Gigabyte GeForce GTX 650 Ti Boost OC WindForce 2X review
MSI Radeon HD 7790 TurboDuo OC review
Metro Last Light VGA Graphics Benchmark performance test
Noctua NH-U12S and NH-U14S review
ASUS GeForce GTX 670 DirectCU Mini review
OCZ Vertex 3.20 SSD review
Gigabyte Radeon HD 7790 2GB OC review
Cooler Master Eisberg 240L Prestige review
Guru3D and OCZ Contest - PC Power 1200W PSU Giveaway
MSI GeForce GTX 650 Ti BOOST OC review
Dodgy Radeon HD 8000 Series GPU Specifications Surface
Posted by Hilbert Hagedoorn on: 12/04/2012 09:09 AM | 12 comment(s) ]
Yesterdau we already reported that AMD is to launch Radeon HD 8000 series in 2Q13. AMD's Radeon HD 8000 "Sea Islands" family of GPUs may arrive some time in Q2-2013, but they won't arrive without a specifications overhaul. While NVIDIA is counting on performance enhancements to come out of higher clock speeds on existing silicon, while maintaining current (or lower) power-draw, with its "Enhanced Kepler" family of GPUs that will include the "GK110" juggernaut, AMD is counting on physically bigger chips with more components. AMD could step up transistor counts of its chips by as much as 20 percent, on existing 28 nm process.
According to TPU, the biggest chip from AMD's fold could pack 5.1 billion transistors, amounting to 2560 stream processors, and an updated raster processing engine, with 48 ROPs ("Tahiti" has 32). While the Radeon HD 8970 maxes out the chip's capabilities, the HD 8950 could feature 256 fewer stream processors, at 2304. It could also go light on the clock speeds. AMD's performance-segment chip, codenamed "Sun" could see a similar stream processor increase to 1792, with the Radeon HD 8870 maxing it out, and HD 8850 featuring 1536. It maintains the memory bus width and raster engine layout of its predecessor.
AMD's mainstream chip, codenamed "Oland," which succeeds "Cape Verde," could address the problem of memory bandwidth, which is giving competing NVIDIA GPUs an upper hand in its segment. It could feature 896 stream processors, and a wider 192-bit GDDR5 memory interface. The Radeon HD 8770 maxes its specifications out, while the Radeon HD 8750 features a slimmer 128-bit memory interface, and 768 stream processors.
The authenticity of this information can't be vouched for, but presents the only way in which AMD can end up with a competitive GPU lineup against NVIDIA, over 2013 - Via TPU.
Dodgy Radeon HD 8000 Series GPU Specifications Surface
tuco
Ancient Guru
Posts: 3036
Joined: 2006-02-27
Ancient Guru
Posts: 3036
Joined: 2006-02-27
#4470942 Posted on: 12/04/2012 04:25 PM
damm, I read it as doggy at first. I'd love an ATI gpu that did it that style.
damm, I read it as doggy at first. I'd love an ATI gpu that did it that style.
morbias
Moderator
Posts: 12399
Joined: 2004-09-20
Moderator
Posts: 12399
Joined: 2004-09-20
#4471079 Posted on: 12/04/2012 06:05 PM
more transistors = lower maximum clock speed
so it will be interesting how this pans out
as a consumer, I'd prefer AMDs route because they should be easier to overclock
more transistors = lower maximum clock speed
so it will be interesting how this pans out
sykozis
Ancient Guru
Posts: 13493
Joined: 2008-07-14
Ancient Guru
Posts: 13493
Joined: 2008-07-14
#4471088 Posted on: 12/04/2012 06:11 PM
as a consumer, I'd prefer AMDs route because they should be easier to overclock, and also get a greater performance improvement for each step up vs nvidia.
I do think its a little dumb that both companies are basically taking the easy way out when it comes to a new generation gpu.
How are they taking the easy way out?
as a consumer, I'd prefer AMDs route because they should be easier to overclock, and also get a greater performance improvement for each step up vs nvidia.
I do think its a little dumb that both companies are basically taking the easy way out when it comes to a new generation gpu.
How are they taking the easy way out?
Xzibit
Banned
Posts: 4383
Joined: 2009-12-11
Banned
Posts: 4383
Joined: 2009-12-11
#4471096 Posted on: 12/04/2012 06:17 PM
more transistors = lower maximum clock speed
Oh no. This is an epic fail.
more transistors = lower maximum clock speed
Oh no. This is an epic fail.
BABA-The Hacker
Master Guru
Posts: 302
Joined: 2008-02-28
Master Guru
Posts: 302
Joined: 2008-02-28
#4471206 Posted on: 12/04/2012 07:31 PM
....dont forget......
http://www.guru3d.com/news_story/eu_could_cripple_future_graphics_cards.html
....dont forget......
http://www.guru3d.com/news_story/eu_could_cripple_future_graphics_cards.html
EU Could Cripple future graphics cards
Posted by: Hilbert Hagedoorn on: 10/15/2012 05:20 AM | 50 comment(s) ]
A very interesting story appeared on a colleague website today. New information about a new energy law that will apply within the EU has surfaced on the web. The new law requires that both discrete and integrated graphics cards live up to certain energy standards. AMD is worried that this will affect next generation graphics cards and have them barred from sales in the EU. There are standardizations that make sure pre-built computers, but also discrete components, achieve a certain level of energy efficiency.
The story is reported by Nordic Hardware so credits and props to them, here's their skinny on it:
---
Exactly how much depends on a row of criteria. These standards also include simple things, such as that after a certain amount of time the computer will enter sleep mode. The idea behind this is to have as energy efficient computers as possible to reduce the overall consumption of energy. The specification for the so called Eco design Lot 3 with the EC can be found here, where there are hundreds of pages to read for those with lots of time to spare.
FB_BW is short for Buffer Bandwidth. What we call memory bandwidth in graphics cards
There are currently seven specifications for graphics cards - G1, G2, G3, G4, G5, G6 and G7. Graphics cards of the G7 classification have a bandwidth of 128 GB/s (GigaByte per Second) and more, without an upper limit today. The category depends on the performance - in this case measured in memory bandwidth. These GPU categories are also paired with a certain level of energy efficiency. If a graphics card doesn't live up to the standard set by the EC it can be removed from all markets within the EU. The rules will now be constricted, which threatens next generation graphics cards.
The commission wants to stop dedicated graphics cards of group G7 from going above 320 GB/s - that is in theory a memory bus at 384-bit connected to memory operating at 6667 MHz or 512-bit with 5001 MHz. This is definitely within reach for the next generation graphics cards. Radeon HD 7970 GHz Edition currently has a bandwidth of 288 GB/s with a 384-bit memory bus and 6000 MHz memory. For notebooks the limit will be only 225 GB/s.
Besides that the energy efficiency requirements will be tighter - in this case the energy consumption of the card in relation to its memory bandwidth. Performance delivered in games or general calculations are irrelevant. according to Lot 3. Exactly what the "performance" and energy consumption quote looks like we don't know at the time of writing, but it will also affect cards in the entry level segment and not just performance and enthusiast cards. The quote is strict enough to worry AMD.
Future generations are in danger
According to data NordicHardware has seen from a high level employee at AMD, current graphics cards are unable to meet with these requirements. This includes "GPUs like Cape Verde and Tahiti",that is used in the HD 7700 and HD 7900 series, and can't meet with the new guidelines, the same goes for the older "Caicos" that is used in the HD 6500/6600 and HD 7500/7600 series. Also "Oland"is mentioned, which is a future performance circuit from AMD, that according to rumors will be used in the future HD 8800 series. What worries AMD the most is how this will affect future graphics cards since the changes in Lot 3 will go into effect soon. The changes will of course affect Nvidia as much as it will AMD.
The commission guidelines could be too strict for next generation graphics cards
Earlier today there were talk about the new restrictions going into effect in early 2013, but now it looks like it will be 2014. This will put nearly unrealistic demands on both AMD and Nvidia. Besides the fact the standardization is not very logical since memory bandwidth does not translate into performance that easily we see it as a great obstacle for future graphics cards, but the revision of "Lot 3" is done and the wheels are set in motion.
According to a report published in August this year the current roadmaps does not support the new requirements up until 30 months into the future. The changes in Lot 3 will therefore be introduced in steps. The first will be in 2013 or 2014 as mentioned above, and thereafter new restrictions will apply in 2015. OEM companies like Dell and HP are well aware of this and worried about how this will affect their operations. The changes should also affect retail graphics cards and home builders.
Graphics card energy consumption has been rising steadily over the last couple of years. Last generation the limit for the PCI Express standard was broken when graphics cards sporting two GPUs consumed well over 300 watt, both from AMD and Nvidia. Both us and our well informed readers think that the way the EC is applying its restrictions is wrong, especially how it has decided to estimate performance. AMD is planning on making an official statement and hopefully the EC will listen. We have not been able to reach Nvidia for a comment on this issue, but hopefully they share AMD's concern.
We definitely feel that restrictions that lead to more efficient hardware is a good thing, but it needs to be done properly with the affected companies being involved in the discussion. We will of course follow up on this and return with more information when possible.
Via Nordic Hardware
Posted by: Hilbert Hagedoorn on: 10/15/2012 05:20 AM | 50 comment(s) ]
A very interesting story appeared on a colleague website today. New information about a new energy law that will apply within the EU has surfaced on the web. The new law requires that both discrete and integrated graphics cards live up to certain energy standards. AMD is worried that this will affect next generation graphics cards and have them barred from sales in the EU. There are standardizations that make sure pre-built computers, but also discrete components, achieve a certain level of energy efficiency.
The story is reported by Nordic Hardware so credits and props to them, here's their skinny on it:
---
Exactly how much depends on a row of criteria. These standards also include simple things, such as that after a certain amount of time the computer will enter sleep mode. The idea behind this is to have as energy efficient computers as possible to reduce the overall consumption of energy. The specification for the so called Eco design Lot 3 with the EC can be found here, where there are hundreds of pages to read for those with lots of time to spare.
FB_BW is short for Buffer Bandwidth. What we call memory bandwidth in graphics cards
There are currently seven specifications for graphics cards - G1, G2, G3, G4, G5, G6 and G7. Graphics cards of the G7 classification have a bandwidth of 128 GB/s (GigaByte per Second) and more, without an upper limit today. The category depends on the performance - in this case measured in memory bandwidth. These GPU categories are also paired with a certain level of energy efficiency. If a graphics card doesn't live up to the standard set by the EC it can be removed from all markets within the EU. The rules will now be constricted, which threatens next generation graphics cards.
The commission wants to stop dedicated graphics cards of group G7 from going above 320 GB/s - that is in theory a memory bus at 384-bit connected to memory operating at 6667 MHz or 512-bit with 5001 MHz. This is definitely within reach for the next generation graphics cards. Radeon HD 7970 GHz Edition currently has a bandwidth of 288 GB/s with a 384-bit memory bus and 6000 MHz memory. For notebooks the limit will be only 225 GB/s.
Besides that the energy efficiency requirements will be tighter - in this case the energy consumption of the card in relation to its memory bandwidth. Performance delivered in games or general calculations are irrelevant. according to Lot 3. Exactly what the "performance" and energy consumption quote looks like we don't know at the time of writing, but it will also affect cards in the entry level segment and not just performance and enthusiast cards. The quote is strict enough to worry AMD.
Future generations are in danger
According to data NordicHardware has seen from a high level employee at AMD, current graphics cards are unable to meet with these requirements. This includes "GPUs like Cape Verde and Tahiti",that is used in the HD 7700 and HD 7900 series, and can't meet with the new guidelines, the same goes for the older "Caicos" that is used in the HD 6500/6600 and HD 7500/7600 series. Also "Oland"is mentioned, which is a future performance circuit from AMD, that according to rumors will be used in the future HD 8800 series. What worries AMD the most is how this will affect future graphics cards since the changes in Lot 3 will go into effect soon. The changes will of course affect Nvidia as much as it will AMD.
The commission guidelines could be too strict for next generation graphics cards
Earlier today there were talk about the new restrictions going into effect in early 2013, but now it looks like it will be 2014. This will put nearly unrealistic demands on both AMD and Nvidia. Besides the fact the standardization is not very logical since memory bandwidth does not translate into performance that easily we see it as a great obstacle for future graphics cards, but the revision of "Lot 3" is done and the wheels are set in motion.
According to a report published in August this year the current roadmaps does not support the new requirements up until 30 months into the future. The changes in Lot 3 will therefore be introduced in steps. The first will be in 2013 or 2014 as mentioned above, and thereafter new restrictions will apply in 2015. OEM companies like Dell and HP are well aware of this and worried about how this will affect their operations. The changes should also affect retail graphics cards and home builders.
Graphics card energy consumption has been rising steadily over the last couple of years. Last generation the limit for the PCI Express standard was broken when graphics cards sporting two GPUs consumed well over 300 watt, both from AMD and Nvidia. Both us and our well informed readers think that the way the EC is applying its restrictions is wrong, especially how it has decided to estimate performance. AMD is planning on making an official statement and hopefully the EC will listen. We have not been able to reach Nvidia for a comment on this issue, but hopefully they share AMD's concern.
We definitely feel that restrictions that lead to more efficient hardware is a good thing, but it needs to be done properly with the affected companies being involved in the discussion. We will of course follow up on this and return with more information when possible.
Via Nordic Hardware
zer0_c0ol
Maha Guru
Posts: 1843
Joined: 2007-04-12
Maha Guru
Posts: 1843
Joined: 2007-04-12
#4471224 Posted on: 12/04/2012 07:46 PM
even the eu energy commission will **** with my gaming..
even the eu energy commission will **** with my gaming..
BABA-The Hacker
Master Guru
Posts: 302
Joined: 2008-02-28
Master Guru
Posts: 302
Joined: 2008-02-28
#4471250 Posted on: 12/04/2012 08:12 PM
.....Indeed, sir....
there are only 2 of them... ATI and NVidia... if anyone dies from them....
It is only (us) would suffer...
.....Indeed, sir....
there are only 2 of them... ATI and NVidia... if anyone dies from them....
It is only (us) would suffer...
Lane
Ancient Guru
Posts: 4875
Joined: 2005-02-25
Ancient Guru
Posts: 4875
Joined: 2005-02-25
#4471268 Posted on: 12/04/2012 08:29 PM
Dont forget thoses are only pure rumors .. Even there DP/SP dont look to match ( 5.38Tflops SP and 1.6 Tflops DP ? ) Too high on DP and with a strange ration between both ) 1: 3.336 )..
Dont forget thoses are only pure rumors .. Even there DP/SP dont look to match ( 5.38Tflops SP and 1.6 Tflops DP ? ) Too high on DP and with a strange ration between both ) 1: 3.336 )..
Titan29
Member Guru
Posts: 147
Joined: 2011-10-21
Member Guru
Posts: 147
Joined: 2011-10-21
#4471361 Posted on: 12/04/2012 10:01 PM
Interesting. Those specs have been on wikipedia for more than a month from what I have seen.
http://en.wikipedia.org/wiki/Sea_Islands_%28GPU_family%29
Interesting. Those specs have been on wikipedia for more than a month from what I have seen.
http://en.wikipedia.org/wiki/Sea_Islands_%28GPU_family%29
schmidtbag
Member Guru
Posts: 97
Joined: 2012-11-10
Member Guru
Posts: 97
Joined: 2012-11-10
#4475133 Posted on: 12/09/2012 02:33 AM
Because instead of finding ways to actually improve performance of each thread and each clock cycle, they're just taking on more of them instead.
How are they taking the easy way out?
Because instead of finding ways to actually improve performance of each thread and each clock cycle, they're just taking on more of them instead.
---TK---
Ancient Guru
Posts: 14712
Joined: 2005-12-10
Ancient Guru
Posts: 14712
Joined: 2005-12-10
#4475138 Posted on: 12/09/2012 02:39 AM
Oh no. Ban Hammer?
Oh no. This is an epic fail.
Oh no. Ban Hammer?
Click here to post a comment for this news story on the message forum.
Member Guru
Posts: 97
Joined: 2012-11-10
as a consumer, I'd prefer AMDs route because they should be easier to overclock, and also get a greater performance improvement for each step up vs nvidia.
I do think its a little dumb that both companies are basically taking the easy way out when it comes to a new generation gpu.