[VC][TT] - [Rumor] Radeon Rx 300: Bermuda, Fiji, Grenada, Tonga and Trinidad

[HurleyBird]

Link?

 

[Cloudfire777]

Link?

"The black market" is the source according to leaker

 

[artivix]

A little bird tweeted this today. Its been posted in several sites so it could be legit.

Quote:
380x 319 달러
390x 499 달러

달러 = Dollar

The check of the Korean RRA product clearance website shows two C87xxx video card products cleared in the past few weeks too. Some people think they are the AMD 370x and 380x for spring release and 390x Bermuda around summer.

http://rra.go.kr/m/status/selist.js..._no=&maker=&nation=&fromdate=&todate=&x=0&y=0

 

[ShintaiDK]

PRINTED CIRCUIT BOARD ASSEMBLY FOR PERSONAL COMPUTER(VIDEO/ GRAPHICS CARD) 102-C87201-00.

PRINTED CIRCUIT BOARD ASSEMBLY (VIDEO GRAPHIC CARD)HF SAMOA 2GB DDR3 900E/1000M P/N:102-C87001-00

So the 87001 is a 2GB DDR3 card. I dont hope that will cost 319$

https://www.zauba.com/

Also seems AMD calls one of them C880:
AMD C880 PRINTED CIRCUIT BOARD ASSEMBLY (VIDEO GRAPHIC CARD)WITH COOLER MASTER HEATSINK P/N.102-C88001-00 (FOC)
PRINTED CIRCUIT BOARD ASSEMBLY (VIDEO GRAPHIC CARD)C880 FIJI XT P/N.102-C88001-00

 

[artivix]

Hardware France is reporting single AMD Fiji (probably meaning 380x instead of 390x) was used for the demos at GDC 2015, so release of new 3xxx cards could be a matter of weeks away:

AMD obviously offered a demonstration LiquidVR based on Oculus Rift Crescent Bay and animation Showdown with an excellent result. Difficult though without comparing other solutions side-by-side to say that LiquidVR provides a significant advantage. When we tested this same demo during an Nvidia event last September, the result was also very good.

Note that the system used by AMD did not use multi-GPU but to a graphics card based on a new GPU. In other words, it was a Radeon R9 390X GPU and Fiji.

https://translate.googleusercontent.com/translate_c?act=url&depth=1&hl=en&ie=UTF8&prev=_t&rurl=translate.google.com&sl=fr&tl=en&u=http://www.hardware.fr/news/14109/gdc-amd-mise-vr-annonce-liquidvr.html&usg=ALkJrhi-0oPLVNlOsQr0TibWOCTVnpd8sA

 

[ShintaiDK]

By the shipping numbers there is no such thing as 390X or 380X. But C880 instead. Perhaps to get the numbering on the same page as CPUs.

 

[Kenmitch]

By the shipping numbers there is no such thing as 390X or 380X. But C880 instead. Perhaps to get the numbering on the same page as CPUs.

Smoke and mirrors. Description could be purposely misleading. After all calling the GTX 970 a 4GB card is perfectly acceptable to you.

Part could also be a low powered laptop gpu I guess.

 

[ShintaiDK]

Smoke and mirrors. Description could be purposely misleading.
Part could also be a low powered laptop gpu I guess.

Its a video card. Its Fiji XT and it contains a cooler master heatsink. We know cooler master will supply it.

And no, its not smoke and mirrors or misleading. Its customs data. One place you dont want to place lies.

 

[stahlhart]

Smoke and mirrors. Description could be purposely misleading. After all calling the GTX 970 a 4GB card is perfectly acceptable to you.

Part could also be a low powered laptop gpu I guess.

This is the last time I'm going to warn you to stop trolling in this forum.
-- stahlhart

 

[Vesku]

PRINTED CIRCUIT BOARD ASSEMBLY FOR PERSONAL COMPUTER(VIDEO/ GRAPHICS CARD) 102-C87201-00.

PRINTED CIRCUIT BOARD ASSEMBLY (VIDEO GRAPHIC CARD)HF SAMOA 2GB DDR3 900E/1000M P/N:102-C87001-00

So the 87001 is a 2GB DDR3 card. I dont hope that will cost 319$

https://www.zauba.com/

Also seems AMD calls one of them C880:
AMD C880 PRINTED CIRCUIT BOARD ASSEMBLY (VIDEO GRAPHIC CARD)WITH COOLER MASTER HEATSINK P/N.102-C88001-00 (FOC)
PRINTED CIRCUIT BOARD ASSEMBLY (VIDEO GRAPHIC CARD)C880 FIJI XT P/N.102-C88001-00

Cooler Master heatsink is consistent with the rumored AIO liquid cooling of the 380/390.

 

[RussianSensation]

$499 for a 390X which based on rumoured paper specs looks to be 45-50% faster than a 290X sounds WAY too good to be true. I would have believed this in the old GPU eras where performance went up 75%+ every 18 months but not anymore. We can reasonably deduce that the top GM200 version won't be $500. Therefore, AMD has no reason to price the 390X below $550, I would say even $650, if the rumoured performance comes to fruition.

Unless GM200 launches first and ends up 20-30% faster than 390X, why does AMD need to low ball so much? They can drop the price to $499 if necessary, only if GM200 smokes their chip badly. Either the 45-50% faster than 290X is wishful thinking or the prices are just made up. I mean 980 is less than 15% faster (just 10-11% faster at 4k) than a 290X today and it sells well at $550-600, with not a single price drop so far.

If 390X is $500 and is up to 50% faster than a 290X, it would sell out for months unless GM200 is out of this world amazing. Why would AMD go for the single chip performance crown but sell their card for only $500 after NV sold a mid-range next gen chip for $550? NV already proved that gamers are willing to pay $550+ for 10-15% more performance over a 290X. If 390X is truly a winner, AMD would be losing a lot of profits not selling this card at $600+. The reason I say this is AMD's price performance strategy has failed to make them money so the next reasonable strategy to go for is --> Premium performance @ a premium price.

Also, the $319 price for 380X seems odd because rumours have that chip similar in performance to a 980. Even at $399 it would sell.

If 390X is actually 45-50% faster than a 290X and is $499, that would be one of the most bizarre cases. As a consumer I would love that but as far as AMD's profitability strategy, it would leave me scratching my head. I think Lisa Su needs to realize that maintaining market share but not making much money is a waste of time. Might as well be like iOS with <15% market share but healthy profits. I think a lot of gamers today would choose a profitable AMD with low market share than AMD close to bankruptcy with 40-50% market share because they are "giving away" their products at low margins that aren't even enough to cover expenses. All things considered, I will be shocked if 390X lives up to the performance claims but is only $500, and manages to actually make AMD money, something that HD7000/200 series failed to do.

I have to say this may be one of AMD's most secretive launches in a while. We still do not have a single legitimate benchmark leak 1.5 years since 290X launched. The specs might match but no one really knows what the performance increase is over the 290X - 20%, 30%, 65%? It's pure guessing. It would also be odd to go with expensive HBM and > 500GB/sec memory bandwidth and end up < 30% faster than a 290X.

 

[destrekor]

By the shipping numbers there is no such thing as 390X or 380X. But C880 instead. Perhaps to get the numbering on the same page as CPUs.

That is just internal designation for the specific part, not the entire graphics card as sold to consumers.

The Hawaii XT board was C671:
PRINTED CIRCUIT BOARD ASSEMBLY (VIDEO GRAPHIC CARD)C671 HAWAII XT P/N.102-C67101-00

 

[nvgpu]

AMD is willing to subsidize and lose money on expensive 500mm2++ die size GPU, 4GB HBM memory(less production quantity than GDDR5 and SK Hynix certainly will want higher price for it), 10 layer or more PCB & 300W water cooling solution, so what are you questioning them for? Just buy it and enjoy it.

 

[raghu78]

$499 for a 390X which based on rumoured paper specs looks to be 45-50% faster than a 290X sounds WAY too good to be true. I would have believed this in the old GPU eras where performance went up 75%+ every 18 months but not anymore. We can reasonably deduce that the top GM200 version won't be $500. Therefore, AMD has no reason to price the 390X below $550, I would say even $650, if the rumoured performance comes to fruition.

Unless GM200 launches first and ends up 20-30% faster than 390X, why does AMD need to low ball so much? They can drop the price to $499 if necessary, only if GM200 smokes their chip badly. Either the 45-50% faster than 290X is wishful thinking or the prices are just made up. I mean 980 is less than 15% faster (just 10-11% faster at 4k) than a 290X today and it sells well at $550-600, with not a single price drop so far.

If 390X is $500 and is up to 50% faster than a 290X, it would sell out for months unless GM200 is out of this world amazing. Why would AMD go for the single chip performance crown but sell their card for only $500 after NV sold a mid-range next gen chip for $550? NV already proved that gamers are willing to pay $550+ for 10-15% more performance over a 290X. If 390X is truly a winner, AMD would be losing a lot of profits not selling this card at $600+. The reason I say this is AMD's price performance strategy has failed to make them money so the next reasonable strategy to go for is --> Premium performance @ a premium price.

Also, the $319 price for 380X seems odd because rumours have that chip similar in performance to a 980. Even at $399 it would sell.

If 390X is actually 45-50% faster than a 290X and is $499, that would be one of the most bizarre cases. As a consumer I would love that but as far as AMD's profitability strategy, it would leave me scratching my head. I think Lisa Su needs to realize that maintaining market share but not making much money is a waste of time. Might as well be like iOS with <15% market share but healthy profits. I think a lot of gamers today would choose a profitable AMD with low market share than AMD close to bankruptcy with 40-50% market share because they are "giving away" their products at low margins that aren't even enough to cover expenses. All things considered, I will be shocked if 390X lives up to the performance claims but is only $500, and manages to actually make AMD money, something that HD7000/200 series failed to do.

I have to say this may be one of AMD's most secretive launches in a while. We still do not have a single legitimate benchmark leak 1.5 years since 290X launched. The specs might match but no one really knows what the performance increase is over the 290X - 20%, 30%, 65%? It's pure guessing. It would also be odd to go with expensive HBM and > 500GB/sec memory bandwidth and end up < 30% faster than a 290X.

There is a strong reason for AMD to push for aggressive pricing on the next gen GPUs. Nvidia has taken a massive chunk of market share from AMD. AMD lost close to 33% of GPU sales in Q4 2015(down from 35% to 24% marketshare). Q1 might not look any better. The sales momentum is with Nvidia and AMD needs to do very well with the next gen GPU product stack. They really need to knock this one out of the park . The rumoured pricing reminds me of the HD 4870. AMD needed to get back from the HD 2900XT failure. This time the situation is similar though not the same. AMD is late by a solid 9 months. They know Nvidia is really on top and has GM200 waiting. So matching Nvidia's pricing won't do AMD much good as Nvidia will not let go of the market share they gained so easily. So its definitely going to take some really aggressive pricing from AMD to reverse the market share loss. I also think that AMD will match or beat Nvidia for the single GPU crown perf. I do not expect a GM200 to be more than 35% faster than GM204.

I have a feeling AMD will start off an aggressive bid to gain the lost market share. I think they will succeed too. GM200 is going to be around 20% bigger than R9 390X in terms of die size. I don't think Nvidia will want a price war which will crater their margins. Nvidia will allow the situation to return to the traditional position, which means pricing in a manner such that Nvidia still retain 60 - 65% of the market share. AMD have a golden opportunity to gain high end notebook GPU share with HBM based R9 380x mobile products and probably take the notebook GPU crown decisively.

btw why do you assume AMD cannot be profitable selling a 500 sq mm GPU with HBM for $500. 28nm is cheap and a mature process. GF wafer pricing would be even better than TSMC given that AMD is their No.1 customer and AMD's next gen GPU products are going to be made at GF. Each GPU AMD makes at GF is a GPU not made at TSMC. GF will not spare any effort in getting AMD to move more of their products to their fabs from TSMC. AMD will have multiple salvage SKUs to handle any yield issues with the big die or with a large interposer and HBM.

btw you have to understand that Nvidia's brand is superior than AMD's. So if a R9 390X is as fast or 5-10% faster than full fat GM200 and is sold for same price the GM200 will still sell much more. AMD needs to strike a balance between per unit profit and volume. Right now their number one priority is gaining unit market share from Nvidia especially at USD 200 - USD 500. A aggressive R9 3xx price stack is a good strategy to achieve that. :thumbsup:

 

[Headfoot]

If the 390x is $500, I'm day one buying it. Triple Monitor must be fed

 

[RussianSensation]

AMD is willing to subsidize and lose money on expensive 500mm2++ die size GPU, 4GB HBM memory(less production quantity than GDDR5 and SK Hynix certainly will want higher price for it), 10 layer or more PCB & 300W water cooling solution, so what are you questioning them for? Just buy it and enjoy it.

That's not even logical. That means the more units you sell, the more money you lose and the closer you are to bankruptcy. In that case it's better to not release anything at all. Also, I may or may not upgrade to R9 390X/GM200. I might skip them entirely as my 7970s are plenty fast for the games I play. I am interested in hardware but it doesn't mean every new release I will just upgrade for the sake of upgrading.

Also, why would I buy a $500 390X without knowing the performance and price of GM200? That makes no sense.

I do not expect a GM200 to be more than 35% faster than GM204.

It's possible it will be faster. NV already announced Titan X which means they have another 2-4 months to bin GM200 to ramp up clocks. If GM200 overclocks similarly to the 980, it could be up to 50% faster.

GM200 is going to be around 20% bigger than R9 390X in terms of die size.

We don't know that. If GM200 is 600mm2, based on your estimate R9 390X is only 500mm2. How do you expect AMD to get 50% more performance over 290X with only a 500mm2 die size on 28nm?

I don't think Nvidia will want a price war which will crater their margins. Nvidia will allow the situation to return to the traditional position, which means pricing in a manner such that Nvidia still retain 60 - 65% of the market share.

More reason that if R9 390X is 50% faster than R9 290X, it's not necessary to price it at $499. Titan X will likely end up more expensive than Titan/Titan Black (rumours point to $1299-1349). Do you honestly believe that consumer GM200 will be less than $699?

Your example of 4870 isn't valid because HD4800 generation did not gain AMD any market share --> AMD didn't gain much market share with HD4800 generation (38% -> 40%). In the process, their GPU division barely made $.

AMD have a golden opportunity to gain high end notebook GPU share with HBM based R9 380x mobile products and probably take the notebook GPU crown decisively.

Laptops are completely different. We are not discussing those since 380X @ $319 and 390X @ $499 were referenced with regard to desktop parts. We don't know anything about laptop chips as 0 rumours came up for them.

btw why do you assume AMD cannot be profitable selling a 500 sq mm GPU with HBM for $500.

I am not saying they can't be profitable but given NV's pricing for 780, 780Ti and mid-range 680/980, why price a card way faster than 980 at only $499? Just go gain market share? Market share without profits is rather meaningless -- just ask Android manufacturers.

btw you have to understand that Nvidia's brand is superior than AMD's. So if a R9 390X is as fast or 5-10% faster than full fat GM200 and is sold for same price the GM200 will still sell much more.

I already know that. Even if 390X is 30% faster than GM200 at $500, GM200 will sell more. That's why AMD needs to focus on balancing profits and market share with brand agnostic users. I wouldn't even bother trying to convert NV's loyal customers. They are probably similar to most of iPhone's customers where they wouldn't even consider a competing product regardless or price, performance or features. Why would you think NV's loyal customers would convert for price/performance when for 5 generations in a row that didn't happen? (HD4000/5000/6000/7000/200)

AMD needs to strike a balance between per unit profit and volume. Right now their number one priority is gaining unit market share from Nvidia especially at USD 200 - USD 500. A aggressive R9 3xx price stack is a good strategy to achieve that. :thumbsup:

I disagree. #1 priority is profits, not market share. I'd rather have 1% market share but 99% of the industry's profits than 99% market share and 1% of the profits.

Focusing on market share at the expense of lost profits is not a strategy of successful firms.

http://www.macrumors.com/2015/02/26/ios-vs-android-profits-q4-2014/

If the 390x is $500, I'm day one buying it. Triple Monitor must be fed

What if it's a flop with just 10% faster performance over the 980? If anything the low price of $500 sounds to me like it probably won't live up to expectations. If you were the CEO of AMD and you had a chip 50% faster than R9 290X on your hands, would you price it only at $500 when your competitor's $550 chip is just 10% faster (!!!)?

980 is 10% faster than 290X at 1440P, 8% faster at 4K. Ya, that's right, the gap has shrunk that much.
http://www.techpowerup.com/reviews/Gigabyte/GTX_960_G1_Gaming/28.html

Something doesn't quite add up here....unless GM200 is an uber monster and will smack 390X by 20%+ and/or 390X is nowhere near 50% faster than a 290X.

 

[Headfoot]

What if it's a flop with just 10% faster performance over the 980? If anything the low price of $500 sounds to me like it probably won't live up to expectations. If you were the CEO of AMD and you have a chip 50% faster than R9 290X on your hands, would you price it only at $500? Something doesn't quite add up here....unless GM200 is an uber monster and will smack 390X by 20%+.

True, if it's only 10% over a 980 I'll just get a second 290. I prefer to stay single GPU when I can but the single 290 is requiring me to turn down settings quite a bit to maintain good frame rates on 3x1080p. The real tricky issue would be if it ends up being 20% faster than a 980 and doesn't overclock much. I think I gotta see at least 30% before I'll seriously consider a 390x over dual-290.

My guess if it ends up being $550 for real is that GM200 is a beast and the 390x will represent the best value high end, but not single GPU king. Maybe issues with HBM?

 

[96Firebird]

4GB VRAM doesn't concern you on 3x1080p? Do you run into VRAM trouble with your current 290?

 

[digitaldurandal]

4GB VRAM doesn't concern you on 3x1080p? Do you run into VRAM trouble with your current 290?

VRAM doesn't scale linearly with resolution. I have not had VRAM problems on my 290s.

It is about 33% less pixels than 4k.

 

[Headfoot]

4GB VRAM doesn't concern you on 3x1080p? Do you run into VRAM trouble with your current 290?

4GB of VRAM definitely concerns me for 3x1080p going forward, I would really like to see 8GB, even 6 is good... but if the choice is between $1350 w/ 12 GB or $500-550 for 4 GB and 10-20% less speed I'd go with the $550. I can swing that kind of cash, but anything around $1000 is just too rich for my blood. I can compromise by turning down textures from Ultra to High or Medium if I absolutely must. If a 990/1080/980Ti comes out with 6GB of RAM around ~$600 I'd seriously consider that too. Hopefully we see some leaked info on the inevitable cut-down Titan X card soon

I'm currently not having VRAM issues on the 4GB 290, so realistically it would only have to last me until the 14nm GPUs come out as I will likely upgrade again at that time. The only thing so far I can't run texture wise at max that I've played is Titanfall, and the max texture pack is visually indistinguishable from one step down to me. I can't do triple on the hard games to play like Crysis 3 right now due to lack of GPU power, so I run into that before I run into VRAM limitations usually.

 

[thilanliyan]

4GB VRAM doesn't concern you on 3x1080p?

It's okay...it's full speed vram...

I haven't games much lately tbh...are many games breaking the 4gb barrier (at 3x1080p)?

 

[Cloudfire777]

Great deal or what...
http://www.newegg.com/Product/Produc...-710-_-Product

 

[Silverforce11]

If the 390x is $500, I'm day one buying it. Triple Monitor must be fed

$499 for the potential 50% increase over R290X is ridiculously good (very skeptical though!). I'm in for 2 to CF!

 

[StinkyPinky]

4GB of VRAM is fine with me. I currently game at 1080p. I may upgrade to 1440p but I won't be doing 4K for some years.

Is there any news on the 380 and 390?

 

[raghu78]

We don't know that. If GM200 is 600mm2, based on your estimate R9 390X is only 500mm2. How do you expect AMD to get 50% more performance over 290X with only a 500mm2 die size on 28nm?

architectural improvements combined with HBM which moves the majority of memory controller complexity to the HBM stack thus freeing area, transistors and power on the GPU die. HBM cuts memory controller + memory chips power by 65 - 70%. Where do you think that saved power combined with improved architectural efficiency is being invested ? Towards driving performance up in a big way.

Your example of 4870 isn't valid because HD4800 generation did not gain AMD any market share --> AMD didn't gain much market share with HD4800 generation (38% -> 40%). In the process, their GPU division barely made $.

Sorry please refresh your memory with the truth. AMD gained market share and Nvidia were actually losing GPU revenue in a big way.

http://archive.news.softpedia.com/news/AMD-Gains-Market-Share-from-NVIDIA-97704.shtml

"The difference was made by the aggressive pricing approach AMD's ATI graphics unit has taken, as the price for all add-in cards has been brought down. &#8220;Priced aggressively yet delivering solid performance, AMD's new line not only took back some market share - jumping up to 40 percent from 35 percent the quarter prior - it forced Nvidia (and partners) to cut prices on its recently released GTX 200 series product,&#8221; said JPR."

http://www.nvidia.com/object/io_1226081888345.html

"For the third quarter of fiscal 2009, revenue was $897.7 million compared to $1.12 billion for the third quarter of fiscal 2008, a decrease of 20 percent. For the nine months ended October 26, 2008, revenue increased to $2.94 billion compared to $2.90 billion for the nine months ended October 28, 2007, an increase of 2 percent."

http://www.nvidia.com/object/io_1234300989054.html

"For the fourth quarter of fiscal 2009, revenue was $481.1 million compared to $1.2 billion for the fourth quarter of fiscal 2008, a decrease of 60 percent. For the twelve months ended January 25, 2009, revenue was $3.4 billion compared to $4.1 billion for the twelve months ended January 27, 2008, a decrease of 16 percen"

so please stop denying that HD 4800 series did not help AMD gain market share. HD 4800 series was the second most successful generation behind the HD 5800 generation when AMD became the largest GPU vendor (desktop + notebook combined)

http://www.xbitlabs.com/news/graphi..._on_Discrete_GPU_Market_Mercury_Research.html

I am not saying they can't be profitable but given NV's pricing for 780, 780Ti and mid-range 680/980, why price a card way faster than 980 at only $499? Just go gain market share? Market share without profits is rather meaningless -- just ask Android manufacturers.

You know AMD lost massive market share in Q4 2014 and will do in Q1 2015. How do you think AMD can gain back that market share if they just match the price of Nvidia's equivalent performing cards. Assuming a desktop Geforce GTX 990 or 980 Ti at USD 650 - 700 what do you think AMD should price the R9 390x assuming it matches the full fat GM200 perf.
At USD 700 reception would not be great and it will be difficult to win back market share. At USD 500 - 550 the press reception would be awesome and the sales momentum will be back.

I already know that. Even if 390X is 30% faster than GM200 at $500, GM200 will sell more. That's why AMD needs to focus on balancing profits and market share with brand agnostic users. I wouldn't even bother trying to convert NV's loyal customers. They are probably similar to most of iPhone's customers where they wouldn't even consider a competing product regardless or price, performance or features. Why would you think NV's loyal customers would convert for price/performance when for 5 generations in a row that didn't happen? (HD4000/5000/6000/7000/200)

You won't bother converting the loyal NV customer but what about the massive 30% market share (wrt AMD's unit volume) AMD lost. How do you think they can regain that back ? Not with parity pricing

I disagree. #1 priority is profits, not market share. I'd rather have 1% market share but 99% of the industry's profits than 99% market share and 1% of the profits.

You have a horrible understanding of the GPU market. Unit volume and unit market share across the entire GPU segment should be balanced with profit margin per GPU and ASP. The goal is to maximize revenue and unit market share and not one at the expense of other. What good is it if you have 20% of the GPU market and are having good margins but lesser revenue and overall dollar profit. Assume a GPU ASP and unit volume scenario as explained below

Eg: USD 100 ASP * 1,000,000 units = 100 million USD
USD 80 ASP * 2,000,000 units = 160 million USD

If we assume the average cost to AMD per unit was USD 50 the profit is higher with the second scenario (160 - 100 = 60 ) compared to the 1st scenario (100 - 50 = 50).

So what matters is to maximize overall revenue and profits and not just margins. What you need to do is generate more cash overall. Pursuing the highest margins while losing unit market share and revenue is a losing proposition :thumbsdown:

Focusing on market share at the expense of lost profits is not a strategy of successful firms.

Again overall revenue and free cash flow(profit after manufacturing cost is accounted) is more important than margins per unit. I think you need to go back to some basic study of volume economics.

What if it's a flop with just 10% faster performance over the 980? If anything the low price of $500 sounds to me like it probably won't live up to expectations. If you were the CEO of AMD and you had a chip 50% faster than R9 290X on your hands, would you price it only at $500 when your competitor's $550 chip is just 10% faster (!!!)?

Something doesn't quite add up here....unless GM200 is an uber monster and will smack 390X by 20%+ and/or 390X is nowhere near 50% faster than a 290X.

It adds up very well. Do you think Nvidia is going to give back the 11% unit market share easily. AMD will have to fight hard to earn it back. You are basically ignoring the ground reality of the market situation.