They are looking more and more likely to be the real deal now.are these the actual specs? There's alot of mixed reactions going around.
ReadingRainbow4
This topic is locked from further discussion.
- Internal mass storage
- One SKU at launch: 500 GB HDD
- There may also be a Flash drive SKU in the future
Blazerdt47
Don't care about the rest of the rumor poo but I really want this to be true. I think Nintendo made the best decision with having one SKU last generation for the longest time, and I really hope Microsoft and Sony can follow suit with that same line of thinking. I hate this differentiation between a "basic" and "premium," or like last generation... arcade and pro.
[QUOTE="ReadingRainbow4"]They are looking more and more likely to be the real deal now.are these the actual specs? There's alot of mixed reactions going around.
clyde46
Yeah, reading through the neogaf thead and the Durango specs seem to be along the same lines although weaker.
Kinect Box is looking more and more likely.
[QUOTE="ReadingRainbow4"]They are looking more and more likely to be the real deal now. Long time no post.are these the actual specs? There's alot of mixed reactions going around.
clyde46
They are looking more and more likely to be the real deal now.[QUOTE="clyde46"][QUOTE="ReadingRainbow4"]
are these the actual specs? There's alot of mixed reactions going around.
ReadingRainbow4
Yeah, reading through the neogaf thead and the Durango specs seem to be along the same lines although weaker.
Kinect Box is looking more and more likely.
Maybe, I personally think MS would lose tons of market share they just went after Nintendo's pot of gold. Casuals don't really buy all that many games.[QUOTE="RR360DD"]This thread sucks until tormento turns up *waits*Tessellationhe was celebrating the 192GB/s of bandwidth and now that they changed it to 176GB/s is around durango's territory,of course he is going to deny these rumors because they don't make his future plastic girlfriend any better :cool: Actually i wasn't celebrating but if you most know the PS4 one still quite faster,the 720 on is 68GB/s + 102GB/s but that doesn't mean they should be accounted like 170GB/s,go search on Beyond or Neogaf you can't combine those numbers,is 68 from one point to another and 102 from one point to another.. Unlike the PS4 176GB/s one.
These numbers don't add up for me. 1.6 billion triangles per second seems low while xenon did 500 million per sec. Sony claimed the ps3 could do 2.0 teraflops and this is 1.8? Would it really be that expensive to overclock that CPU to atleast 2.0ghz? Audio processor? Doesn't the CPU usually handle that. Where is ronvalencia when u need him.GotNugzThe RSX did 275 almost half of the 360 ones and we know how that story end,theory numbers means sh**.
[QUOTE="Tessellation"][QUOTE="RR360DD"]This thread sucks until tormento turns up *waits*tormentoshe was celebrating the 192GB/s of bandwidth and now that they changed it to 176GB/s is around durango's territory,of course he is going to deny these rumors because they don't make his future plastic girlfriend any better :cool: Actually i wasn't celebrating but if you most know the PS4 one still quite faster,the 720 on is 68GB/s + 102GB/s but that doesn't mean they should be accounted like 170GB/s,go search on Beyond or Neogaf you can't combine those numbers,is 68 from one point to another and 102 from one point to another.. Unlike the PS4 176GB/s one. we don't know anything about either console,these are just rumors for all we know these two speeds can work together,maybe not? or MS worked it around to fix the bandwidth,remember how they increased the xbox 360 RAM quite late,we don't know anything to base things of just speculations.
I havent been around SW the past few days and honestly, I think its because of the rumors... Im excited like everyone else, but it just feels so useless to specualte considering it means nothing till the company unveils it. We're all stuck in the weird haze between generations.
[QUOTE="Tessellation"][QUOTE="RR360DD"]This thread sucks until tormento turns up *waits*tormentoshe was celebrating the 192GB/s of bandwidth and now that they changed it to 176GB/s is around durango's territory,of course he is going to deny these rumors because they don't make his future plastic girlfriend any better :cool: Actually i wasn't celebrating but if you most know the PS4 one still quite faster,the 720 on is 68GB/s + 102GB/s but that doesn't mean they should be accounted like 170GB/s,go search on Beyond or Neogaf you can't combine those numbers,is 68 from one point to another and 102 from one point to another.. Unlike the PS4 176GB/s one.
You know, the DDRx DRAM has already been outclassed by a new revolutionary DRAM which called "Hybrid Memory Cube" It is a multi-chip module which works in parallel with the multi-core processors. The advantages are enormous over the DDRx. It costs less, 15x more performance, consumes less power and takes less space than a typical DDR3. If Microsoft is going to use this new HMC in its new Xbox (I don't think why they shouldn't), it will be a game changer since Microsoft is currently 1 of the 10 members of the HMC consortium. Just wait for the final specs after E3...;)
"Hybrid Memory Cube is a revolutionary innovation in DRAM memory architecture that sets a new standard for memory performance, power consumption and cost."
- HMC Combines high-speed logic process technology with a stack of through-silicon-via (TSV) bonded memory die.
- HMC delivers dramatic improvements in performance, breaking through the memory wall and enabling dramatic performance and bandwidth improvements -a single HMC can provide more than 15x the performance of a DDR3 module.
- The revolutionary architecture of HMC is exponentially more efficient than current memory, utilizing 70% less energy per bit than DDR3 DRAM technologies..
- Hybrid Memory Cube's increased density per bit and reduced form factor contribute to lower total cost of ownership, by allowing more memory into each machine and using nearly 90% less space than today's RDIMMs.
- With performance levels that break through the memory wall, Hybrid Memory Cube represents the key to extending network system performance to push through the challenges of new 100G and 400G infrastructure growth. Eventually, HMC will drive exascale CPU system performance growth for next generation HPC systems.
- Whereas DDR4 represents an evolutionary standard, HMC is a revolutionary technology that is a complete paradigm shift from current memory architectures.
- Hybrid Memory Cube will redefine memory. By advancing past the traditional DRAM system, HMC is setting a new standard of memory that can keep up with the advancements of CPUs and GPUs.
- Hybrid Memory Cube could be an absolute game changer for applications ranging from high performance computing to consumer technologies like tablets and graphics cards that value a combination of form factor, energy and bandwidth.
http://www.hybridmemorycube.org/technology.html
"Micron Readies Hybrid Memory Cube for Debut"
Hybrid Memory Cube (HMC) technology, a multi-chip module (MCM) that aims to address one of the biggest challenges in high performance computing: scaling the memory wall.
Memory architectures haven't kept pace with the bandwidth requirements of multicore processors. As microprocessor speeds out-accelerated DRAM memory speeds, a bottleneck developed that is referred to as the memory wall. Stacked memory applications, however, enable higher memory bandwidth.
The Hybrid Memory Cube (HMC) is a new memory architecture that combines a high-speed logic layer with a stack of through-silicon-via (TSV) bonded memory die that enables impressive advantages over current technology. According to company figures, a single HMC offers a 15x performance increase and uses 70 percent less energy per bit when compared to DDR3 memory, and takes up 90 percent less space than today's RDIMMs. The Cube is also scalable per application, which is not possible with DDR3 and DDR4. System designers have the option of employing the HMC as near memory for best performance or in a scalable module form factor, as far memory, for optimum power efficiency. This is a huge leap forward from a technology perspective, noted Graham, compared to DDRx and other boutique memory products that are out there.
http://www.hpcwire.com/hpcwire/2013-...for_debut.html
Actually i wasn't celebrating but if you most know the PS4 one still quite faster,the 720 on is 68GB/s + 102GB/s but that doesn't mean they should be accounted like 170GB/s,go search on Beyond or Neogaf you can't combine those numbers,is 68 from one point to another and 102 from one point to another.. Unlike the PS4 176GB/s one.[QUOTE="tormentos"][QUOTE="Tessellation"] he was celebrating the 192GB/s of bandwidth and now that they changed it to 176GB/s is around durango's territory,of course he is going to deny these rumors because they don't make his future plastic girlfriend any better :cool:TheXFiles88
You know, the DDRx DRAM has already been outclassed by a new revolutionary DRAM which called "Hybrid Memory Cube" It is a multi-chip module which works in parallel with the multi-core processors. The advantages are enormous over the DDRx. It costs less, 15x more performance, consumes less power and takes less space than a typical DDR3. If Microsoft is going to use this new HMC in its new Xbox (I don't think why they shouldn't), it will be a game changer since Microsoft is currently 1 of the 10 members of the HMC consortium. Just wait for the final specs after E3...;)
"Hybrid Memory Cube is a revolutionary innovation in DRAM memory architecture that sets a new standard for memory performance, power consumption and cost."
- HMC Combines high-speed logic process technology with a stack of through-silicon-via (TSV) bonded memory die.
- HMC delivers dramatic improvements in performance, breaking through the memory wall and enabling dramatic performance and bandwidth improvements -a single HMC can provide more than 15x the performance of a DDR3 module.
- The revolutionary architecture of HMC is exponentially more efficient than current memory, utilizing 70% less energy per bit than DDR3 DRAM technologies..
- Hybrid Memory Cube's increased density per bit and reduced form factor contribute to lower total cost of ownership, by allowing more memory into each machine and using nearly 90% less space than today's RDIMMs.
- With performance levels that break through the memory wall, Hybrid Memory Cube represents the key to extending network system performance to push through the challenges of new 100G and 400G infrastructure growth. Eventually, HMC will drive exascale CPU system performance growth for next generation HPC systems.
- Whereas DDR4 represents an evolutionary standard, HMC is a revolutionary technology that is a complete paradigm shift from current memory architectures.
- Hybrid Memory Cube will redefine memory. By advancing past the traditional DRAM system, HMC is setting a new standard of memory that can keep up with the advancements of CPUs and GPUs.
- Hybrid Memory Cube could be an absolute game changer for applications ranging from high performance computing to consumer technologies like tablets and graphics cards that value a combination of form factor, energy and bandwidth.
http://www.hybridmemorycube.org/technology.html
"Micron Readies Hybrid Memory Cube for Debut"
Hybrid Memory Cube (HMC) technology, a multi-chip module (MCM) that aims to address one of the biggest challenges in high performance computing: scaling the memory wall.
Memory architectures haven't kept pace with the bandwidth requirements of multicore processors. As microprocessor speeds out-accelerated DRAM memory speeds, a bottleneck developed that is referred to as the memory wall. Stacked memory applications, however, enable higher memory bandwidth.
The Hybrid Memory Cube (HMC) is a new memory architecture that combines a high-speed logic layer with a stack of through-silicon-via (TSV) bonded memory die that enables impressive advantages over current technology. According to company figures, a single HMC offers a 15x performance increase and uses 70 percent less energy per bit when compared to DDR3 memory, and takes up 90 percent less space than today's RDIMMs. The Cube is also scalable per application, which is not possible with DDR3 and DDR4. System designers have the option of employing the HMC as near memory for best performance or in a scalable module form factor, as far memory, for optimum power efficiency. This is a huge leap forward from a technology perspective, noted Graham, compared to DDRx and other boutique memory products that are out there.
http://www.hpcwire.com/hpcwire/2013-...for_debut.html
Thanks for the links. Never heard of it till now, so it was interesting to read.
Also, to pour some gasoline on the fire I found this link as well:
It will have applications in servers but the most obvious application is in games. Faster memory together with faster processors will make for smoother high-resolution graphics. That is where you will notice the most change. It won't make any noticeable difference for simple things like accounting or word processing.
Linky
Don't care. The company with the most games I want to play this fall is the one that will get my money for a new console. Simple as that.
Sounds like specs got better for Orbis :D
Orbis not only has a more powerful GPU,but now also has 4 compute units dedicated for whatever like cloth physics and such :P
On AMD Tahiti's CU, it combines two 32bit ALUs to form one 64bit ALU. AMD Tahiti has a ratio 1:4 on 64bit vs 32bit i.e. 16 ALUs with 64bit mode per CU. From AMD's dev forum, the next flagship GCN has a ratio 1:2 on 64bit vs 32bit i.e. 32 ALUs with 64bit mode per CU. It looks like PS4 would be based on next flagship GCN's CU but cut down i.e. reduced from 40 CU to 18 CU.UPDATE: some people is confused about the GPU, here you have more info about it:
Each CU contains dedicated:
- ALU (32 64-bit operations per cycle)
- Texture Unit
- L1 data cache
- Local data share (LDS)
About 14 + 4 balance:
- 4 additional CUs (410 Gflops) extra ALU as resource for compute
- Minor boost if used for rendering
Dual Shader Engines:
- 1.6 billion triangles/s, 1.6 billion vertices/s
18 Texture units
- 56 billion bilinear texture reads/s
- Can utilize full memory bandwith
8 Render backends:
- 32 color ops/cycle
- 128 depth ops/cycle
- Can utilize full memory bandwith
All this info issubject to changein the future by Sony.
Blazerdt47
Day 1, my GTX 670 will be rapedGioVela2010
On 64bit shader, Liverpool GPU would murder it.
In terms of 64bit shader compute, Liverpool = 32 x 18 CU = 576 stream processors @ 800 Mhz
Tahiti XT2 = 16 x 32 CU = 512 stream processors @ 1Ghz
Tahiti XT = 16 x 32 CU = 512 stream processors @ 925 Mhz
It's x1900/Xenos (full speed 32bit compute) vs G7x again....
PS; Atm, I don't see the benefit of 64bit compute for games.
[QUOTE="loco145"]If Sony expects to sell any systems, it can't be more than $399. It just can't be. The market is different now. Personally I think people are more willing, now, to spend higher amounts than in the past. People spend spend $700 on iPads that are just really expensive toys that begin phasing out just after a few years.$599.
ItsEvolution
[QUOTE="TheXFiles88"]
[QUOTE="tormentos"] Actually i wasn't celebrating but if you most know the PS4 one still quite faster,the 720 on is 68GB/s + 102GB/s but that doesn't mean they should be accounted like 170GB/s,go search on Beyond or Neogaf you can't combine those numbers,is 68 from one point to another and 102 from one point to another.. Unlike the PS4 176GB/s one.SUD123456
You know, the DDRx DRAM has already been outclassed by a new revolutionary DRAM which called "Hybrid Memory Cube" It is a multi-chip module which works in parallel with the multi-core processors. The advantages are enormous over the DDRx. It costs less, 15x more performance, consumes less power and takes less space than a typical DDR3. If Microsoft is going to use this new HMC in its new Xbox (I don't think why they shouldn't), it will be a game changer since Microsoft is currently 1 of the 10 members of the HMC consortium. Just wait for the final specs after E3...;)
"Hybrid Memory Cube is a revolutionary innovation in DRAM memory architecture that sets a new standard for memory performance, power consumption and cost."
- HMC Combines high-speed logic process technology with a stack of through-silicon-via (TSV) bonded memory die.
- HMC delivers dramatic improvements in performance, breaking through the memory wall and enabling dramatic performance and bandwidth improvements -a single HMC can provide more than 15x the performance of a DDR3 module.
- The revolutionary architecture of HMC is exponentially more efficient than current memory, utilizing 70% less energy per bit than DDR3 DRAM technologies..
- Hybrid Memory Cube's increased density per bit and reduced form factor contribute to lower total cost of ownership, by allowing more memory into each machine and using nearly 90% less space than today's RDIMMs.
- With performance levels that break through the memory wall, Hybrid Memory Cube represents the key to extending network system performance to push through the challenges of new 100G and 400G infrastructure growth. Eventually, HMC will drive exascale CPU system performance growth for next generation HPC systems.
- Whereas DDR4 represents an evolutionary standard, HMC is a revolutionary technology that is a complete paradigm shift from current memory architectures.
- Hybrid Memory Cube will redefine memory. By advancing past the traditional DRAM system, HMC is setting a new standard of memory that can keep up with the advancements of CPUs and GPUs.
- Hybrid Memory Cube could be an absolute game changer for applications ranging from high performance computing to consumer technologies like tablets and graphics cards that value a combination of form factor, energy and bandwidth.
http://www.hybridmemorycube.org/technology.html
"Micron Readies Hybrid Memory Cube for Debut"
Hybrid Memory Cube (HMC) technology, a multi-chip module (MCM) that aims to address one of the biggest challenges in high performance computing: scaling the memory wall.
Memory architectures haven't kept pace with the bandwidth requirements of multicore processors. As microprocessor speeds out-accelerated DRAM memory speeds, a bottleneck developed that is referred to as the memory wall. Stacked memory applications, however, enable higher memory bandwidth.
The Hybrid Memory Cube (HMC) is a new memory architecture that combines a high-speed logic layer with a stack of through-silicon-via (TSV) bonded memory die that enables impressive advantages over current technology. According to company figures, a single HMC offers a 15x performance increase and uses 70 percent less energy per bit when compared to DDR3 memory, and takes up 90 percent less space than today's RDIMMs. The Cube is also scalable per application, which is not possible with DDR3 and DDR4. System designers have the option of employing the HMC as near memory for best performance or in a scalable module form factor, as far memory, for optimum power efficiency. This is a huge leap forward from a technology perspective, noted Graham, compared to DDRx and other boutique memory products that are out there.
http://www.hpcwire.com/hpcwire/2013-...for_debut.html
Thanks for the links. Never heard of it till now, so it was interesting to read.
Also, to pour some gasoline on the fire I found this link as well:
It will have applications in servers but the most obvious application is in games. Faster memory together with faster processors will make for smoother high-resolution graphics. That is where you will notice the most change. It won't make any noticeable difference for simple things like accounting or word processing.
Linky
Yeah and this quote:"The Cube is also scalable per application....System designers have the option of employing the HMC as near memory for best performance or in a scalable module form factor, as far memory, for optimum power efficiency"is very interesting since it goes in lieu with that leaked "Scalable Console patent". Here is the chart of how the HMC vs. DRR3 & DDR4:
[QUOTE="SUD123456"]
[QUOTE="TheXFiles88"]
You know, the DDRx DRAM has already been outclassed by a new revolutionary DRAM which called "Hybrid Memory Cube" It is a multi-chip module which works in parallel with the multi-core processors. The advantages are enormous over the DDRx. It costs less, 15x more performance, consumes less power and takes less space than a typical DDR3. If Microsoft is going to use this new HMC in its new Xbox (I don't think why they shouldn't), it will be a game changer since Microsoft is currently 1 of the 10 members of the HMC consortium. Just wait for the final specs after E3...;)
"Hybrid Memory Cube is a revolutionary innovation in DRAM memory architecture that sets a new standard for memory performance, power consumption and cost."
- HMC Combines high-speed logic process technology with a stack of through-silicon-via (TSV) bonded memory die.
- HMC delivers dramatic improvements in performance, breaking through the memory wall and enabling dramatic performance and bandwidth improvements -a single HMC can provide more than 15x the performance of a DDR3 module.
- The revolutionary architecture of HMC is exponentially more efficient than current memory, utilizing 70% less energy per bit than DDR3 DRAM technologies..
- Hybrid Memory Cube's increased density per bit and reduced form factor contribute to lower total cost of ownership, by allowing more memory into each machine and using nearly 90% less space than today's RDIMMs.
- With performance levels that break through the memory wall, Hybrid Memory Cube represents the key to extending network system performance to push through the challenges of new 100G and 400G infrastructure growth. Eventually, HMC will drive exascale CPU system performance growth for next generation HPC systems.
- Whereas DDR4 represents an evolutionary standard, HMC is a revolutionary technology that is a complete paradigm shift from current memory architectures.
- Hybrid Memory Cube will redefine memory. By advancing past the traditional DRAM system, HMC is setting a new standard of memory that can keep up with the advancements of CPUs and GPUs.
- Hybrid Memory Cube could be an absolute game changer for applications ranging from high performance computing to consumer technologies like tablets and graphics cards that value a combination of form factor, energy and bandwidth.
http://www.hybridmemorycube.org/technology.html
"Micron Readies Hybrid Memory Cube for Debut"
Hybrid Memory Cube (HMC) technology, a multi-chip module (MCM) that aims to address one of the biggest challenges in high performance computing: scaling the memory wall.
Memory architectures haven't kept pace with the bandwidth requirements of multicore processors. As microprocessor speeds out-accelerated DRAM memory speeds, a bottleneck developed that is referred to as the memory wall. Stacked memory applications, however, enable higher memory bandwidth.
The Hybrid Memory Cube (HMC) is a new memory architecture that combines a high-speed logic layer with a stack of through-silicon-via (TSV) bonded memory die that enables impressive advantages over current technology. According to company figures, a single HMC offers a 15x performance increase and uses 70 percent less energy per bit when compared to DDR3 memory, and takes up 90 percent less space than today's RDIMMs. The Cube is also scalable per application, which is not possible with DDR3 and DDR4. System designers have the option of employing the HMC as near memory for best performance or in a scalable module form factor, as far memory, for optimum power efficiency. This is a huge leap forward from a technology perspective, noted Graham, compared to DDRx and other boutique memory products that are out there.
http://www.hpcwire.com/hpcwire/2013-...for_debut.html
TheXFiles88
Thanks for the links. Never heard of it till now, so it was interesting to read.
Also, to pour some gasoline on the fire I found this link as well:
It will have applications in servers but the most obvious application is in games. Faster memory together with faster processors will make for smoother high-resolution graphics. That is where you will notice the most change. It won't make any noticeable difference for simple things like accounting or word processing.
Linky
Yeah and this quote:"The Cube is also scalable per application....System designers have the option of employing the HMC as near memory for best performance or in a scalable module form factor, as far memory, for optimum power efficiency"is very interesting since it goes in lieu with that leaked "Scalable Console patent". Here is the chart of how the HMC vs. DRR3 & DDR4:
Another stack memory i.e. HBM http://sites.amd.com/us/Documents/TFE2011_006HYN.pdf Document dated 3/10/2011.
http://semiaccurate.com/2011/10/27/amd-far-future-prototype-gpu-pictured/
http://www.eetimes.com/electronics-news/4369672/Micron-broadening-support-for-memory-cube
Micron is not alone in the pursuit of a 3-D memory stack. The Jedec group is working on a follow on to the 12.8 Gbit/second Wide I/O interface that targets mobile applications processors. The so-called HB-DRAM or HBM effort is said to target a 120-128 Gbyte/second interface and is led by the Jedec JC-42 committee including representatives from Hynix and other companies.
AMD would need stack memory for AMD Kaveri APU i.e. running into the memory bandwidth wall with Trinity.
Sounds like specs got better for Orbis :D
Orbis not only has a more powerful GPU,but now also has 4 compute units dedicated for whatever like cloth physics and such :P
killzowned24
14 + 4 split is most likely using GCN's hardware partitioning feature i.e. a feature not used in DX..
[QUOTE="TheXFiles88"]
[QUOTE="SUD123456"]
Thanks for the links. Never heard of it till now, so it was interesting to read.
Also, to pour some gasoline on the fire I found this link as well:
It will have applications in servers but the most obvious application is in games. Faster memory together with faster processors will make for smoother high-resolution graphics. That is where you will notice the most change. It won't make any noticeable difference for simple things like accounting or word processing.
Linky
ronvalencia
Yeah and this quote:"The Cube is also scalable per application....System designers have the option of employing the HMC as near memory for best performance or in a scalable module form factor, as far memory, for optimum power efficiency"is very interesting since it goes in lieu with that leaked "Scalable Console patent". Here is the chart of how the HMC vs. DRR3 & DDR4:
Another stack memory i.e. HBM http://sites.amd.com/us/Documents/TFE2011_006HYN.pdf Document dated 3/10/2011.
http://semiaccurate.com/2011/10/27/amd-far-future-prototype-gpu-pictured/
http://www.eetimes.com/electronics-news/4369672/Micron-broadening-support-for-memory-cube
Micron is not alone in the pursuit of a 3-D memory stack. The Jedec group is working on a follow on to the 12.8 Gbit/second Wide I/O interface that targets mobile applications processors. The so-called HB-DRAM or HBM effort is said to target a 120-128 Gbyte/second interface and is led by the Jedec JC-42 committee including representatives from Hynix and other companies.
AMD would need stack memory for AMD Kaveri APU i.e. running into the memory bandwidth wall with Trinity.
I recognized AMD also has a similar design for its new GPU architectures. However, I am more interested of its "scalability per application" such as that leaked "scalable console patent" from the new Xbox's road map.
These numbers don't add up for me. 1.6 billion triangles per second seems low while xenon did 500 million per sec. Sony claimed the ps3 could do 2.0 teraflops and this is 1.8? Would it really be that expensive to overclock that CPU to atleast 2.0ghz? Audio processor? Doesn't the CPU usually handle that. Where is ronvalencia when u need him.GotNugz
AMD Xenos couldn't sustain 500 million triangles per second ie.g. it needs GCN's geometry design fixes.
http://www.digitalartsonline.co.uk/news/creative-hardware/siggraph-2012-amd-launches-firepro-w5000-w7000-w8000-w9000-graphics-card/
AMD FirePro W5000 has 1.65 billion triangles per second and it's based on Radeon HD 7850 with 768 stream processors (12 CU).
PS4's Liverpood GPU seems to be GCN v2.0's CU design with 7850's front-end and back-end(1) i.e. it's semi-customised.
1. http://www.amd.com/us/products/desktop/graphics/7000/7850/Pages/radeon-7850.aspx#3
[QUOTE="ronvalencia"]
[QUOTE="TheXFiles88"]
Yeah and this quote:"The Cube is also scalable per application....System designers have the option of employing the HMC as near memory for best performance or in a scalable module form factor, as far memory, for optimum power efficiency"is very interesting since it goes in lieu with that leaked "Scalable Console patent". Here is the chart of how the HMC vs. DRR3 & DDR4:
TheXFiles88
Another stack memory i.e. HBM http://sites.amd.com/us/Documents/TFE2011_006HYN.pdf Document dated 3/10/2011.
http://semiaccurate.com/2011/10/27/amd-far-future-prototype-gpu-pictured/
http://www.eetimes.com/electronics-news/4369672/Micron-broadening-support-for-memory-cube
Micron is not alone in the pursuit of a 3-D memory stack. The Jedec group is working on a follow on to the 12.8 Gbit/second Wide I/O interface that targets mobile applications processors. The so-called HB-DRAM or HBM effort is said to target a 120-128 Gbyte/second interface and is led by the Jedec JC-42 committee including representatives from Hynix and other companies.
AMD would need stack memory for AMD Kaveri APU i.e. running into the memory bandwidth wall with Trinity.
I recognized AMD also has a similar design for its new GPU architectures. However, I am more interested of its "scalability per application" such as that leaked "scalable console patent" from the new Xbox's road map.
If Microsoft funds HMC for AMD GPUs, then AMD could benefit from it e.g. AMD's unified shader tech was partly funded by MS.Uncharted 3 @ 60 fps http://www.eurogamer.net/articles/digitalfoundry-what-if-uncharted-3-ran-at-60-fps
There are some indications Uncharted 3 @60 fps version could be running on a dev PC.
BC (via small exe replacement download) shouldn't be problem with Sony's 1st party games on AMD "Gaming Evolved" based PC box.
[QUOTE="GotNugz"][QUOTE="campzor"]I cant understand tech mumbo jumbo.. just tell me with dbz power levels how it compares to durango (rumored) + wii u + ps3/360campzorPlaystation 4 = Goku Xbox 720 = Vegeta Wii U = Krillin PS3/360 = Saibamen im literally lol'n with the ps3/360 - wii u comparison..
Yea that was pretty good lol
Or they just rendered it out at 60fps from a replay like your link implied.Uncharted 3 @ 60 fps http://www.eurogamer.net/articles/digitalfoundry-what-if-uncharted-3-ran-at-60-fps
There are some indications Uncharted 3 @60 fps version could be running on a dev PC.
BC (via small exe replacement download) shouldn't be problem with Sony's 1st party games on AMD "Gaming Evolved" based PC box.
ronvalencia
[QUOTE="ronvalencia"]Or they just rendered it out at 60fps from a replay like your link implied. If its captured at 30FPS, rendering it again at 60 is going to do nothing.Uncharted 3 @ 60 fps http://www.eurogamer.net/articles/digitalfoundry-what-if-uncharted-3-ran-at-60-fps
There are some indications Uncharted 3 @60 fps version could be running on a dev PC.
BC (via small exe replacement download) shouldn't be problem with Sony's 1st party games on AMD "Gaming Evolved" based PC box.
ferret-gamer
Actually i wasn't celebrating but if you most know the PS4 one still quite faster,the 720 on is 68GB/s + 102GB/s but that doesn't mean they should be accounted like 170GB/s,go search on Beyond or Neogaf you can't combine those numbers,is 68 from one point to another and 102 from one point to another.. Unlike the PS4 176GB/s one.[QUOTE="tormentos"][QUOTE="Tessellation"] he was celebrating the 192GB/s of bandwidth and now that they changed it to 176GB/s is around durango's territory,of course he is going to deny these rumors because they don't make his future plastic girlfriend any better :cool:TheXFiles88
You know, the DDRx DRAM has already been outclassed by a new revolutionary DRAM which called "Hybrid Memory Cube" It is a multi-chip module which works in parallel with the multi-core processors. The advantages are enormous over the DDRx. It costs less, 15x more performance, consumes less power and takes less space than a typical DDR3. If Microsoft is going to use this new HMC in its new Xbox (I don't think why they shouldn't), it will be a game changer since Microsoft is currently 1 of the 10 members of the HMC consortium. Just wait for the final specs after E3...;)
"Hybrid Memory Cube is a revolutionary innovation in DRAM memory architecture that sets a new standard for memory performance, power consumption and cost."
- HMC Combines high-speed logic process technology with a stack of through-silicon-via (TSV) bonded memory die.
- HMC delivers dramatic improvements in performance, breaking through the memory wall and enabling dramatic performance and bandwidth improvements -a single HMC can provide more than 15x the performance of a DDR3 module.
- The revolutionary architecture of HMC is exponentially more efficient than current memory, utilizing 70% less energy per bit than DDR3 DRAM technologies..
- Hybrid Memory Cube's increased density per bit and reduced form factor contribute to lower total cost of ownership, by allowing more memory into each machine and using nearly 90% less space than today's RDIMMs.
- With performance levels that break through the memory wall, Hybrid Memory Cube represents the key to extending network system performance to push through the challenges of new 100G and 400G infrastructure growth. Eventually, HMC will drive exascale CPU system performance growth for next generation HPC systems.
- Whereas DDR4 represents an evolutionary standard, HMC is a revolutionary technology that is a complete paradigm shift from current memory architectures.
- Hybrid Memory Cube will redefine memory. By advancing past the traditional DRAM system, HMC is setting a new standard of memory that can keep up with the advancements of CPUs and GPUs.
- Hybrid Memory Cube could be an absolute game changer for applications ranging from high performance computing to consumer technologies like tablets and graphics cards that value a combination of form factor, energy and bandwidth.
http://www.hybridmemorycube.org/technology.html
"Micron Readies Hybrid Memory Cube for Debut"
Hybrid Memory Cube (HMC) technology, a multi-chip module (MCM) that aims to address one of the biggest challenges in high performance computing: scaling the memory wall.
Memory architectures haven't kept pace with the bandwidth requirements of multicore processors. As microprocessor speeds out-accelerated DRAM memory speeds, a bottleneck developed that is referred to as the memory wall. Stacked memory applications, however, enable higher memory bandwidth.
The Hybrid Memory Cube (HMC) is a new memory architecture that combines a high-speed logic layer with a stack of through-silicon-via (TSV) bonded memory die that enables impressive advantages over current technology. According to company figures, a single HMC offers a 15x performance increase and uses 70 percent less energy per bit when compared to DDR3 memory, and takes up 90 percent less space than today's RDIMMs. The Cube is also scalable per application, which is not possible with DDR3 and DDR4. System designers have the option of employing the HMC as near memory for best performance or in a scalable module form factor, as far memory, for optimum power efficiency. This is a huge leap forward from a technology perspective, noted Graham, compared to DDRx and other boutique memory products that are out there.
http://www.hpcwire.com/hpcwire/2013-...for_debut.html
Shoud i do that a go with GDDR5 best and fastest solution.? No that would be call wishful thinking,the fact that the ESRAM exist basically confirms that you will not get that Ram on 720,ESRAM is there for a reason,to help the starved DDR3 memory.[QUOTE="ferret-gamer"][QUOTE="ronvalencia"]Or they just rendered it out at 60fps from a replay like your link implied. If its captured at 30FPS, rendering it again at 60 is going to do nothing. Uncharted 3 has a replay editor, it wouldn't be that hard for them render it out at 60fps, any decent replay system in a game could do it. I'm pretty sure there is a youtube video of someone who used Source engine's replay system to make a slow mo 10,000fps vid.Uncharted 3 @ 60 fps http://www.eurogamer.net/articles/digitalfoundry-what-if-uncharted-3-ran-at-60-fps
There are some indications Uncharted 3 @60 fps version could be running on a dev PC.
BC (via small exe replacement download) shouldn't be problem with Sony's 1st party games on AMD "Gaming Evolved" based PC box.
clyde46
Please Log In to post.
Log in to comment