Yeah, it would negate power savings. I was just wondering out loud if it was possible, maybe as a stopgap. It's not happening, but it would be interesting to see.Would that not negate the main use of HBM as a power saver?
Also not a single person has commented on this by Ryan Smith.
http://www.anandtech.com/print/9266/amd-hbm-deep-dive
" HBM in turn allows from 2 to 8 stacks to be used, with each stack carrying 1GB of DRAM. AMDs example diagrams so far (along with NVIDIAs Pascal test vehicle) have all been drawn with 4 stacks,"
Or it could be real. Sweclockers is not just a random news site.
There is no way 8 stacks is coming for Fiji XT.
We know its 4096bit. 8 stacks would give 8192bit.
I think they are referring to the technical limits of HBM in general. 8 stacks is what the dual Fiji is suppose to have. 4 stacks per GPU. Still 4GB/4096bit per GPU but 8 stacks in total.
We have known it is 4096bit/4GB for a long time.Again, does R9 290X with 8GB GDDR5 have 1024 bit memory bus?
Also, where you have have the official Fiji specs from?
Cloudfire, that looks like a 280X Tri-X...
http://www.sapphiretech.com/presentation/product/?cid=1&gid=3&sgid=1227&pid=2456&psn=&lid=1&leg=0
I think the middle symbol is photoshopped onto the card. It is too dark in the angled picture, while the other two have a lighter-colored black because of the reflectance of the light.
I think the actual pictures used were from the 280X Toxic, which has the closest fan sticker at an angle.
Yeah, it would negate power savings. I was just wondering out loud if it was possible, maybe as a stopgap. It's not happening, but it would be interesting to see.
I didn't think 8-high stacks were possible on HBM1.
http://techreport.com/review/28294/amd-high-bandwidth-memory-explained/2This first-gen HBM stack will impose at least one limitation of note: its total capacity will only be 4GB. At first blush, that sounds like a limited capacity for a high-end video card.
Now that GPU makers are selling high-end cards on the strength of their performance at 4K resolutions, one might expect more capacity from a brand-new flagship graphics card.
When I asked Macri about this issue, he expressed confidence in AMD's ability to work around this capacity constraint. In fact, he said that current GPUs aren't terribly efficient with their memory capacity simply because GDDR5's architecture required ever-larger memory capacities in order to extract more bandwidth. As a result, AMD "never bothered to put a single engineer on using frame buffer memory better," because memory capacities kept growing. Essentially, that capacity was free, while engineers were not. Macri classified the utilization of memory capacity in current Radeon operation as "exceedingly poor" and said the "amount of data that gets touched sitting in there is embarrassing."
Strong words, indeed.
With HBM, he said, "we threw a couple of engineers at that problem," which will be addressed solely via the operating system and Radeon driver software. "We're not asking anybody to change their games."
We have known it is 4096bit/4GB for a long time.
http://www.sisoftware.eu/rank2011d/...dbe6deedd4e7d3f587ba8aacc9ac91a187f4c9f9&l=en
Yup, I said that in my earlier postIsn't it more accurate to say that one 4096 shader card has 4GB. There is no way to infer that 8 GB does not exist.
Unknown unknowns do exist.
Isn't it more accurate to say that one 4096 shader card has 4GB. There is no way to infer that 8 GB does not exist.
Unknown unknowns do exist.
AMD might have a plan for 4GB Fiji after all. And a confirmation about AMD releasing a 4GB card indeed
http://techreport.com/review/28294/amd-high-bandwidth-memory-explained/2This first-gen HBM stack will impose at least one limitation of note: its total capacity will only be 4GB. At first blush, that sounds like a limited capacity for a high-end video card.
Now that GPU makers are selling high-end cards on the strength of their performance at 4K resolutions, one might expect more capacity from a brand-new flagship graphics card.
When I asked Macri about this issue, he expressed confidence in AMD's ability to work around this capacity constraint. In fact, he said that current GPUs aren't terribly efficient with their memory capacity simply because GDDR5's architecture required ever-larger memory capacities in order to extract more bandwidth. As a result, AMD "never bothered to put a single engineer on using frame buffer memory better," because memory capacities kept growing. Essentially, that capacity was free, while engineers were not. Macri classified the utilization of memory capacity in current Radeon operation as "exceedingly poor" and said the "amount of data that gets touched sitting in there is embarrassing."
Strong words, indeed.
With HBM, he said, "we threw a couple of engineers at that problem," which will be addressed solely via the operating system and Radeon driver software. "We're not asking anybody to change their games."
This makes as much sense as anything else, but it probably won't help much from the marketing angle. Sad to think that even if the performance is there and the memory is being used efficiently, some idiots won't buy because of what will then be an arbitrary distinction- 4 vs 8.
Try reading my quote again. AMD`s engineer explains it:Why do you call people with valid concerns idiots? What is wrong with you?
I would hardly call this arbitrary. True, there may be few games that benefit from more than 4 GB today, but they will only grow in number. Add to that CF/SLI and 4K and one can certainly understand why someone would not want a 4 GB enthusiast card. Especially not one for $849 (if that price holds true).
Sapphire Black Diamond 280-X:
http://www.aliexpress.com/item-img/...ard-with-LED-light-3-fans-3G/1889604944.html#
Try reading my quote again. AMD`s engineer explains it:
If it was 4GB GDDR5 vs 8GB GDDR5 it would have been valid concern. But since its 4GB HBM with high bandwidth that doesnt need to cache the data to reach similar bandwidth as GDDR5, they can overcome the capacity problem with AMD`s own drivers.
He says the current memory management is very poor with current cards and that only a very small portion of the cached data in the VRAM is actually used.
I`m sure the Fiji reviews will cover this fully when the card is finally launched. Its good to hear AMD have a solution and a better way to use data with memory.
Im guessing a part of it is that they found a way to determine what is actively read from the cached data and what needs to be flushed more regularly?I did and I don't trust AMD to fix this, at least not at launch. And what does that even mean "doesn't need to cache the data"? Of course data needs to be cached, otherwise it will be loaded from slower memory pools (RAM, SSD/HDD) via PCIe.
Sadly I see many apologists now that AMD has less memory than expected/wanted. I did not buy NVs 1.5 GB GTX 580, 3 GB 780 (Ti) or 6 GB 980 Ti. And I'm glad I didn't.
I can see people being concerned about it only being 4GB - but what if benchmarks up to 4K still show Fiji ahead. Having more leeway with more VRAM is nice, but if it ends up being faster in every scenario than the competition, it's hard to justify going with something else just because it has 4GB. It's kind of funny that people are just fine buying 980's with 4GB right now but suddenly the sky is falling if something faster launches with only 4GB.
It's a persistent idea that somehow the speed of HBM means you can get away with less memory.
I'm not sure that's going to work.
But I'm willing to learn...
Or it's just worthless scuttlebutt...
http://arstechnica.com/information-...hbm-why-amds-high-bandwidth-memory-matters/2/"You're not limited in this world to any number of stacks, but from a capacity point of view, this generation-one HBM, each DRAM is a two-gigabit DRAM, so yeah, if you have four stacks you're limited to four gigabytes. You could build things with more stacks, you could build things with less stacks. Capacity of the frame buffer is just one of our concerns. There are many things you can do to utilise that capacity better. So if you have four stacks you're limited to four [gigabytes], but we don't really view that as a performance limitation from an AMD perspective."
"If you actually look at frame buffers and how efficient they are and how efficient the drivers are at managing capacities across the resolutions, you'll find that there's a lot that can be done. We do not see 4GB as a limitation that would cause performance bottlenecks. We just need to do a better job managing the capacities. We were getting free capacity, because with [GDDR5] in order to get more bandwidth we needed to make the memory system wider, so the capacities were increasing. As engineers, we always focus on where the bottleneck is. If you're getting capacity, you don't put as much effort into better utilising that capacity. 4GB is more than sufficient. We've had to go do a little bit of investment in order to better utilise the frame buffer, but we're not really seeing a frame buffer capacity [problem]. You'll be blown away by how much [capacity] is wasted.