Overclock.net › Forums › Industry News › Hardware News › [BusinessKorea]Samsung and SK Hynix to Supply High Bandwidth Memory to Nvidia
New Posts  All Forums:Forum Nav:

[BusinessKorea]Samsung and SK Hynix to Supply High Bandwidth Memory to Nvidia - Page 4

post #31 of 55
I don't see why it even matters who had it first. What I am curious about is who will implement it best, not first. It is quite clear that both companies will have HBM equipped video cards next year, whose will be best is yet to be determined.




Quote:
Originally Posted by prjindigo View Post

I wouldn't put it past nV to mount 8 HBM1 on a poser and run a fake 2048 bit wide bus with a 1024 bit wide controller. Whatever they release will be 12.2/12.0 compliant and run faster than a 970. Since the 980/970/X arent even fully 12.0 compliant they'll disable the software work-arounds for unsupported features and leave all you poor ducks who bought lies to buy new $500 cards.




To you nutbags that think nV is winning OR that intel and nV control the standards; AMD is about to change the way computers work AGAIN. intel and nV have had very very little input this time.

That article is full of "dominance talk" but is total BS. AMD removed north-bridge/south-bridge. AMD implemented hypertransport. AMD implemented sub-core. AMD made the first APU's. AMD made DX12 happen and provided 60+% of the code for it. AMD proved HBM to the market. AMD was the first to near linear performance from multiple cards. AMD was the first to six cores.

If AMD hadn't kicked intel's kanasta in court, nVidia wouldn't have even existed. AMD is responsible directly or through their actions for 75% of what we have in our boxes today.

Don't believe me? Look up BTX. Read up on the fight 3Dfx had with intel over SLi that AMD sent lawyers to help 3Dfx win.

None of which has led to a market dominance. There is nothing wrong with saying that Nvidia is dominating the discrete gpu martket, it is quite clear that they are. This does not minimize the importance of AMD nor does it discount their past success, but their past success or their importance does not minimize the fact that Nvidia is currently dominating the discrete GPU market.
Computer
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7-4790k Gigabyte Z97X Gaming 7 Nvidia GTX 1080ti FE (EK block) CORSAIR Vengeance 
Hard DriveHard DriveCoolingOS
Samsung 850 Evo Mushkin Reactor Custom Loop Windows 10 
MonitorKeyboardPowerCase
Ben Q BL3201PT Logitech G110 EVGA 1300w Thermatake Core P5 
MouseAudio
LOGITECH G502 PROTEUS CORE Corsair 2100 
  hide details  
Reply
Computer
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7-4790k Gigabyte Z97X Gaming 7 Nvidia GTX 1080ti FE (EK block) CORSAIR Vengeance 
Hard DriveHard DriveCoolingOS
Samsung 850 Evo Mushkin Reactor Custom Loop Windows 10 
MonitorKeyboardPowerCase
Ben Q BL3201PT Logitech G110 EVGA 1300w Thermatake Core P5 
MouseAudio
LOGITECH G502 PROTEUS CORE Corsair 2100 
  hide details  
Reply
post #32 of 55
Quote:
Originally Posted by Serios View Post

The thing is HBM was specifically developed by AMD.
Nvidia most likely had proper access to HBM after it became a standard.

So what? Welcome to how computers and standards work.

Did AMD invent PCI? Are you suggesting they shouldn't be allowed to use it because Intel invented it and made it a standard?
post #33 of 55
Quote:
Originally Posted by Serios View Post

I don't jump to any conclusions, you are simply wrong, Samsung was not SUPPLYING Nvidia whit HBM, some samples here and there can't be considered "supplying", Hynix is supplying AMD whit HBM right now for example.

So when Samsung supplies (or, need some other word here?) Nvidia with sample chips for product design and validation, are the chips magically appearing out of thin air at Nvidia's lap? Are they donations through 3rd party companies? Are the chips make believe and non existent? You are just trying to take one definition of the word supply and put it in a context when I am using the word in another. Either way, Samsung and probably Hynix have both somehow managed to create some HBM chips and somehow managed to make those chips arrive at Nvidia, the word you choose to describe such an action is up to you I guess.

I choose to use the wording "going in to production" to mean a product is moving from testing and validation samples into final production samples and capacity, and the word "supplying" to mean one company has furnished, or given, or equipped another company with some product to do initial designs and testing with even while the product in question is not fully ready for final production yet.
Gaming
(17 items)
 
Gaming PC
(20 items)
 
 
CPUMotherboardGraphicsRAM
7700K AS Rock Z170 OC Formula Titan X Pascal 2050MHz 64GB DDR4-3200 14-14-14-34-1T 
Hard DriveHard DriveHard DriveCooling
950 EVO m.2 OS drive 850 EVO 1TB games drive Intel 730 series 500GB games drive Custom water cooling 
OSMonitorKeyboardPower
Win 10 Pro x64 AMH A399U E-Element mechanical, black switches, Vortex b... EVGA G3 1kw 
CaseMouseAudioAudio
Lian-Li PC-V1000L Redragon M901 LH Labs Pulse X Infinity DAC Custom built balanced tube amp with SS diamond ... 
Audio
MrSpeakers Alpha Prime 
  hide details  
Reply
Gaming
(17 items)
 
Gaming PC
(20 items)
 
 
CPUMotherboardGraphicsRAM
7700K AS Rock Z170 OC Formula Titan X Pascal 2050MHz 64GB DDR4-3200 14-14-14-34-1T 
Hard DriveHard DriveHard DriveCooling
950 EVO m.2 OS drive 850 EVO 1TB games drive Intel 730 series 500GB games drive Custom water cooling 
OSMonitorKeyboardPower
Win 10 Pro x64 AMH A399U E-Element mechanical, black switches, Vortex b... EVGA G3 1kw 
CaseMouseAudioAudio
Lian-Li PC-V1000L Redragon M901 LH Labs Pulse X Infinity DAC Custom built balanced tube amp with SS diamond ... 
Audio
MrSpeakers Alpha Prime 
  hide details  
Reply
post #34 of 55
Thread Starter 
Quote:
Originally Posted by Serios View Post

The thing is HBM was specifically developed by AMD.
Nvidia most likely had proper access to HBM after it became a standard.

No it was developed by Hynix because AMD demanded something that was efficient yet could give bandwidth enough for future use.
People may give way too much credit to AMD for this. I bet AMDs contribution was supplying chips that could test the HBM stacks for Hynix.

Like the article says, Hynix have been supplying Nvidia too for a good while now. There goes the HBM exclusivity BS thats been floating around for far too long and was put out there by one of the worst rumor sites out there yet people gave it far too much value.
Right now AMD and Nvidia both are regular customers to Hynix. The difference is that AMD have a product that use HBM now while Nvidia is currently stockpiling HBM and building upcoming Pascal cards that will use it.
Edited by iLeakStuff - 9/22/15 at 12:54pm
post #35 of 55
Quote:
Originally Posted by EniGma1987 View Post

So when Samsung supplies (or, need some other word here?) Nvidia with sample chips for product design and validation, are the chips magically appearing out of thin air at Nvidia's lap? Are they donations through 3rd party companies? Are the chips make believe and non existent? You are just trying to take one definition of the word supply and put it in a context when I am using the word in another. Either way, Samsung and probably Hynix have both somehow managed to create some HBM chips and somehow managed to make those chips arrive at Nvidia, the word you choose to describe such an action is up to you I guess.

I choose to use the wording "going in to production" to mean a product is moving from testing and validation samples into final production samples and capacity, and the word "supplying" to mean one company has furnished, or given, or equipped another company with some product to do initial designs and testing with even while the product in question is not fully ready for final production yet.
Most likely Samsung gives samples to anybody who wants to buy their chips, even AMD.
You said Samsung was already supplying Nvidia with HBM which implies the fact that Nvidia has been buying HBM chips from Samsung since July-June an that is simply not true.
post #36 of 55
Quote:
Originally Posted by 47 Knucklehead View Post

So what? Welcome to how computers and standards work.

Did AMD invent PCI? Are you suggesting they shouldn't be allowed to use it because Intel invented it and made it a standard?
Learn to read the context, the user I responded suggested that Nvidia had even from the start equal knowledge about HBM compared to AMD the company that actually developed it. That most definitely was not the case.
Edited by Serios - 9/22/15 at 1:11pm
post #37 of 55
Quote:
Originally Posted by iLeakStuff View Post

No it was developed by Hynix because AMD demanded something that was efficient yet could give bandwidth enough for future use.
People may give way too much credit to AMD for this. I bet AMDs contribution was supplying chips that could test the HBM stacks for Hynix.

Like the article says, Hynix have been supplying Nvidia too for a good while now. There goes the HBM exclusivity BS thats been floating around for far too long and was put out there by one of the worst rumor sites out there yet people gave it far too much value.
Right now AMD and Nvidia both are regular customers to Hynix. The difference is that AMD have a product that use HBM now while Nvidia is currently stockpiling HBM and building upcoming Pascal cards that will use it.
Hynix was involved in the final stages of the development, really all the info is on the internet what doesn't anybody read it?
Quote:
Macri said the HBM development effort started inside of AMD seven years ago, so not long after GDDR5 was fully baked. He and his team were concerned about the growing proportion of the total PC power budget consumed by memory, and they suspected that memory power consumption would eventually become a limiting factor in overall performance.
post #38 of 55
Quote:
Originally Posted by iLeakStuff View Post

No it was developed by Hynix because AMD demanded something that was efficient yet could give bandwidth enough for future use.
People may give way too much credit to AMD for this. I bet AMDs contribution was supplying chips that could test the HBM stacks for Hynix.

rolleyes.gif

Nvidia and AMD will use HBM 2, end of story. But implying that AMD had hardly anything to do with HBM's development is just crazy talk..
post #39 of 55
Thread Starter 
Quote:
Originally Posted by Serios View Post

Hynix was involved in the final stages of the development, really all the info is on the internet what doesn't anybody read it?

"Development inside AMD"
Translation: We want HBM for our GPUs. Who can we contact to make it happen? tongue.gif

Quote:
Originally Posted by GorillaSceptre View Post

rolleyes.gif

Nvidia and AMD will use HBM 2, end of story. But implying that AMD had hardly anything to do with HBM's development is just crazy talk..
"End of story"? No this isnt the end.

  • The OP says Hynix have been supplying Nvidia with HBM1 for a while. There is a reason for that.
  • We have read the rumors many months ago that a Pascal chip had taped out and was ready for Q1 release. With HBM2 that doesnt exist...?
  • Then its the fact that Hynix and Samsung doesnt begin HBM2 mass production until sometime early next year according to the article. Will Nvidia sit idling and wait for summer to finally have enough HBM2 to release a card, or will they release a card with HBM1 thats been available for many months so they can get the first Pascal out the door?

All we have is rumors about Pascal using HBM2 in general. So don`t be so sure.
post #40 of 55
Quote:
Originally Posted by Serios View Post

Most likely Samsung gives samples to anybody who wants to buy their chips, even AMD.
duh? ...

Quote:
Originally Posted by Serios View Post

You said Samsung was already supplying Nvidia with HBM which implies the fact that Nvidia has been buying HBM chips from Samsung since July-June an that is simply not true.
You know Nvidia actually held a Pascal prototype validation board on stage in front of hundreds of thousands of people with HBM chips on the card back in March right?
Samsung has had samples of their chips ready for months now, and Hynix had samples ready for quite some time last year.

Gaming
(17 items)
 
Gaming PC
(20 items)
 
 
CPUMotherboardGraphicsRAM
7700K AS Rock Z170 OC Formula Titan X Pascal 2050MHz 64GB DDR4-3200 14-14-14-34-1T 
Hard DriveHard DriveHard DriveCooling
950 EVO m.2 OS drive 850 EVO 1TB games drive Intel 730 series 500GB games drive Custom water cooling 
OSMonitorKeyboardPower
Win 10 Pro x64 AMH A399U E-Element mechanical, black switches, Vortex b... EVGA G3 1kw 
CaseMouseAudioAudio
Lian-Li PC-V1000L Redragon M901 LH Labs Pulse X Infinity DAC Custom built balanced tube amp with SS diamond ... 
Audio
MrSpeakers Alpha Prime 
  hide details  
Reply
Gaming
(17 items)
 
Gaming PC
(20 items)
 
 
CPUMotherboardGraphicsRAM
7700K AS Rock Z170 OC Formula Titan X Pascal 2050MHz 64GB DDR4-3200 14-14-14-34-1T 
Hard DriveHard DriveHard DriveCooling
950 EVO m.2 OS drive 850 EVO 1TB games drive Intel 730 series 500GB games drive Custom water cooling 
OSMonitorKeyboardPower
Win 10 Pro x64 AMH A399U E-Element mechanical, black switches, Vortex b... EVGA G3 1kw 
CaseMouseAudioAudio
Lian-Li PC-V1000L Redragon M901 LH Labs Pulse X Infinity DAC Custom built balanced tube amp with SS diamond ... 
Audio
MrSpeakers Alpha Prime 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Hardware News
Overclock.net › Forums › Industry News › Hardware News › [BusinessKorea]Samsung and SK Hynix to Supply High Bandwidth Memory to Nvidia