• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Possible Radeon 390X / 390 and 380X Spec / Benchmark (do not hotlink images!!!!!!)

Status
Not open for further replies.
4GB HBM does not = 4GB GDDR5

It has up to 4 times the bandwidth, is up to 50% more efficient for power and more importantly increases the amount of power available for the GPU.

As bandwidth increases, frame buffer sizes can reduce and you will still get the same performance. You need only look back 7 years to when GDDR3 and GDDR5 were about, the same scenario happened, GDDR3 2GB was significantly slower than GDDR5 1GB with IDENTICAL GPU's and clock speeds.
 
Last edited:
Of course there is also talk of DX12 allowing memory sharing, so VRAM on multiple cards will stack, which could help.
I mean chances are you'll need multiple cards to run at the settings that would make VRAM an issue anyway.

Whether this feature is something you just get or something developers can implement if the mind to could determine if this will come into play at all.

Until then, I guess you may just have to turn settings down on your brand new cards...
 
MadMossy said:
4GB HBM does not = 4GB GDDR5

It has up to 4 times the bandwidth, is up to 50% more efficient for power and more importantly increases the amount of power available for the GPU.
Click to expand...

Yes but 4GB of textures is still 4GB of textures.

Or put it this way, a pint of water is a pint of water, whether it is in a bottle or a bucket, or even a super high pressure hose system, you may be able to empty some Quicker than others but if all three only hold 1 pint then you cannot squeeze more than that into it. all three will still be filled by the same tap as well, in this case the HDD/system ram.
 
bru said:
Yes but 4GB of textures is still 4GB of textures.

Or put it this way, a pint of water is a pint of water, whether it is in a bottle or a bucket, or even a super high pressure hose system, you may be able to empty some Quicker than others but if all three only hold 1 pint then you cannot squeeze more than that into it. all three will still be filled by the same tap as well, in this case the HDD/system ram.
Click to expand...

Trains A and B are traveling in the same direction on parallel tracks. Train A is traveling at 60 mph and train B is traveling at 70 mph. Train A passesa station at 12:20 P.M. If train B passes the same station at 12:32 P.M., at what time will train B catch up to train A?"

:D
 
JediFragger said:
Trains A and B are traveling in the same direction on parallel tracks. Train A is traveling at 60 mph and train B is traveling at 70 mph. Train A passesa station at 12:20 P.M. If train B passes the same station at 12:32 P.M., at what time will train B catch up to train A?"

:D
Click to expand...

1:44pm, but they still both have 4 carriages and can only carry the same number of passengers. ;)
 
africanos23 said:
I know how it depends on your total amount of VRAM and the "extra" frame buffer if you have it available (I have used a vast array of GPUs) but let's not forget most people wanting to drive 1440p or 4K with good FPS (60FPS) are looking at this card or (the small number) of 1080p user that are looking for 120FPS+ or/and futureproofing for newer games.

Adding a second card is going to allow you to have that FPS (60+) on the highest settings with MSAA so why should you need to drop something due to a VRAM bottleneck ? (most SLI/CFX rigs are the medium to high end tier not the low end)

When you shell out cash on a new toy ($XXX if not more) on the highest end I doubt you want to turn down settings.

3GB is borderline up until 1440p (for newer titles) you are going to have to lower quite a few details.

4GB is ok for 1440p (for most games) some you are going to have to sacrifice a detail or two.

4GB is ok in some scenarios for 4K but don't be surprised if you hit that limit when going over medium-high settings (even without AA) in AAA titles.

I also want to add something, the difference moving from SLI GTX970 and CFX R9-290s 4GB to 12GB VRAM TX (single and SLI) has a lot to do with the frame buffer.

More = less swaps = smoother gaming experience
Click to expand...

Why did you bother writing any of this? 1) If you're going CF, all DX12 / Vulkan games should pool memory. Therefore 2x Fury = 8GB. 2) HBM can swap larger files and more of them than GDDR5, and at lower latency, and DX12 / Vulkan can do the same, plus tiled resources. When the option is to go 4GB HBM or GDDR5 (and much slower GPU and less bandwidth), the answer is pretty damned obvious ... plus they make their transition to interposer, new memory, and new memory controller on a familiar and reliable process - they don't have THAT much to do for Arctic Islands ... NVIDIA have their work cut out with everything new on an unfamiliar process, and coming much later to the stacked memory (and certainly HBM) game.

Add to this GCN is far better at parallelisation than Maxwell or Kepler, and Vulkan & DX12 are all about parallelisation, and HBM and increased bandwidth suit parallelisation far more than a larger frame buffer ...... it's a total no brainer.
 
Last edited:
pmc25 said:
Why did you bother writing any of this? 1) If you're going CF, all DX12 / Vulkan games should pool memory. Therefore 2x Fury = 8GB. 2) HBM can swap larger files and more of them than GDDR5, and at lower latency, and DX12 / Vulkan can do the same, plus tiled resources. When the option is to go 4GB HBM or GDDR5 (and much slower GPU and less bandwidth), the answer is pretty damned obvious.
Click to expand...

First of dx12/vulcan, Can pool memory but that doesnt mean that all if any games will do so. IIRC was also a feature of mantle yet wasnt used.
Second point it wont affect older games of which most of us have so pooling is imo irrelevant atm as its a possible feature that hasnt been used and wont work for non dx12/vulcan stuff
Hopefully HBM will help with the Vram issues but atm we dont know
 
pmc25 said:
Why did you bother writing any of this? 1) If you're going CF, all DX12 / Vulkan games should pool memory. Therefore 2x Fury = 8GB. 2) HBM can swap larger files and more of them than GDDR5, and at lower latency, and DX12 / Vulkan can do the same, plus tiled resources. When the option is to go 4GB HBM or GDDR5 (and much slower GPU and less bandwidth), the answer is pretty damned obvious ... plus they make their transition to interposer, new memory, and new memory controller on a familiar and reliable process - they don't have THAT much to do for Arctic Islands ... NVIDIA have their work cut out with everything new on an unfamiliar process, and coming much later to the stacked memory (and certainly HBM) game.
Click to expand...

There won't be any pooling of memory with DX12 or anything else for a number of years to come as PC hardware is nowhere near up to the speed required across the PCI-E slots.

Even AMD are hinting that this is true as the 390X comes with 8gb which would be unneeded with pooled memory.
 
triss said:
First of dx12/vulcan, Can pool memory but that doesnt mean that all if any games will do so. IIRC was also a feature of mantle yet wasnt used.
Second point it wont affect older games of which most of us have so pooling is imo irrelevant atm as its a possible feature that hasnt been used and wont work for non dx12/vulcan stuff
Hopefully HBM will help with the Vram issues but atm we dont know
Click to expand...

It wasn't implemented into the Mantle drivers.

Oh, it won't effect older games? You mean the ones that never need that much memory anyway? Oh, right.

Yes we do know.
 
Status
Not open for further replies.
Back
Top Bottom