Overclock.net › Forums › Industry News › Hardware News › [Guru3D] SK Hynix Launches Fastest 8Gb Graphics DRAM (GDDR6) - Adopted by 2018
New Posts  All Forums:Forum Nav:

[Guru3D] SK Hynix Launches Fastest 8Gb Graphics DRAM (GDDR6) - Adopted by 2018 - Page 9

post #81 of 108
Not sure if this was posted here or not but AMD have been working with SK-Hynix on GDDR6 since 2015: https://www.linkedin.com/in/daehyun-jun-bb135a22


Most folks find it hard to believe but it is AMD who has their hands in memory design, not nVIDIA. nVIDIA just swoops in and uses the tech... they don't design anything memory related afaik.
Edited by Mahigan - 4/25/17 at 6:09am
Kn0wledge
(20 items)
 
Pati3nce
(14 items)
 
Wisd0m
(10 items)
 
Reply
Kn0wledge
(20 items)
 
Pati3nce
(14 items)
 
Wisd0m
(10 items)
 
Reply
post #82 of 108
AMD are involved in a lot more than GPU designs, just like nVidia.

They both are major players in development simply because of their inherent need in cutting edge technologies.
   
AGP bencher
(14 items)
 
CPUMotherboardGraphicsRAM
Ryzen R7 1700 Gigabyte GA-AX370-Gaming 5 Sapphire HD 6950 2GiB 2x8GB KFA2 HOF DDR4-3600 
Hard DriveHard DriveHard DriveHard Drive
Crucial MX100 256GB Seagate 600 Series 240GB Seagate 7200.14 2TB Samsung F3 1TB 
CoolingCoolingCoolingCooling
EKWB Supreme HF XSPC Rasa GPU EK XT360 EK 4.0 
OSMonitorMonitorKeyboard
W10 Pro LG IPS235 LG E2250V KUL ES-87 
PowerCaseMouseAudio
SF Leadex II 650W Lian Li PC-A05NB Logitech G9 Xonar DX 
AudioAudio
SMSL SA-S3+Technics CB-250 Sennheiser HD555 
CPUMotherboardRAMHard Drive
AMD A10-5700 Gigabyte F2A75M-HD2 G.SKILL Ares 2133 CL9 Hitachi 5K750 
Hard DriveCoolingOSMonitor
Momentus .7 200GB Noctua NH-L9a Server 2012 R2 Standard AUO B156HW01 
PowerCaseOther
PicoPSU-80-WI-25V AIO Aluminium Handmade TP-Link Archer Something Something Wi-Fi AC 
CPUCPUCPUMotherboard
Core2Duo E6400 Core2Quad Q6600 Pentium Dual Core E5200 AsRock 4COREDUAL-SATA2 R2.0 
GraphicsRAMHard DriveOptical Drive
A dumpload of ancient AGP cards Kingston Value DDR2-667 CL4 2T @CL3 1T Seagate 160GB 7200.10 LG IDE DVD-ROM 
CoolingCoolingOSMonitor
Ghettomade CPU waterblock 49cc 2stroke engine copper radiator WinXP SP2 32bit ProView 17" 
PowerCase
Tacens Radix V 550W Ghetto aluminium bench 
  hide details  
Reply
   
AGP bencher
(14 items)
 
CPUMotherboardGraphicsRAM
Ryzen R7 1700 Gigabyte GA-AX370-Gaming 5 Sapphire HD 6950 2GiB 2x8GB KFA2 HOF DDR4-3600 
Hard DriveHard DriveHard DriveHard Drive
Crucial MX100 256GB Seagate 600 Series 240GB Seagate 7200.14 2TB Samsung F3 1TB 
CoolingCoolingCoolingCooling
EKWB Supreme HF XSPC Rasa GPU EK XT360 EK 4.0 
OSMonitorMonitorKeyboard
W10 Pro LG IPS235 LG E2250V KUL ES-87 
PowerCaseMouseAudio
SF Leadex II 650W Lian Li PC-A05NB Logitech G9 Xonar DX 
AudioAudio
SMSL SA-S3+Technics CB-250 Sennheiser HD555 
CPUMotherboardRAMHard Drive
AMD A10-5700 Gigabyte F2A75M-HD2 G.SKILL Ares 2133 CL9 Hitachi 5K750 
Hard DriveCoolingOSMonitor
Momentus .7 200GB Noctua NH-L9a Server 2012 R2 Standard AUO B156HW01 
PowerCaseOther
PicoPSU-80-WI-25V AIO Aluminium Handmade TP-Link Archer Something Something Wi-Fi AC 
CPUCPUCPUMotherboard
Core2Duo E6400 Core2Quad Q6600 Pentium Dual Core E5200 AsRock 4COREDUAL-SATA2 R2.0 
GraphicsRAMHard DriveOptical Drive
A dumpload of ancient AGP cards Kingston Value DDR2-667 CL4 2T @CL3 1T Seagate 160GB 7200.10 LG IDE DVD-ROM 
CoolingCoolingOSMonitor
Ghettomade CPU waterblock 49cc 2stroke engine copper radiator WinXP SP2 32bit ProView 17" 
PowerCase
Tacens Radix V 550W Ghetto aluminium bench 
  hide details  
Reply
post #83 of 108
Quote:
Originally Posted by BiG StroOnZ View Post

Not sure what to make of the, "not enough performance advantage," considering Vega still has yet to be launched and from the most recent leaks we have of it, it is only 10% faster than a GTX 1080. We have no other information regarding what succeeds Vega, whereas we know of Volta. Giving me assurance that again, it will be another generation where NVIDIA can do anything they want. We still have yet to see the 1060 and 1080 memory revision released, this will probably put the 1060 ahead of the 580, and of course then the 1080 will be prepared to take on Vega.

The 1060 9Gbps and the 1080 11Gbps are nothing more than a gap-filler. They will not be able to keep Vega off for very long. Nvidia has to bypass the time until Volta finall comes. If Q3 to Q4 2017 is remotely true, that can be another 6-8 months until any successor arrives. To see which card series comes out first is quite exciting. In the past Nvidia started with the x70 and x80 and brought the x60 a few months later, then the big chip like a Titan slightly behind.

The battle against Vega will not be about performance but also price. The 1080 had price cuts recently. Since it is a middle-class sized chip, Nvidia might be able to go down way more. Vega can never win a price-war with expensive HBM2, which is limited to 8GB and this HBCC nonsense. Doesn't sell well on a GPU wrapping.

Quote:
Originally Posted by BiG StroOnZ View Post

GDDR6 is going to be pretty expensive to use (while cheaper than HBM2), and GDDR5X will have dropped in price even though speeds will have increased. I mean we are talking about 2018-2019 here, a two year passing of time for prices to change. The memory has the potential to go up to 16 Gbps. I see it making it to at least 14 Gbps as well, and this allows manufacturers like NVIDIA to use the GDDR6 option for 16 Gbps. The only reason why I used 15 Gbps as an option is because the configuration looks more uniform when comparing the possible future GPU SKUs.

On the manufacturing side Nvidia could use tricks like with the 1080 Ti. Deactivating a cluster for odd or uneven memory configurations should be easier to simulate beforehand. The Geforce cards can get higher bandwidth / throughput without touching the memory speeds themselve. Although I wonder if this 16 Gbps number is fixed or if there are further improvements in the near future we don't know about yet.

Quote:
Originally Posted by geoxile View Post

I think people are mistaking bottlenecks in the front end for a memory bandwidth bottleneck. GCN doesn't need more bandwidth it needs to be revamped. Hopefully Vega will do it. I really doubt memory will have little if any impact on the performance of Vega, especially since it'll supposedly have 2 stacks of HBM2, which means 512GB/s again like Fury X.

Look at the amount of shaders GCN has to feed. There is a reason why the 4096 shaders of Fiji needed HBM. Vega seems to stay at the same amount. All Radeons run into heavy bandwidth limitations with standard GDDR5. The GCN archictecture has some bottlenecks too. But 512GB/s will not be enough for middle to late 2017. The GTX 1080 Ti already has 484 GB/s and Nvidia is known for their better memory management including memory complression.

What Vega needs is pure rawpower. Otherwise it will starve under all possibilities like Fiji did before.

Quote:
Originally Posted by Alwrath View Post

Volta with ddr6 will be like when I opened up my radeon 4870 with it being the first card having ddr5 ram biggrin.gif

Funny, reminded me of the jump from going 512MB to a 1GB HD4850 model. Might have been still GDDR3, but the smaller models were so much slower it didn't matter. Looked it up a few moments ago. Those were 8x 1 Gbps chips. How fast time passes by....

Quote:
Originally Posted by guttheslayer View Post

Gddr6 is still way too early. Pascal almost miss its g5x adoption. I would say volta will be facing same issue.

Did Pascal really miss it? I always remind about the old HMC contracts Nvidia, Samsung and Micron had a few years ago. After it was clar Nvidia would switched to HBM, there is no way this contracts wouldn't have been altered for all parties. There are so many legal claused included, the matter could have been expensive otherise. Micron offered GDDR5X and Samung agreed to deliver HBM2 for the Nvidia DXG-1 .. and maybe a chip here and the with Samsungs 14-nm-LPP (Low Power Plus). Happened in the low-end, but the last part is just speculation from me.

To this point no known issues with Volta are known. The rumoured release is still several quarters away.

Quote:
Originally Posted by guttheslayer View Post

I wouldnt count on G6 this year. If they release volta this year. It will be 12gbps g5x

Again, I refer to the announcements about Micron being ahead of schedule and will be delivering GDDR6 till the end of 2017. Taking SK Hynix in 2018 aside, Nvidia could already start the pre-production a few months earlier to ensure availability at launch. Not impossible Volta will have some GDDR5X configuration again, yet I don't believe it will be the high-end to enthusiast once more.

Getting a node-shrink even if it would be a half step as 12nm practically is, Nvidia wil not be so leaned back to not get what they can get in terms of memory configurations. The change is soo much of a low-hanging fruit.
post #84 of 108
Quote:
Originally Posted by Wishmaker View Post

Funny how all these new things appear when HBM was the tech of the future biggrin.gif. Yet again, AMD has no power here and NVIDIA is pushing for tech where they do not need to pay royalties to AMD.

Someone didn't do their homework and decided to trash AMD for no reason lol

AMD doesn't own HBM nor would Nvidia pay them royalties for it. Nvidia also doesn't own GDDR6, Hynix is pushing things here entirely separate from Nvidia.
Besides the fact that AMD has been the one to adopt new memory on their GPU first for a while, Nvidia hasn't been pushing anything memory wise they always wait on VRam tech so idk what you're getting at.
Toxic-DT
(19 items)
 
Toxic-Mini
(6 items)
 
Toxic-SV
(10 items)
 
CPUMotherboardGraphicsRAM
i7 4770k Gigabyte Z87X-UD4H Sapphire Nitro + RX 470 Gskill Ares 
Hard DriveHard DriveHard DriveOptical Drive
Mushkin Eco3 Kingdian  Corsair LE  Lite On 
CoolingOSMonitorKeyboard
Cryorig H5 Windows 10 pro Dell S2415h Azio MK Retro 
PowerCaseMouseAudio
Seasonic G Series 650w Define R5  Logitech G500s Asus Xonar DSX 
AudioAudioAudio
Monoprice Retro Heaphones Monoprice Vocal Dynamic Mic Edifier r980t  
CPUMotherboardGraphicsRAM
Rk3288 Firefly  Mali t760 mp4 Ddr3 1600 
Hard DriveOS
Internal EMMC Android 
CPUMotherboardGraphicsRAM
Xeon e5649 HP z400 FirePro v4800 12gb Gkill Ripjaw 
Hard DriveHard DriveCoolingOS
Hitachi Ultrastar Western Digital 320gb HP High Performance Cooler Ubuntu Gnome 16.10 
PowerCase
HP z400 HP z400 
  hide details  
Reply
Toxic-DT
(19 items)
 
Toxic-Mini
(6 items)
 
Toxic-SV
(10 items)
 
CPUMotherboardGraphicsRAM
i7 4770k Gigabyte Z87X-UD4H Sapphire Nitro + RX 470 Gskill Ares 
Hard DriveHard DriveHard DriveOptical Drive
Mushkin Eco3 Kingdian  Corsair LE  Lite On 
CoolingOSMonitorKeyboard
Cryorig H5 Windows 10 pro Dell S2415h Azio MK Retro 
PowerCaseMouseAudio
Seasonic G Series 650w Define R5  Logitech G500s Asus Xonar DSX 
AudioAudioAudio
Monoprice Retro Heaphones Monoprice Vocal Dynamic Mic Edifier r980t  
CPUMotherboardGraphicsRAM
Rk3288 Firefly  Mali t760 mp4 Ddr3 1600 
Hard DriveOS
Internal EMMC Android 
CPUMotherboardGraphicsRAM
Xeon e5649 HP z400 FirePro v4800 12gb Gkill Ripjaw 
Hard DriveHard DriveCoolingOS
Hitachi Ultrastar Western Digital 320gb HP High Performance Cooler Ubuntu Gnome 16.10 
PowerCase
HP z400 HP z400 
  hide details  
Reply
post #85 of 108
Quote:
Originally Posted by Hardware Hoshi View Post


Look at the amount of shaders GCN has to feed. There is a reason why the 4096 shaders of Fiji needed HBM. Vega seems to stay at the same amount. All Radeons run into heavy bandwidth limitations with standard GDDR5. The GCN archictecture has some bottlenecks too. But 512GB/s will not be enough for middle to late 2017. The GTX 1080 Ti already has 484 GB/s and Nvidia is known for their better memory management including memory complression.

What Vega needs is pure rawpower. Otherwise it will starve under all possibilities like Fiji did before.

It's known that Fiji had a weak rasterizer and inefficient cache. If memory bandwidth was the real problem we wouldn't be hearing of a 2-stack HBM2 Vega, which would only amount to 512GB/s yet again.
New and Shiny
(18 items)
 
  
CPUMotherboardGraphicsRAM
Rynze 7 1700 ASUS ROG Crosshair VI Hero Sapphire HD7950 G.SKILL TridentZ F4-3200C14D 
Hard DriveHard DriveHard DriveHard Drive
Kingston HyperX 3K Crucial MX300 Western Digital Black Western Digital Green 
Hard DriveCoolingOSMonitor
Western Digital Red Noctua NH-U14S Windows 10 Pro Dell U2414H 
MonitorKeyboardPowerCase
Dell P2414H Ducky One Corsair RM650x NZXT H440 White 
MouseAudio
Logitech G502 Proteus Spectrum Xonar DX 
  hide details  
Reply
New and Shiny
(18 items)
 
  
CPUMotherboardGraphicsRAM
Rynze 7 1700 ASUS ROG Crosshair VI Hero Sapphire HD7950 G.SKILL TridentZ F4-3200C14D 
Hard DriveHard DriveHard DriveHard Drive
Kingston HyperX 3K Crucial MX300 Western Digital Black Western Digital Green 
Hard DriveCoolingOSMonitor
Western Digital Red Noctua NH-U14S Windows 10 Pro Dell U2414H 
MonitorKeyboardPowerCase
Dell P2414H Ducky One Corsair RM650x NZXT H440 White 
MouseAudio
Logitech G502 Proteus Spectrum Xonar DX 
  hide details  
Reply
post #86 of 108
Quote:
Originally Posted by Hardware Hoshi View Post

The 1060 9Gbps and the 1080 11Gbps are nothing more than a gap-filler. They will not be able to keep Vega off for very long. Nvidia has to bypass the time until Volta finall comes. If Q3 to Q4 2017 is remotely true, that can be another 6-8 months until any successor arrives. To see which card series comes out first is quite exciting. In the past Nvidia started with the x70 and x80 and brought the x60 a few months later, then the big chip like a Titan slightly behind.

The battle against Vega will not be about performance but also price. The 1080 had price cuts recently. Since it is a middle-class sized chip, Nvidia might be able to go down way more. Vega can never win a price-war with expensive HBM2, which is limited to 8GB and this HBCC nonsense. Doesn't sell well on a GPU wrapping.

On the manufacturing side Nvidia could use tricks like with the 1080 Ti. Deactivating a cluster for odd or uneven memory configurations should be easier to simulate beforehand. The Geforce cards can get higher bandwidth / throughput without touching the memory speeds themselve. Although I wonder if this 16 Gbps number is fixed or if there are further improvements in the near future we don't know about yet.

This is their reasoning behind releasing the 9Gbps and 11Gbps enhancements to these cards. The RX 580 is already out and ties, slightly wins or slightly loses (TPU has it 7% faster on average) versus the 1060. Finally, the RX 580 is actually a viable option whereas with the 480 most people said, "might as well get a 1060," because the performance difference was substantial enough to dismiss the 480 as an option (and many other variables). Whereas today, with the 580, the card is incredibly competitive in comparison to the 1060. Here though, NVIDIA doesn't want to leave that performance parity open and by simply increasing the memory speed on the revamp; the 1060 can be back to being 5-10% ahead of the 580. While you describe this as a gap-filler, it is still a swift move or play by NVIDIA. Sure it isn't a game changing move, but it is enough to carry them over to the next generation. They already made the sales they needed to make and with even a slight performance advantage, alongside brand recognition, it will allow NVIDIA to continue to make those impulse buying sales from potential customers. With the 1080 11 Gbps, again the same applies (the videos I saw of the comparisons between Vega and the 1080 showed them basically having identical performance, with maybe a slight advantage going to Vega). Most recent leaks have Volta at around Q3 2017 which is around September/October right in time for holiday. Recently, we have seen that release leak estimations and dates have been pretty accurate. That gives them half of a year to use Pascal to compete with the Vega stack. Now here's where things get problematic. If this is the case and we see a 2080 card dropped before 2018 or at least we have all the information available about it (like NVIDIA has been doing lately with their press releases) with availability around Q4 2017 / Q1 2018 that means that now, AMD has a whole slew of problems. This is because the 2080 is more than likely going to be Titan Xp performance for around $600. Now if the Vega card is also $600 because of having HBM2 and only is around 1080 performance, you can see how detrimental this could be; because as you say with Vega it is going to have HBM2, and today, if it only competes with the 1080, it will be difficult for them to get people to go over to the Red Team if the pricing isn't aggressive. AMD still has the issue of being the value option, or price to performance option. People only want to go AMD if it makes sense from a monetary standpoint. Aside from AMD enthusiasts, it is hard for them to convince potential buyers because their cards and products are still stigmatized. So here, while Pascal is still available, this can become an issue and then even worse if Volta is indeed a 2017 product.

There is a possibility that the card AMD demonstrated is not their fasted card, but based on math it seems unlikely:

An RX 580 has 2306 SPs, as far as I'm aware the maximum amount of SPs possible on GCN is 4096. That's a 78% increase in SPs. If we are to estimate that 78% increase in SPs means 78% linear increase in performance; we arrive at 12.7% faster than a GTX 1080 @ 1440p (we saw with the Fury X that HBM didn't really affect performance as much as we thought it would, when compared to the 980 Ti. Therefore, I'm not adding more performance percentage points for HBM 2 for that exact reason):



This unfortunately means, that AMD doesn't have a competitor against the 1080 Ti / Titan Xp which again means they aren't actually competing in the high end segment for yet another generation. Translating to, again, NVIDIA doing anything they freaking want with prices for the rest of Pascal's life and for Volta's lifespan whether that means for better or for worse (undercutting AMD or price gouging).

Now that we see where Vega will most likely end up, even though the 1080 is a mid-sized chip, it still is going to easily compete with full Vega, especially after adding the 11 Gbps memory. Hopefully AMD has some magic sauce that allows them to price Vega @ $550 or something along those lines and retain a 5-8% performance advantage over the 1080 @ $500 where then it could be a pretty hopeful battle. However, again, NVIDIA will not leave that performance delta open for too long, they must always retain the performance crown. Bringing us back to our Volta release estimations in Q3/Q4 2017 as even more probable. As a 2080 that is as fast as a Titan Xp for say $600-650 means that the 2080 will be around 30% faster than the best Vega SKU for $550-600, while only being 10% more expensive. I mean there is a chance that AMD blows us away and can offer the biggest Vega SKU for 1070 prices, but then that leaves no price available for cut down products that are supposed to compete with the 1070 instead of the 1080. I mean the best versions of the RX 580 are already $300. So what would a Vega card that is a 1070 and 1080 competitor cost. Doesn't leave much room for price wars.

They will most likely manipulate the amount of VRAM additionally, as I agree with that and they could utilize more memory compression technology, but of course that isn't as good as simply increasing the memory bandwidth by increasing the speed. Of course there are future improvements with the 16 Gbps; GDDR5 started @ 900MHz and today it is @ 8000-9000MHz. Surely there is advancements to be made with the technology. Which brings us back to discussions we had around 2 years ago, is HBM going to survive or ever going to even be necessary for consumer GPUs? Or is this going to be a case where Radeon Technologies Group consistently uses HBM technology in their GPUs and NVIDIA goes a different route by sticking to modern day GDDR technology; whether GDDR6 or G5X.
Edited by BiG StroOnZ - 4/25/17 at 1:11pm
post #87 of 108
Quote:
Originally Posted by BiG StroOnZ View Post

GDDR5 started @ 900MHz and today it is @ 1800-2200MHz.

I had to correct this because the 900mhz isnt the data rate but the base clock. The 900mhz gddr5 was 3600mhz data rate, now we've double or slightly more the base clock to 1800-2200mhz.
Toxic-DT
(19 items)
 
Toxic-Mini
(6 items)
 
Toxic-SV
(10 items)
 
CPUMotherboardGraphicsRAM
i7 4770k Gigabyte Z87X-UD4H Sapphire Nitro + RX 470 Gskill Ares 
Hard DriveHard DriveHard DriveOptical Drive
Mushkin Eco3 Kingdian  Corsair LE  Lite On 
CoolingOSMonitorKeyboard
Cryorig H5 Windows 10 pro Dell S2415h Azio MK Retro 
PowerCaseMouseAudio
Seasonic G Series 650w Define R5  Logitech G500s Asus Xonar DSX 
AudioAudioAudio
Monoprice Retro Heaphones Monoprice Vocal Dynamic Mic Edifier r980t  
CPUMotherboardGraphicsRAM
Rk3288 Firefly  Mali t760 mp4 Ddr3 1600 
Hard DriveOS
Internal EMMC Android 
CPUMotherboardGraphicsRAM
Xeon e5649 HP z400 FirePro v4800 12gb Gkill Ripjaw 
Hard DriveHard DriveCoolingOS
Hitachi Ultrastar Western Digital 320gb HP High Performance Cooler Ubuntu Gnome 16.10 
PowerCase
HP z400 HP z400 
  hide details  
Reply
Toxic-DT
(19 items)
 
Toxic-Mini
(6 items)
 
Toxic-SV
(10 items)
 
CPUMotherboardGraphicsRAM
i7 4770k Gigabyte Z87X-UD4H Sapphire Nitro + RX 470 Gskill Ares 
Hard DriveHard DriveHard DriveOptical Drive
Mushkin Eco3 Kingdian  Corsair LE  Lite On 
CoolingOSMonitorKeyboard
Cryorig H5 Windows 10 pro Dell S2415h Azio MK Retro 
PowerCaseMouseAudio
Seasonic G Series 650w Define R5  Logitech G500s Asus Xonar DSX 
AudioAudioAudio
Monoprice Retro Heaphones Monoprice Vocal Dynamic Mic Edifier r980t  
CPUMotherboardGraphicsRAM
Rk3288 Firefly  Mali t760 mp4 Ddr3 1600 
Hard DriveOS
Internal EMMC Android 
CPUMotherboardGraphicsRAM
Xeon e5649 HP z400 FirePro v4800 12gb Gkill Ripjaw 
Hard DriveHard DriveCoolingOS
Hitachi Ultrastar Western Digital 320gb HP High Performance Cooler Ubuntu Gnome 16.10 
PowerCase
HP z400 HP z400 
  hide details  
Reply
post #88 of 108
Quote:
Originally Posted by Echoa View Post

Besides the fact that AMD has been the one to adopt new memory on their GPU first for a while, Nvidia hasn't been pushing anything memory wise they always wait on VRam tech so idk what you're getting at.

In the past AMD or rather Ex-ATI had alot of involvement in the first memory adoption, that's true. Seems to be no longer the case. Because the AMD management's pipedream of "Fusion", the APU / CPU side needs faster internal memory. Might explain why AMD is so fixated on HBM. Their choices have not been economical in the recent few years. GDDR5X was the first memory type they haven't used anywhere first. GDDR6 might be the next one. Losing the initiative decreased the chance to react. With such tight ressources AMD has at the moment, this could turn out critical in the long run.

Quote:
Originally Posted by geoxile View Post

It's known that Fiji had a weak rasterizer and inefficient cache. If memory bandwidth was the real problem we wouldn't be hearing of a 2-stack HBM2 Vega, which would only amount to 512GB/s yet again.

Fiji was a total disaster on the engineering level. So many things did not work out and nothing worked together expect in the idial scenario. The bandwidth was not the issue with HBM itself. The lower total HBM capacity of 4GB instead was a real dealbreaker. Since stories pop up that say Vega wll be HBM2 in 4GB or 8GB configurations only, I am not very positive about the future here. Sounds way to much of a repeat of Fiji.

Quote:
Originally Posted by BiG StroOnZ View Post

There is a possibility that the card AMD demonstrated is not their fasted card, but based on math it seems unlikely

Before we go into details, let's be said that AMD has the habit of presenting their hardware in the best light. They showed doom roughly at 1080 level or slightly higher. That demo is already the best case scenario for them. You can compare the type of results with similar announcements back with Fiji and Polaris. Alot of that stuff is pure marketing and actually not worth a single second of attantion. We will have to wait for final reviews of non-biased sources.

Quote:
Originally Posted by BiG StroOnZ View Post

An RX 580 has 2306 SPs, as far as I'm aware the maximum amount of SPs possible on GCN is 4096. That's a 78% increase in SPs. If we are to estimate that 78% increase in SPs means 78% linear increase in performance; we arrive at 12.7% faster than a GTX 1080 @ 1440p (we saw with the Fury X that HBM didn't really affect performance as much as we thought it would, when compared to the 980 Ti. Therefore, I'm not adding more performance percentage points for HBM 2 for that exact reason)

Too bad that shaders don't scale linear. GCN has problems to utilized them to full load. One of the reasons why AMD needs to use tricks like "Asynchornous Compute Shaders" to fire up anything with workload. If there are other techniques like the HBCC for the memory comes into play, it just shows the difficulty of utilizing the full potential of the theoretical rawpower.

To b feared is the amount of optimizations the game developers have to put in again. If said features have to be supported full time by dedicated programmers, alot of unused potential could lie waste in general. Only the AMD house and garden studios will be able to fully bring out Vega's power. Therefor I see ithe results really dependent on the scenario.

We both can do number games all day long, but never reach a complusive outcome. Vega is and stays a mystery. Yet when I ever learned something about AMD in the last few years, all they could accomplish at best was parity with the competition.

Quote:
Originally Posted by BiG StroOnZ View Post

This unfortunately means, that AMD doesn't have a competitor against the 1080 Ti / Titan Xp which again means they aren't actually competing in the high end segment for yet another generation. Translating to, again, NVIDIA doing anything they freaking want with prices for the rest of Pascal's life and for Volta's lifespan whether that means for better or for worse (undercutting AMD or price gouging).

From everything we know so far, the enthusiast segment is out of AMD's reach. Although it doesn't have to be bad per se. There is a small possiblity that Vega can at least swirl up everything at 1080 level or lower. For the prices we have to see how the market reacts. I am more worried about what happends when Volta really arrives with GDDR6. The time between Volta and Navi release could bring another round of absurd prices, if AMD is not fully competitive.
post #89 of 108
Quote:
Originally Posted by Echoa View Post

I had to correct this because the 900mhz isnt the data rate but the base clock. The 900mhz gddr5 was 3600mhz data rate, now we've double or slightly more the base clock to 1800-2200mhz.

Yeah, that was my error, I was using Newegg as my source and they label it "effective memory clock" without realizing Gbps was right next to it (not wearing my glasses tongue.gif):





Regardless, that's still a 122% increase in memory speed from the start. thumb.gif
Edited by BiG StroOnZ - 4/25/17 at 2:27pm
post #90 of 108
Quote:
Originally Posted by Shogon View Post

I've heard countless times Windows is a walled garden of an OS, has questionable tactics of forcing people to Windows 10 (especially if you care about dx12, or new hardware with your older OS) and yet we still use it.

Seeing a monopoly at all is a bad thing in any market; you should know that even if it conflicts emotionally with our favorite corporations.

But hey, this is ocn at the end of the day.

I can't really say Microsoft has done anything horribly offensive there. New versions of their API only supported with new versions of their operating system? Gee, what a surprise! Not supporting new hardware on old operating systems (which, incidentally, the hardware manufacturers aren't doing either)? Shocking! They deserve to be criticized but not for these reasons, at least not to any significant extent.

You misinterpreted what I said. I don't want a monopoly. However, if we must have one, I can tell you now based on what Nvidia is doing with only a decent lead that the market would quickly turn to crap. Contrast with, for example, Intel, who practically has had a monopoly in the professional, server, and high-end markets for years yet has kept prices at a given tier consistent each generation. Nvidia meanwhile is charging more for their Gx104-tier GPUs today than they did for their Gx100-tier GPUs a few years ago. That's completely ridiculous.

It's pretty rude to insinuate I'm a fanboy. But hey, this is ocn at the end of the day. wink.gif
Edited by CynicalUnicorn - 4/25/17 at 2:09pm
Triumvirate
(20 items)
 
Osmium
(8 items)
 
EleMINT 24
(8 items)
 
CPUMotherboardGraphicsGraphics
Xeon E3-1226 v3 ASRock Z97M OC Formula Sapphire RX 480 (reference) MSI Low-Profile 750Ti 
RAMRAMHard DriveHard Drive
Kingston Blu Red - 2x4GB ADATA XPG - 2x4GB Crucial M500 960GB Samsung 840 250GB 
Hard DriveCoolingCoolingCooling
WD Scorpio Blue 1TB Cooler Master 212+ PNY Quadro 600 blower Corsair H55 + Kraken G10 
OSMonitorMonitorKeyboard
Windows 10 Pro ASUS VE247H - 1080p Gateway FPD1960 - 1280x1024 Rosewill RK9000I - Cherry MX Blue 
PowerCaseMouseOther
Rosewill HIVE 650W Fractal Design Arc Mini R2 Razer DeathAdder 2013 CyberPower 1500PFCLCD 
CPUMotherboardRAMHard Drive
AMD A10-7870K ASRock FM2A88X-ITX+ G.Skill Ripjaws X - 2x4GB @ 2400MT/s Samsung 840 120GB 
CoolingOSPowerCase
Evercool EC-HPS-810CP Windows 7 Pro 64-bit PicoPSU-120-WI-25 Mini-Box M350 
CPUMotherboardGraphicsRAM
Intel i3-4005U OEM Intel HD 4200 4GB DDR3 (soldered) 
Hard DriveOSMonitorPower
MyDigitalSSD Super Boot Eco 120GB Linux Mint 17.1 Cinnamon 1366x768 TN panel :'( brick 
  hide details  
Reply
Triumvirate
(20 items)
 
Osmium
(8 items)
 
EleMINT 24
(8 items)
 
CPUMotherboardGraphicsGraphics
Xeon E3-1226 v3 ASRock Z97M OC Formula Sapphire RX 480 (reference) MSI Low-Profile 750Ti 
RAMRAMHard DriveHard Drive
Kingston Blu Red - 2x4GB ADATA XPG - 2x4GB Crucial M500 960GB Samsung 840 250GB 
Hard DriveCoolingCoolingCooling
WD Scorpio Blue 1TB Cooler Master 212+ PNY Quadro 600 blower Corsair H55 + Kraken G10 
OSMonitorMonitorKeyboard
Windows 10 Pro ASUS VE247H - 1080p Gateway FPD1960 - 1280x1024 Rosewill RK9000I - Cherry MX Blue 
PowerCaseMouseOther
Rosewill HIVE 650W Fractal Design Arc Mini R2 Razer DeathAdder 2013 CyberPower 1500PFCLCD 
CPUMotherboardRAMHard Drive
AMD A10-7870K ASRock FM2A88X-ITX+ G.Skill Ripjaws X - 2x4GB @ 2400MT/s Samsung 840 120GB 
CoolingOSPowerCase
Evercool EC-HPS-810CP Windows 7 Pro 64-bit PicoPSU-120-WI-25 Mini-Box M350 
CPUMotherboardGraphicsRAM
Intel i3-4005U OEM Intel HD 4200 4GB DDR3 (soldered) 
Hard DriveOSMonitorPower
MyDigitalSSD Super Boot Eco 120GB Linux Mint 17.1 Cinnamon 1366x768 TN panel :'( brick 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Hardware News
Overclock.net › Forums › Industry News › Hardware News › [Guru3D] SK Hynix Launches Fastest 8Gb Graphics DRAM (GDDR6) - Adopted by 2018