuff that limited a lot of people to no more than 512MB of RAM in the early 2000's due to the sheer expense associated with it.

[Notional]

35 minutes ago, Eibe said:

There is demand for more RAM, did you even read the article?

 

I have already said that they are not going to increase prodiction as it would create too many risks. 

 

And I have already stated the assumption of elastic demand, and refuted it in the same sentence saying why it is like it is. 

 

Read all my posts please... I really do not need economics teachings. 

Well, you seem to, if you don't understand a higher demand than supply results in higher prices.

 

Price elasticity is not particularly interesting here, as it is quite low. With a 100+% price increase, they are still selling everything they make. An increase in production would not increase demand all that much, just lower the profit margin of the ram itself.

[leadeater]

40 minutes ago, Dabombinable said:

Which isn't helped by the faster/up to spec HBM2 being released far later than it was supposed to be. I swear HBM might end up being another case of RDRAM the way its going.

Edit: FYI, RDRAM is the extremely costly stuff that limited a lot of people to no more than 512MB of RAM in the early 2000's due to the sheer expense associated with it.

Rambus DRAM users club wooo, PS2 and N64 had it too .

[Nocte]

21 minutes ago, Notional said:

Well, you seem to, if you don't understand a higher demand than supply results in higher prices.

 

Price elasticity is not particularly interesting here, as it is quite low. With a 100+% price increase, they are still selling everything they make. An increase in production would not increase demand all that much, just lower the profit margin of the ram itself.

Lol?

 

Okay, where did I say there is enough supply? Nowhere. 

 

Where did I say it was elastic? I said "assuming it was elastic, but in tuth it is not". 

 

So unless you are just trying to throw attacks around, then get your facts straight. 

 

How would an increase in production increase demand? That must be some sort of sci-fi. I said that if the demand was elastic (which it is for the average consumer, not firms) they would make more money by increasing production. But as the demand of individuals is nothig compared to the one of firms, then it is obvious that they are operating on the inelastic side of demand. Still, an increase in production would yield them higher profits especially if they operate on the inelastic side of demand as they can keep the prices high despite higher production. Someone else here needs an economics lesson here, not me. 

[Notional]

25 minutes ago, Eibe said:

Lol?

 

Okay, where did I say there is enough supply? Nowhere. 

 

Where did I say it was elastic? I said "assuming it was elastic, but in tuth it is not". 

 

So unless you are just trying to throw attacks around, then get your facts straight. 

 

How would an increase in production increase demand? That must be some sort of sci-fi. I said that if the demand was elastic (which it is for the average consumer, not firms) they would make more money by increasing production. But as the demand of individuals is nothig compared to the one of firms, then it is obvious that they are operating on the inelastic side of demand. Still, an increase in production would yield them higher profits especially if they operate on the inelastic side of demand as they can keep the prices high despite higher production. Someone else here needs an economics lesson here, not me. 

If supply wasn't limited, we wouldn't see a 100+% price increase, so it's a given.

 

You assumed price elasticity, don't try and change your argument. Including b2b market now makes little sense, as they are fundamentally different than consumer markets, and also use different products. So you gotta find out which market you are focusing on here (could be both, sure).

 

Consumer market (including this one) is price elastic to some degree. Increased production could result in people buying larger kits (16GB instead of 8GB, or 32 instead of 16GB for instance) if pricing lowered enough. And people buying a system now instead of waiting. So the market is quite elastic in the bottom end, but in the higher end much less so. After all, you need at least 8GB ram in a new system, whether that costs 30$ or 300$.

But one of the most basic rules in supply/demand theory is that selling more units doesn't necessarily result in higher profits. Often you make more by selling less, as you can increase profit margins. Your assumptions seem to be that an increase in production would be fully demanded at current prices, as perfectly elastic supply. But that is hardly the case.

 

If we assume an indirect cartel is happening, due to no one entering competitive pricing, and keeping production volume below optimal, we will see a price increase that explains massive profits and low availability. We see P₀ pricing rather than the competitive P_c pricing. This graph also shows why increased sales might not necessarily result in higher profits in the MC curve.

 

[Nocte]

23 minutes ago, Notional said:

You assumed price elasticity, don't try and change your argument.

If I assumed price elasticitiy my argument would be right. 

I also refuted elasticity in reality... I introduced the b2b market long before that reply? It seems you are trying to change what I said to fit your answer? I don't see your point here...

 

I know you are right, as I know you know what you are talking about, I was trying to make a point against hordes of quotes against me of people who only know "demand and supply" for what concerns economics. I just do not like people throwing arguments against me with little knowledge of the subject after studying economics for almost half of my life time.

 

EDIT: It's late night here. I am going to reply tomorrow.