- Joined
- Mar 10, 2010
- Messages
- 11,878 (2.20/day)
- Location
- Manchester uk
System Name | RyzenGtEvo/ Asus strix scar II |
---|---|
Processor | Amd R5 5900X/ Intel 8750H |
Motherboard | Crosshair hero8 impact/Asus |
Cooling | 360EK extreme rad+ 360$EK slim all push, cpu ek suprim Gpu full cover all EK |
Memory | Corsair Vengeance Rgb pro 3600cas14 16Gb in four sticks./16Gb/16GB |
Video Card(s) | Powercolour RX7900XT Reference/Rtx 2060 |
Storage | Silicon power 2TB nvme/8Tb external/1Tb samsung Evo nvme 2Tb sata ssd/1Tb nvme |
Display(s) | Samsung UAE28"850R 4k freesync.dell shiter |
Case | Lianli 011 dynamic/strix scar2 |
Audio Device(s) | Xfi creative 7.1 on board ,Yamaha dts av setup, corsair void pro headset |
Power Supply | corsair 1200Hxi/Asus stock |
Mouse | Roccat Kova/ Logitech G wireless |
Keyboard | Roccat Aimo 120 |
VR HMD | Oculus rift |
Software | Win 10 Pro |
Benchmark Scores | 8726 vega 3dmark timespy/ laptop Timespy 6506 |
I suspect that as process advancement slows down the there will be a permanent move toward two GPU generations on the same silicon process; the initial one right after a shrink being with small dies and the one later in the process's life being with large dies. This allows manufacturers to avoid making large dies early in the life of a manufacturing process when defects are the most common.
Case in point: Think of what NVidia did. They had lots of trouble getting GF100 (a huge die) to work in the early life of 40nm, and they similarly had to cancel the huge GK100 on 28nm and wait for GK110 a year later. AMD has never released a huge die early in the life of a process, and I suspect it's exactly for that reason. The last huge die AMD/ATI released was R600, and that was made on a pretty old (at the time) 80nm process while the smaller chips in the 2000 series were based on the new (at the time) 65nm.
My prediction is that with their experiences on 40nm and 28nm both AMD and NVidia will initally favor small dies on 20nm to avoid defects. This will mean in 2014/2015 the AMD Rx 300 series & NVidia 800 series being only about 20% faster than the AMD Rx 200 and Nvidia 700 series, Then in 2015/2016 the AMD Rx 400 and Nvidia 900 series will be released with big dies on 20nm resulting in another 20% performance improvement. Everyone skipping upgrading this generation expecting a 50% performance improvement next generation is fooling themselves. Those days are over.
I think this decidedly marks a trend toward marketing based on features rather than performance, much like AMD and Intel have done in the CPU space, and toward an increasing focus on software optimization to improve performance when hardware cannot advance.
i allways skip a few gens and got my 50+ yet again
3870 - 3870x2 - 5870 - 5870/5850 - 7970 it did involve some common sense though and being skint a lot or there would be more in that list
anyway,my main point is that 50% is easy doable next arch change for either vendor, both will be hitting out with some interesting tech, maxwell and its 1tb/s memory cache and as for pirate islands, anyones guess there , my guess would include on die l4 cache 256-512mb and 4-8 x86 cores probably jaguer(probably high end gpu though) and both these hit that 50% plus potential you speak of, Tsv(through silicon vias) could shake up the game if used right and creatively.