• We've upgraded our forums. Please post any issues/requests in this thread.

AMD "Greenland" Vega10 Silicon Features 4096 Stream Processors?

btarunr

Editor & Senior Moderator
Staff member
Joined
Oct 9, 2007
Messages
34,335 (9.23/day)
Likes
17,427
Location
Hyderabad, India
System Name Long shelf-life potato
Processor Intel Core i7-4770K
Motherboard ASUS Z97-A
Cooling Xigmatek Aegir CPU Cooler
Memory 16GB Kingston HyperX Beast DDR3-1866
Video Card(s) 2x GeForce GTX 970 SLI
Storage ADATA SU800 512GB
Display(s) Samsung U28D590D 28-inch 4K
Case Cooler Master CM690 Window
Audio Device(s) Creative Sound Blaster Recon3D PCIe
Power Supply Corsair HX850W
Mouse Razer Abyssus 2014
Keyboard Microsoft Sidewinder X4
Software Windows 10 Pro Creators Update
#1
The LinkedIn profile of an R&D manager at AMD discloses key details of the company's upcoming "Greenland" graphics processor, which is also codenamed Vega10. Slated for an early-2017 launch, according to AMD's GPU architecture roadmap, "Greenland" will be built on AMD's "Vega" GPU architecture, which succeeds even the "Polaris" architecture, which is slated for later this year.

Show full news post
 
Joined
Dec 14, 2009
Messages
6,592 (2.25/day)
Likes
5,826
Location
Glasgow - home of formal profanity
System Name New Ho'Ryzen
Processor Ryzen 1700X @ 3.82Ghz
Motherboard Asus Crosshair VI Hero
Cooling TR Le Grand Macho & custom GPU loop
Memory 16Gb G.Skill 3200 RGB
Video Card(s) GTX1080ti (Heatkiller WB) @ 2Ghz core/1.5(12)Ghz mem
Storage Samsumg 960 Pro m2. 512Gb
Display(s) Dell Ultrasharp 27" (2560x1440)
Case Lian Li PC-V33WX
Audio Device(s) On Board
Power Supply Seasonic Prime TItanium 850
Software W10
Benchmark Scores Look, it's a Ryzen on air........ What's the point?
#2
Anyone not seeing the elephant in the room? Polaris releases late 2016, generously maybe by autumn/fall. A few months later, new architecture is released again. Makes zero sense unless they're doing a reverse Nvidia. Vega becomes the Titan style part, with Polaris 10 & 11 being the x80 and x70 parts.

FUD I say.
 
Joined
Sep 2, 2011
Messages
964 (0.42/day)
Likes
554
Location
Porto
System Name No name / Purple Haze
Processor Phenom II 1100T @ 3.8Ghz / Pentium 4 3.4 EE Gallatin @ 3.825Ghz
Motherboard MSI 970 Gaming/ Abit IC7-MAX3
Cooling CM Hyper 212X / Scythe Andy Samurai Master (CPU) - Modded Ati Silencer 5 rev. 2 (GPU)
Memory 8GB GEIL GB38GB2133C10ADC + 8GB G.Skill F3-14900CL9-4GBXL / 2x1GB Crucial Ballistix Tracer PC4000
Video Card(s) Asus R9 Fury X Strix (4096 SP's/1050 Mhz)/ PowerColor X850XT PE @ (600/1230) AGP + (HD3850 AGP)
Storage Samsung 250 GB / WD Caviar 160GB
Display(s) Benq XL2411T
Audio Device(s) motherboard / Creative Sound Blaster X-Fi XtremeGamer Fatal1ty Pro + Front panel
Power Supply Tagan BZ 900W / Corsair HX620w
Mouse Zowie AM
Keyboard Qpad MK-50
Software Windows 7 Pro 64Bit / Windows XP
Benchmark Scores 64CU Fury: http://www.3dmark.com/fs/11269229 / X850XT PE http://www.3dmark.com/3dm05/5532432
#3
Anyone not seeing the elephant in the room? Polaris releases late 2016, generously maybe by autumn/fall. A few months later, new architecture is released again. Makes zero sense unless they're doing a reverse Nvidia. Vega becomes the Titan style part, with Polaris 10 & 11 being the x80 and x70 parts.

FUD I say.
These are my predicitons:

Polaris (10 and 11) - May / July - Expect at least R9 390 X class performance cards for cheap. (Much like the HD2900 XT to the HD 3870 transition) R9 470 and R9 480 parts.
Vega - ( September 2016 to January 2017) I expect these to be the R9 490 X cards.

Or you could tier these one up being Polaris 11/10 the 480 and 490 class of cards and Vega be the Fiji successor.
 
Joined
Dec 3, 2009
Messages
1,205 (0.41/day)
Likes
364
Location
The Netherlands
System Name PC || Laptop
Processor AMD FX-8350 @ 4,8GHz || i7 6700HQ
Motherboard Asus Sabertooth 990FX || default
Cooling Corsair H100i Push-Pull|| Stock
Memory Samsung 30nm 4x4GB @ 1930MHz|| 2x8GB LVDDR3 @ 1600MHz
Video Card(s) AMD R9 290 @ 1060MHz / 1600MHz || Nvidia Geforce 960M
Display(s) Samsung S23A750D 120Hz + others|| 15,6" 1080P
Case Corsair 500R Fans Galore ||
Audio Device(s) Asus Xonar DX/XD ||
Power Supply Corsair AX860i || 120Watt 19V OEM adaptor
Software Win 10 64bit || Win 10 64bit
#4
Going by what Raja told Ryan from PCPer, I expect RTG to use two chips (on an interposer maybe?) on their top Vega part.
 
Joined
Sep 6, 2013
Messages
1,185 (0.76/day)
Likes
572
Location
Athens, Greece
System Name 3 systems: Gaming/Internet/HTPC
Processor 1075T @ 3.7GHz, undervolted @ 1.32V /1055T @ 3.5GHz, undervolted @ 1.27V/A6 5400K
Motherboard Gigabyte GA-990XA-UD3/ASRock 970 Extreme3 R2.0/ASUS FM2+
Cooling CoolerMaster TX3 (2fans)/Zalman CNPS9500 (Nvidia version)/CoolerMaster
Memory 16GB Adata 2133MHz/16GB DDR3 Kingston 2400MHz/8GB Corsair 1600MHz
Video Card(s) ASUS HD7850 2GB /Gigabyte GT620/FM2 iGPU
Storage OCZ ARC 100 + HDDs/Samsung 840 EVO + HDDs/OCZ Agility 3
Display(s) Samsung LE32D550 32'' TV(2 systems connected)/LG 42''
Case Sharkoon Rebel 12/Sharkoon Rebel 9/Xigmatek Midguard
Audio Device(s) onboard
Power Supply 850W Chieftec/560W Cheieftec/400W
Mouse CoolerMaster/Rapoo/Logitech
Keyboard CoolerMaster/Microsoft/Logitech
Software Windows 10 64bit TP + Windows 7 64bit + Windows XP (triple boot,stupidity all over the drives)
#5
I think I was expecting this. Who knows how many years companies will have to stay at 14nm. Who knows how much more expensive will be the transition to 10nm.

So they will give a 15-25% performance improvement, mostly from higher frequencies and architectural changes, for 3/4 of the power consumption thanks to the 14nm/16nm process, compared to today's models and that's it for this summer.

Add to that GDDR5/X and not HBM, the rumors that Polaris remains a feature level 12_0 card and Pascal still doesn't know what Async compute is, and we already start to question if this summer's models are the cards we are waiting for and not those that will come latter.
 
Joined
Oct 30, 2008
Messages
1,529 (0.46/day)
Likes
376
System Name Lailalo / Edelweiss
Processor FX 8320 @ 4.5Ghz / i7 3610QM @2.3-3.2Ghz
Motherboard ASrock 990FX Extreme 4 / Lenovo Y580
Cooling Cooler Master Hyper 212 Plus / Big hunk of copper
Memory 16GB Samsung 30nm DDR3 1600+ / 8GB Hyundai DDR3 1600
Video Card(s) XFX R9 390 / GTX 660M 2GB
Storage Seagate 3TB/1TB + OCZ Synapse 64GB SSD Cache / Western Digital 1TB 7200RPM
Display(s) LG Ultrawide 29in @ 2560x1080 / Lenovo 15.6 @ 1920x1080
Case Coolermaster Storm Sniper / Lenovo Y580
Audio Device(s) Asus Xonar DG / Whatever Lenovo used
Power Supply Antec Truepower Blue 750W + Thermaltake 5.25in 250W / Big Power Brick
Software Windows 10 Pro / Windows 10 Home
#6
Anyone not seeing the elephant in the room? Polaris releases late 2016, generously maybe by autumn/fall. A few months later, new architecture is released again. Makes zero sense unless they're doing a reverse Nvidia. Vega becomes the Titan style part, with Polaris 10 & 11 being the x80 and x70 parts.

FUD I say.
Probably because most likely Polaris is just Fiji 2.0. Fury with a die shrink. Could be the entire Polaris line will just be variations of Fury with different stream processor or memory counts.

Things really don't seem to get interesting till Vega. The last time AMD hyped up "performance per watt" we got Fury and were underwhelmed. 2.5x means squat. It's just more of that hype train trying to make Polaris look real great when its Vega folks should be more keen on.

AMD needs something to get them through till next year. Polaris will likely do. I just don't feel so bad about having to jump on a 390 before I planned to spend.
 
Joined
Apr 8, 2008
Messages
127 (0.04/day)
Likes
15
#7
I think this relates also with the delay of HBM2, as both AMD and NV are choosing GDDR5X for their next gen. high-end... while in the same time, AMD thinks that HBM2 holds a great potential that they're eager to release a product with it as soon as possible...
 
Joined
Feb 8, 2012
Messages
2,678 (1.25/day)
Likes
1,914
Location
Zagreb, Croatia
System Name Windows 7 64-bit Core i5 3570K
Processor Intel Core i5 3570K @ 4.2 GHz, 1.26 V
Motherboard Gigabyte GA-Z77MX-D3H
Cooling Scythe Katana 4
Memory 4 x 4 GB G-Skill Sniper DDR3 @ 1600 MHz
Video Card(s) Gainward NVIDIA GeForce GTX 970 Phantom
Storage Western Digital Caviar Blue 1 TB, Seagate Baracuda 1 TB
Display(s) Dell P2414H
Case CoolerMaster Silencio 550
Audio Device(s) VIA HD Audio
Power Supply Corsair TX v2 650W
Mouse Steelseries Sensei
Keyboard CM Storm Quickfire Pro, Cherry MX Reds
Software MS Windows 7 Enterprise 64-bit SP1
#8
Ah, R&D Manager spilling out company secrets on his LinkedIn profile ... I wish him good luck on the job market
 
Joined
Jul 18, 2007
Messages
2,425 (0.64/day)
Likes
645
System Name panda
Processor 6700k
Motherboard sabretooth s
Cooling raystorm block<black ice stealth 240 rad<ek dcc 18w 140 xres
Memory 32gb ripjaw v
Video Card(s) 290x gamer<ntzx g10<antec 920
Storage 950 pro 250gb boot 850 evo pr0n
Display(s) QX2710LED@110hz lg 27ud68p
Case 540 Air
Audio Device(s) nope
Power Supply 750w superflower
Mouse g502
Keyboard shine 3 with grey, black and red caps
Software win 10
Benchmark Scores http://hwbot.org/user/marsey99/
#9
Vega becomes the Titan style part, with Polaris 10 & 11 being the x80 and x70 parts.

seems vega_num is their new internal code naming scheme.

but i think they want to switch to a more nvidia style too with a tier of card between the gamers cards and firepro too.

as for greenland next year, i doubt it unless polaris fails.
 

FordGT90Concept

"I go fast!1!11!1!"
Joined
Oct 13, 2008
Messages
20,922 (6.24/day)
Likes
10,023
Location
IA, USA
System Name BY-2015
Processor Intel Core i7-6700K (4 x 4.00 GHz) w/ HT and Turbo on
Motherboard MSI Z170A GAMING M7
Cooling Scythe Kotetsu
Memory 2 x Kingston HyperX DDR4-2133 8 GiB
Video Card(s) PowerColor PCS+ 390 8 GiB DVI + HDMI
Storage Crucial MX300 275 GB, Seagate 6 TB 7200 RPM
Display(s) Samsung SyncMaster T240 24" LCD (1920x1200 HDMI) + Samsung SyncMaster 906BW 19" LCD (1440x900 DVI)
Case Coolermaster HAF 932 w/ USB 3.0 5.25" bay
Audio Device(s) Realtek Onboard, Micca OriGen+
Power Supply Enermax Platimax 850w
Mouse SteelSeries Sensei RAW
Keyboard Tesoro Excalibur
Software Windows 10 Pro 64-bit
Benchmark Scores Faster than the tortoise; slower than the hare.
#10
4096 is pretty disappointing for 14nm. The only incentive you'd have to buy Vega10 would be power consumption compared to Fiji (and minor stuff like DP 1.3, HDMI 2.0a, D3D 12_1 feature level, etc.). They better have something more potent in the works (and not dual GPU) if they want to compete with NVIDIA's top cards. Vega10 should be a upper mid-range card.
 
Joined
Apr 30, 2011
Messages
809 (0.33/day)
Likes
751
Location
Greece
Processor AMD FX-8350 4GHz@1.3V
Motherboard Gigabyte GA-970A UD3 Rev3.0
Cooling Zalman CNPS5X Performa
Memory 2*4GB Patriot Venom RED DDR3 1600MHz CL9
Video Card(s) Sapphire 7950 3GB
Storage Sandisk SSD 120GB, 2 Samsung F1 & F3 (1TB)
Display(s) LG IPS235
Case Zalman Neo Z9 Black
Audio Device(s) Via 7.1 onboard
Power Supply OCZ Z550
Mouse Zalman ZM-M401R
Keyboard Logitech K120
Software Win 7 sp1 64bit
Benchmark Scores CB R15 64bit: single core 95p, multicore 626p, OpenGL 91.61 WPrime 1.55 (8 cores): 9.0 secs
#11
4096 is pretty disappointing for 14nm. The only incentive you'd have to buy Vega10 would be power consumption compared to Fiji (and minor stuff like DP 1.3, HDMI 2.0a, D3D 12_1 feature level, etc.). They better have something more potent in the works (and not dual GPU) if they want to compete with NVIDIA's top cards. Vega10 should be a upper mid-range card.
Vega 11 (or 9 if reversed naming) anyone? ;)
 
Joined
Jan 23, 2011
Messages
304 (0.12/day)
Likes
96
Location
St. Louis, MO
System Name Alienware 17R4
Processor Intel Core i7-6820HK @ 3.6 Ghz
Motherboard Alienware 17R4
Cooling Laptop Cooling
Memory 32GB Crucial Ballistix LT DDR4-2400
Video Card(s) Nvidia Geforce 1070 GTX (Internal)/ASUS Geforce GTX 1080 Ti (in AMP)
Storage 500GB RD400 RevoDrive m2/ 1TB Cruical BX200
Display(s) ASUS Swift PG279
Case Alienware 17R4
Audio Device(s) Sennhesier GSX1000
Power Supply 330W Dell PSU
Mouse Razer Mamba TE Chroma
Keyboard Razer Blackwidow Chroma
#12
Considering the games coming down the pipe and how the porting to PC has become the norm in some cases.... if you haven't bought a new GPU in the last 3 years, anything from the Fury line or the next two AMD lines should be more than enough to hold you over for awhile.

Same on the NV side, if you have a Maxwell or a Pascal and only at 1080p possibly 1440p, you should be good for awhile. I think GPUs are finally falling in line with CPUs, they hitting that size wall and are more about cutting down power/energy efficiency. Sure a die shrink will help performance but they are only going to get so fast.

Developers are lazy and don't really optimize their code. Hence the ports we have been getting on PC lately. They optimize for PS4 and XBone because the hardware is static. With PC they have all the different configurations and unless either side does something to get developers off their ass and actually code for the PC hardware, it is going to be the same dog and pony show it has been for the past 6-8 years.

We've had pretty awesome hardware for awhile now and it just isn't utilized properly.

For some of us we are still on first gen I7's and FX8350's. Unless you synthetic benchmark all day, you'd be hard pressed for a difference. Hardware is starting to get to a limit and finally all this bloat and laziness with developers is going to bite them in their ass.

Now I know some studios do much better than others with their optimization and getting their products to run on various platforms but, a fair majority don't.


I am not too worried though, my 2x Nanos should be more than plenty for awhile.
 
Joined
Jun 10, 2014
Messages
763 (0.59/day)
Likes
325
#13
With the top chip in Vega/"5th generation GCN" holding 4096 shader processors (the same as Fiji), let's at least hope that it brings greater architectural changes than Polaris, which AMD themselves considers a minor change (except for the shrink of course).

It may be a smart change for AMD to target the upper mid range rather than the high-end, after the Fiji blunder where they spent all the resources on a high-end product that didn't sell well. The $300-550 market is after all the most profitable market, and this should make AMD able to cover most of the market share they can cover with their limited resources.
 
Joined
Jun 10, 2014
Messages
763 (0.59/day)
Likes
325
#14
Considering the games coming down the pipe and how the porting to PC has become the norm in some cases.... if you haven't bought a new GPU in the last 3 years, anything from the Fury line or the next two AMD lines should be more than enough to hold you over for awhile.

Same on the NV side, if you have a Maxwell or a Pascal and only at 1080p possibly 1440p, you should be good for awhile. I think GPUs are finally falling in line with CPUs, they hitting that size wall and are more about cutting down power/energy efficiency. Sure a die shrink will help performance but they are only going to get so fast.
Unlike CPUs, GPUs can continue to scale efficiently provided we can put more shader cores on the dies, so for a while GPUs have benefited a lot from shrinks while CPUs have really not improved a lot since Sandy-Bridge. But as we all know, the shrinks are fewer and farther between. We might be looking at two shrinks in the next decade or so, and the benefits from the shrinks will also decline.

But in terms of demand, the demand is actually increasing at a higher rate than in the last ten years, since gamers now want higher resolutions and higher frame rates at the same time. And we are still not at the point where GPUs are "powerful enough" so game developers can achieve everything they want, and we can expect performance requirements to continue to increase for new games.

The jump from 28nm to 14/16nm is actually "two steps", except for the interconnect which is still at 20nm. So Pascal is probably going to be the largest performance gain we have seen for a long time, and we are probably not going to see a similar increase for Volta, and post-Volta. Currently a single GTX 980 Ti is still not "powerful enough" for 4K gaming, and is not close to 60 FPS in all games at stock speed. And for those who want higher frame rates, even though GTX 980 Ti is OK for 1440p, it's not enough to push 120-144 FPS in all games at stock speed. With Pascal probably increasing the gaming performance by >60%, we will still not be at 4K 120 FPS. If Volta(2018) is not going to be another shrink, then we might get as little as 20% more performance, which is not going to keep up with the demand.

Developers are lazy and don't really optimize their code. Hence the ports we have been getting on PC lately. They optimize for PS4 and XBone because the hardware is static. With PC they have all the different configurations and unless either side does something to get developers off their ass and actually code for the PC hardware, it is going to be the same dog and pony show it has been for the past 6-8 years.

We've had pretty awesome hardware for awhile now and it just isn't utilized properly.
You are touching a very important subject. Game developers have gotten used to performance leaps every two years or so, so by the time a game is released they expect people to by more powerful hardware than it was developed on. We all know that performance gains in hardware is going to decrease over time, so writing good code is going to become increasingly important.

The gaming consoles are a big problem, which uses outdated low-end hardware. And as long as developers keep making games for these machines and porting them to PCs by cranking up the model details, they are going to continue to suck. The current API call mania (Direct3D 12/Vulkan/etc.) is not going to help the situation. Every developer knows that batching the draws is the only efficient way to render, and when doing efficient batching the API overhead is low anyway.

Game engines are using way too much abstraction to use GPUs efficiently. If the API overhead is a problem for a game, then the engine CPU overhead is going to be even larger. Doing all kinds of GPU manipulation through thousands of API calls is a step backwards. Scaling with API calls is not going to work well with the raw performance of Pascal, Volta, post-Volta and so on.
 
Joined
Nov 4, 2005
Messages
9,950 (2.25/day)
Likes
2,309
System Name MoFo 2
Processor AMD PhenomII 1100T @ 4.2Ghz
Motherboard Asus Crosshair IV
Cooling Swiftec 655 pump, Apogee GT,, MCR360mm Rad, 1/2 loop.
Memory 8GB DDR3-2133 @ 1900 8.9.9.24 1T
Video Card(s) HD7970 1250/1750
Storage Agility 3 SSD 6TB RAID 0 on RAID Card
Display(s) 46" 1080P Toshiba LCD
Case Rosewill R6A34-BK modded (thanks to MKmods)
Audio Device(s) ATI HDMI
Power Supply 750W PC Power & Cooling modded (thanks to MKmods)
Software A lot.
Benchmark Scores Its fast. Enough.
#15
More interested in the SOC/IP side of the conment, perhaps we will see the first hardware driver acceleration with compute and X86-64 core on a GPU? The time line fits, one or two HMA Zen cores coupled to a GPU die sharing resources? All the performance with no driver issues from different hardware configs, a simple scalable architecture.

Perhaps I'm all wrong here, but what else does SOC mean?
 

AsRock

TPU addict
Joined
Jun 23, 2007
Messages
15,316 (4.00/day)
Likes
4,724
Location
US
Processor 2500k \ 3770k
Motherboard ASRock Z68 \ Z77
Memory Samsung low profile 1600
Video Card(s) XFX 6770 \ XFX R9 290X
Storage Intel 80Gb (SATA2) WD 250Gb \ Team SSD+Samsung Evo 250Gb+500Gb+ 2xCorsair Force+WD250GbHDD
Display(s) Samsung 1080P \ Toshiba HDTV 1080P
Case HTPC400 \ Thermaltake Armor case ( original ), With Zalman fan controller ( wattage usage ).
Audio Device(s) Yamaha RX-V475 \ Marantz SR5008 Tannoy Mercury MKII Paradigm 5SE + Tannoy Mercury F4
Power Supply PC&Power 750w \ Seasonic 750w MKII
Mouse MS intelimouse \ Logitech G700s + Steelseries Sensei wireless
Keyboard Logitech K120 \ ROCCAT MK Pro ( modded amber leds )
Benchmark Scores Meh benchmarks.
#16
System\System on chip ?, but a few things but that's what i first think when the term is used
 
Joined
Oct 22, 2014
Messages
5,294 (4.60/day)
Likes
3,225
Location
Sunshine Coast
System Name Black Box
Processor Intel i5 3570K @ 3.8GHz
Motherboard Asrock Z77 Extreme 4 ATX
Cooling Antec C400
Memory G. Skill 8Gb (2x4Gb) 2133Mhz
Video Card(s) iGPU
Storage Sandisk X 400 256Gb
Case Antec Sonata (don't laugh)
Audio Device(s) No need.
Software W7 Home Premium 64 bit
#17
Silicon On Chip?
 
Joined
Feb 8, 2012
Messages
2,678 (1.25/day)
Likes
1,914
Location
Zagreb, Croatia
System Name Windows 7 64-bit Core i5 3570K
Processor Intel Core i5 3570K @ 4.2 GHz, 1.26 V
Motherboard Gigabyte GA-Z77MX-D3H
Cooling Scythe Katana 4
Memory 4 x 4 GB G-Skill Sniper DDR3 @ 1600 MHz
Video Card(s) Gainward NVIDIA GeForce GTX 970 Phantom
Storage Western Digital Caviar Blue 1 TB, Seagate Baracuda 1 TB
Display(s) Dell P2414H
Case CoolerMaster Silencio 550
Audio Device(s) VIA HD Audio
Power Supply Corsair TX v2 650W
Mouse Steelseries Sensei
Keyboard CM Storm Quickfire Pro, Cherry MX Reds
Software MS Windows 7 Enterprise 64-bit SP1
#18
Game engines are using way too much abstraction to use GPUs efficiently.
I'd say it's rather that game engines are using abstractions that allow very efficient use of GPUs if you know how to prepare content for it.
Modern engines batch draw calls automatically if different surfaces share textures, materials or shaders. When designing optimal art for 3d games it's all about reusing stuff while making it look like you are not reusing stuff.
Optimizing on a shader level is done one time for all eternity ... essentially there is one optimal "physical" based lighting shader all games use these days with diffuse, gloss/specular, emission, occlusion, normal and displacement textures (with additional detail diffuse+normal textures on top that are visible when close up) that allows sky based global illumination. Very little optimization room left there.
All optimization on cpu side is basically how to feed gpu command queues while minimizing number of context switching on gpu.
The hidden part that can make every game look unoptimized is occlusion culling algorithm which importance is often wrongly underestimated. Too many engines are used in a way they unnecessarily draw occluded objects.
The real problem starts when devs push consoles to the limits, while very much relying on low api overhead and low latencies heterogeneous memory design allows, only to reach locked 30 fps ... then do a straight port on pci-e bus induced latencies and higher overhead api. It may be feasible for jaguar core in ps4 to directly write something in video ram every frame, doing that over pci-e on pc would introduce extra latency.
Once you start using benefits of hsa on consoles, you get less scalable port to pc simply because of the modular nature of the pc.
The opposite way would be: develop optimally for pc, then leverage use of hsa on consoles to get acceptable performance... but I'm digressing and borderline rambling
It's system on chip, every soc is asic but not every asic is soc.
 
Joined
Jun 10, 2014
Messages
763 (0.59/day)
Likes
325
#19
I'd say it's rather that game engines are using abstractions that allow very efficient use of GPUs if you know how to prepare content for it.
Modern engines batch draw calls automatically if different surfaces share textures, materials or shaders. When designing optimal art for 3d games it's all about reusing stuff while making it look like you are not reusing stuff.
No, when game engines adds an abstraction layer above the actual API and creates a structure which calls upon each object to render itself we end up with the opposite of batching. As you can see with new games like Ashes of the Singularity keep bragging about the amount of API calls they are able to push through, which is evidence of inefficient coding.

If you try to render a bunch of meshes in a single API call the GPU works way more efficient than if you do it by thousands of small API calls. Even an old GTX 680 is able to render millions of polygons at the screen at a high frame rate, but no game is pushing through geometry at that level due to inefficient usage of the GPU.

All optimization on cpu side is basically how to feed gpu command queues while minimizing number of context switching on gpu.
Well, not quite. The GPU itself is way better at scheduling it's GPU threads/batches and even out the load, way better than an infinite powerful CPU could ever do.

The hidden part that can make every game look unoptimized is occlusion culling algorithm which importance is often wrongly underestimated. Too many engines are used in a way they unnecessarily draw occluded objects.
The GPUs themselves are to a large extent able to automatically cull a lot, that was actually one of the hardware improvements between GF100 and GF110.

Still both vertex shaders and compute shaders can be utilized for efficient culling. It's actually way more efficient to do the fine detailed culling on the GPU in the shader, rather than calculating it in the CPU and passing each part of a mesh as separate API calls.
 
Joined
Nov 4, 2005
Messages
9,950 (2.25/day)
Likes
2,309
System Name MoFo 2
Processor AMD PhenomII 1100T @ 4.2Ghz
Motherboard Asus Crosshair IV
Cooling Swiftec 655 pump, Apogee GT,, MCR360mm Rad, 1/2 loop.
Memory 8GB DDR3-2133 @ 1900 8.9.9.24 1T
Video Card(s) HD7970 1250/1750
Storage Agility 3 SSD 6TB RAID 0 on RAID Card
Display(s) 46" 1080P Toshiba LCD
Case Rosewill R6A34-BK modded (thanks to MKmods)
Audio Device(s) ATI HDMI
Power Supply 750W PC Power & Cooling modded (thanks to MKmods)
Software A lot.
Benchmark Scores Its fast. Enough.
#20
As you can see with new games like Ashes of the Singularity keep bragging about the amount of API calls they are able to push through, which is evidence of inefficient coding.

If you try to render a bunch of meshes in a single API call the GPU works way more efficient than if you do it by thousands of small API calls. Even an old GTX 680 is able to render millions of polygons at the screen at a high frame rate, but no game is pushing through geometry at that level due to inefficient usage of the GPU.
Its not the number of API calls as much as the use of existing objects (instancing) being called on again without the call needing to ask the CPU for geometry information and cull status (batch process) to render the object, previous (to DX12/Mantle/Vulkan) the basic scene geometry was done on the CPU and the majority of culls and texture data was handled and shuffled to the GPU along with the scene render instructions from the CPU, which is exactly why a faster processor could render more FPS, and why benchmarking a CPU in smaller pixel counts (lower resolution) was a normally accepted way to do things, but modern GPU's are aware of texture location through drivers and hardware and can hold more textures in memory (onus on developer of the 3D application to have the correct textures in DX12) instead of drivers that have to be optimized for every game, making drivers 10% actual software, and 90% game optimization data in all versions prior.

You are mistaking hardware efficiency and muscle for poor coding, when its actually showing off the hardware and software improvements.
 
Joined
Oct 9, 2009
Messages
675 (0.23/day)
Likes
656
Location
Finland
System Name :P~
Processor Intel Core i7-5930K (ES)
Motherboard Asus Rampage V Extreme/3.1
Cooling Phanteks PH-TC14PE
Memory 32GB Corsair Vengeance LPX 2400 MHz
Video Card(s) Asus GTX 1080 Strix
Storage 400GB Intel 750 PCI-E SSD, 512GB Crucial MX100 SSD, 3TB WD RED HDD
Display(s) QNIX QX2710LED OC @ 96 Hz 27"
Case Corsair Obsidian 750D
Audio Device(s) Audioquest Dragon Red + Sennheiser HD 650
Power Supply Corsair HX1000i + Cablemod sleeved cables kit
Mouse Logitech G500s
Keyboard Logitech Ultra X Flat Premium
Software Windows 10 64-bit
#21
Ah, R&D Manager spilling out company secrets on his LinkedIn profile ... I wish him good luck on the job market
There has been so many of these lately that I call the bluff and say some of it might be planted there in purpose to create hype. By this P&R team manager. :p
 
Joined
Oct 15, 2010
Messages
534 (0.20/day)
Likes
101
Processor Intel Core I5 2500K @ 4.5 Ghz
Motherboard ASUS P8Z68-V
Cooling Zalman CNPS 9900Max
Memory Corsair Vengeance & Kingston Black 4x4GB DDR3 1600 Mhz
Video Card(s) SAPPHIRE Dual-X OC R9 280X 3GB
Storage OCZ Vertex 4 256GB Sata 3 / Ultrastar 2TB / Seagate 2TB
Display(s) Samsung UN32EH5300 / LG 39LN5700
Case Lian Li PC-K63
Power Supply OCZ ZT 750W
Software Windows 7 Ultimate SP1 64 Bits
#22
So they will give a 15-25% performance improvement, mostly from higher frequencies and architectural changes, for 3/4 of the power consumption thanks to the 14nm/16nm process, compared to today's models and that's it for this summer.
I don't think anything below a 50% perf increase is worthwhile.