Equipped with 2 pure copper heat-pipes and 90 mm unique-blade fan with 3D active functionality, and heat pipes direct touch technology, the cooler can dissipate heat effectively from the GPU while keeping the fan at lower speed and noise.
Cooling System-Fan
Unique Blade Fan Design
The airflow is spilt by the triangle fan edge, and guided smoothly through the 3D stripe curve on the fan surface, effectively enhancing the air flow by 23% over traditional fans.
3D Active Fan
The semi-passive fans will remain off when the GPU is under a set loading or temperature for low power gaming. It allows gamers to enjoy gameplay in complete silence when the system is running light or idle.
Cooling System- Heat Pipes
Composite heat-pipes
The composite heat-pipes combines both thermal conductivity and phase transition for efficiently managing the transfer of heat between two solid interfaces which increases 29% of cooling capacity.
Heat Pipe Direct Touch
The pure copper heat pipes are shaped to maximize the direct contact area to the GPU.
Compact Card Size
Measuring merely 17cm in length, the card, perfectly compatible for building most of the PCs from ATX to mini-ITX.
One-click Super overclocking
With a simple click on XTREME engine utility, gamers can easily tune the card to meet their various gaming requirements without any overclocking knowledge, while saving the hassle of manual adjustment.
Ultra Durable Graphic Components
Engineered with the highest-grade chokes and capacitors, this graphic card delivers outstanding performance and durable system lifespan.
Other Characteristics:
CUDA: Yes
CUDA cores: 1152
Colour of product: Black
Cooling type: Active
DVI-D ports quantity: 2
Data transfer rate: 8 Gbit/s
Depth: 169 mm
DirectX version: 12.0
Discrete graphics adapter memory: 6 GB
DisplayPort version: 1.4
DisplayPorts quantity: 1
Dual Link DVI: Yes
Form factor: ATX
Graphics adapter memory type: GDDR5
Graphics processor: GeForce GTX 1060
Graphics processor family: NVIDIA
HDMI ports quantity: 1
HDMI version: 2.0b
Height: 37 mm
Interface type: PCI Express x16 3.0
Maximum digital resolution: 7680 x 4320 pixels
Maximum displays per videocard: 4
Maximum resolution: 7680 x 4320 pixels
Memory bandwidth (max): 192 GB/s
Memory bus: 192 bit
Memory clock speed: 8008 MHz
Minimum system power supply: 400 W
NVIDIA G-SYNC: Yes
NVIDIA GameWorks VR: Yes
Number of slots: 2
OpenGL version: 4.5
Power consumption (typical): 120 W
Processor boost clock speed: 1708 MHz
Processor frequency: 1506 MHz
Supplementary power connectors: 1x 6-pin
Width: 131 mm
Windows operating systems supported: Yes
Likely to be an Error
- Axeboy
Latest comments (52)
rev6
15 Dec 16#52
All this talk about memory is killing it :smile:
smellyonion
15 Dec 16#51
You
"You also seem to fail to understand how memory allocation and bandwidth works. PS4 Pro still only has 5.5GB TOTAL usable memory,which has to be split video/system and it also has to share that bandwidth with the CPU and everything else. It still has major overheads." :smile:
I am actually shocked.
vulcanproject
15 Dec 16#50
You stopped understanding you mean, AND you cannot read.
You don't even know what is meant when I said video/system resources sharing. The console has 5.5GB of memory available for developers, and no, that doesn't mean it has 5.5GB just for video memory. Because you still have to run system processes related to the game in that space such as the game assets......
I wasn't talking about OS or UI overheads. The developers get 5.5GB for their whole game. That's it. They have to split that resource. Maybe they can use 3GB or so for video, and 2.5GB for game assets. System processes for their game etc. Developers can alter that ratio a bit each way, but not massively. Which is partly why PS4 Pro is unsuited to native 4K. It barely has enough memory for one thing.
It is honestly mind boggling how little you know about how any of this works and yet you continue to amaze with your ignorance.
Aaaaaand you're STILL talking utter poop about a console in a deal for a PC video card. Honestly you should be banned and probably reported for doing this in this thread.
cookied
14 Dec 16#49
now girls put your handbags away... :smiley: can anyone tell me is the 6gb worth the £50 or so over the 3gb gtx 1060 cards ?
smellyonion
14 Dec 16#48
I stopped reading when I read that pro has access to 5.5 gb, you said it's shared by the system! I said it's purely for games. Learn to read. Scroll back. You have absolutely no clue and demonstrated your inability to read... You can't comprehend words.
vulcanproject
14 Dec 16#47
Direct3D is inefficient? Says who? You? I assume you can clear up why exactly if you are qualified to judge?
Don't talk nonsense. Xbox One runs on a superset of Direct3D, it's obviously not that inefficient on PC either because a GTX1060 STILL beats PS4 Pro on games at the same resolutions it can do more frames or higher quality effects.
Peak theoretical FLOPS is not a very useful measurement of graphics performance for a GPU. Especially compared between different architectures as you have been using it. I already said this. You fail to understand even now!!! It's a measurement of compute performance, NOT overall graphics rendering performance.
PS4 Pro does only have access to 5.5GB of memory for developers. Suck it and see. Unlike you I actually read and know about hardware. Mark Cerny (the guy who oversaw the hardware) outlined this fact himself. Unified memory IS shared. Get a dictionary and look up the word unified. It ISN'T the optimal way for the most powerful systems because the CPU is constrained to poor latency memory, slower memory access. It's just cheaper and easier than having a separate chip with it's own dedicated memory like in a PC which is the best way.
This is hilarious. You're seriously out of your depth here. You don't know about anything you are talking about.
Clueless! Bet you regret talking crap and being called out on don't you? Learn your lesson and stop doing it.
smellyonion
14 Dec 16#46
Nope, what I do not understand is your lack of logic.....
Youve been incorrect on several occasions and you keep going on.
I used Teraflops as my metric, you use 2 x desktop GPUs on a inefficient desktop api that have nothing to do with consoles vs desktop.
As I keep saying, consoles will use those flops, the pc never will. Amd cards are under performing on desktops despite having more objective performance..simples.
I know it sucks to think that your 200 pound graphics will perform very similar to a games console and you will make horrendous claims like ps4 hae access to 5,5 gb for the games and system or that 8gb gddr5 was the cheap choice :smile: in 2013.....unified is NOT shared memory, don't confuse the two. Unified is better for gaming unfortunately... Have a read, do some research and talk when you know what you're talking about https://en.m.wikipedia.org/wiki/Heterogeneous_System_Architecture.
Use some logic and have a deep think.
vulcanproject
14 Dec 16#45
I know you don't understand or you wouldn't have been disagreeing with proven facts or talking nonsense :stuck_out_tongue:
smellyonion
14 Dec 16#44
Disagree and that's the end of it. I didmt say it was the best the best way. Very cost efficient and it will run games similar to and better down the line. Everybody has built a PC and do not need a lecture on things that you think that I do not understand.
alltaken123
13 Dec 16#43
vulcanproject
13 Dec 16#42
No, YOU started the comparison between a console and a desktop card, which isn't relevant to this thread!!:laughing::laughing: The Last of Us isn't a PC game. Irrelevant.
You also don't know what you are talking about. AMD's desktop performance is fine...please stop talking utter nonsense.
Yes there is a reason why PS4/Xbone have gone AMD, and that is because AMD are the only company that can offer a complete SoC with an x86 based CPU and a modern gaming level GPU of the performance required. Nobody else can do that, and nobody else can build it for the price AMD bid.
Yes 8GB of GDDR5 is cheap- compared to having more complicated separate chips, extra buses to feed them, separate RAM modules, more space on the motherboard for it, better power supply because consumption would be higher, more expensive/advanced cooling design to support it all etc etc...
Once you decide to have an SoC with unified memory (even GDDR5) because it's cheaper, simpler, smaller and uses less power, then it comes out costing less and makes it easier to manufacture. NOT because it's the fastest way to build a gaming system.
It's still a compromise. It's still not the best way for maximum performance.
You know literally nothing about the computer industry. You are still talking about consoles in a thread about PC graphics cards. GTX1060 is STILL faster than PS4's Pro GPU.
Next?
smellyonion
13 Dec 16#41
Terraflops is raw performance. You are comparing a Nvidia card to an AMD card in a desktop with an api that underperforms on amd on desktops.
There's a reason all consoles have gone amd.
So wrong, oh well. I tried. 8gb gddr5 is cheap.... One of the most expensive components in the ps4. Like I said, you run last of us on a 7800GTX and get back to me.
You've been corrected on so many levels.
vulcanproject
13 Dec 16#40
Your initial comment was that PS4 Pro's GPU was as fast as this. It isn't. Not in raw graphics performance. Not even after we take into consideration something you only brought up when I corrected you the first time. Nice try though.
Even with better coding to the metal, PS4 Pro is STILL not as fast as this.
Unified memory in consoles exists because it's simple and cheap. Not because it is the best solution for ultimate performance. Not when you have to apportion it, share the memory AND the bandwidth between all aspects of the system. The idea that a single unified bank of memory is anything but a compromise to save money (and is somehow better than large split pools dedicated to their tasks) is hilarious.
Physics calculations done only on the CPU are so 2005. Most games have their own methods for physics but modern APIs can do most of it on the GPU because it is much better for that kind of thing, as GPUs are already hugely parallelised. Besides this, PC CPUs are so preposterously faster than PS4's CPU that memory is neither the point nor the bottleneck.
You really don't know what you are talking about do you? Please stop. Staahhhppp!
smellyonion
13 Dec 16#39
You ignored my point about api. You admit consoles are coded to the metal yet ignore raw performance when in reality, the pro will use those flops. It will be able to utilise it that's the end of it.
Unified memory is a huge advantage, both cpu and GPU and manipulate the same memory pool as the same time. Things like physics, that require the superior calculations of the CPU can be written over the same memory live.Something no pc can do. Yes it's worse for non gaming applications but it's a console and for gaming, it is a far superior set up than separate ddr3 and gddr5. No transfer time when the GPU needs cpu work like the pc. Also, 5.5gb, is the amount reserved, just for games...almost double this card. System is side lined.
vulcanproject
13 Dec 16#38
I think you need to understand that measuring a GPU's performance in single precision FLOPS when it's application in the actual case you stated is to do graphics rendering is entirely pointless.
Which is what I outlined to you. If you buy a console to do very specific compute based scientific calculations then you might have a point.
You don't.
You also seem to fail to understand how memory allocation and bandwidth works. PS4 Pro still only has 5.5GB TOTAL usable memory, which has to be split video/system and it also has to share that bandwidth with the CPU and everything else. It still has major overheads. Whereas a PC graphics card has all that memory and bandwidth entirely for video, and the CPU has all it's own memory bandwidth and a bunch load of it's own lower latency memory superior for CPU performance. You are so far out on these points I can't even be bothered to correct the rest of your errors.
This is the only thing you need to know though:
GTX1060 has superior graphics rendering performance to PS4 Pro's GPU, end of story. Besides, PC has a few more than about 10 games that you can actually play 'at larger resolutions.' unlike PS4 Pro. It has nearly everything from the past 15 years :stuck_out_tongue:
smellyonion
13 Dec 16#37
Power can have various ways of measuring it.
In this case, I chose Teraflops and thats the end of it.
Wiki definition:
In computing, FLOPS or flops (an acronym for FLoating-point Operations Per Second) is a measure of computer performance, useful in fields of scientific calculations that make heavy use of floating-point calculations. For such cases it is a more accurate measure than the generic instructions per second.
I brought up api because you referred to how the 1060 beats the 480 in desktop benchmarks yet has lower objective perfomance (FLOPS). Therefore, I was referring to how poorly optimized AMD cards to Direct X, PS4 does not use Direct X as its language so will not have this bottleneck that AMD cards have on desktops.
Then, we havent begun to tallk about 3gb GDDR5 @ 192 gbps vs the pros 8gb GDDR @ 218gbps, the 1060p wont scale at larger resolutions like the pro.
vulcanproject
12 Dec 16#36
Pro's GPU isn't equal to a GTX1060 anyway I already told you that.
You said that Pro had the same power of this GPU and then quoted teraflops to back it up, nothing about what API the two platforms use. So I corrected you by saying no, Pro's GPU is not as fast as a GTX1060. It isn't.
Coding to the metal on consoles always helps their performance but Pro still won't be faster than a GTX1060 in a decent PC build.
I'll repeat, this isn't a thread for you to talk about a games console which is essentially built as a toy and limited to that. This is a GPU for a PC which can be used to do a ridiculous array of extra things, and in 18 months when you feel like even more GPU performance you can take this out, sell it and trade up for something twice as fast again should you so wish.
Axeboy
12 Dec 16#35
Not that its confirmed exactly to be a 480 spec, but its still 30% lower clocks. 30% is a lot.
Not that it matters, ps4 pro vs PC isn't really the right comparison anyways :smiley:
smellyonion
12 Dec 16#34
I get your point, again you're adding cost, losing reesaleabilty without physical media. Your equivalent price pc will not perform better than the pro, simple as. The Ps3 featured something like a 7800GTX yet it pumped out God of war, uncharted, last of us. Impossible on a PC with that card. The efficiency gains are immense.
Again, the pro features a card similar to those currently at £200 on the market.
smellyonion
12 Dec 16#33
Pro GPU is basically a slightly underclocked 480 rx since the pro features 36 compute units. Obviously, with no dx api and unified gddr5 memory pool for efficiency gains, it will perform very well.
dudedude
12 Dec 16#32
blu-ray drive is a relic for Sony to sell overpriced games. PC uses Steam downloads. Wifi is not necessary but you can add it on the motherboard easily enough. Windows keys are around £5 on ebay.
You can build a much faster PC for the cost of a PS4 pro if you want to shop around. I5-2500k £50 on ebay for example.
dudedude
12 Dec 16#31
Right now the graphics market consists of:
RX 460 2GB £100 - terrible compared with the GTX 1050
RX 460 4GB £110 - the 4GB is not worth it compared with the extra speed of the 1050
GTX 1050 2GB £110
GTX 1050 TI (4GB) £130
RX 470 (4GB) £175 - even though it's not much slower than the RX480, the price difference is too small at the moment. It's been sold as low as £150 though, at which price it's a no-brainer.
RX 480 4GB £188 I tend to prefer the GTX 1060 3GB, because it's generally faster today, and lower power.
GTX 1060 (SE) 3GB £189
RX 480 8GB £200
GTX 1060 6GB £220
GTX 1070 £380
GTX 1080 £500
The PS4 is around the RX 470 level. Which means it's only mid-range with the likes of the GTX 1070, 1080, and soon 1080Ti far ahead.
smellyonion
12 Dec 16#30
No. Even with your underpricing. Please include your bluray drive, WiFi, Windows and it will still perform worse...
smellyonion
12 Dec 16#29
Errr, no.
Pro does not use direct x api. Gains will be better than any comparable card on the pro.
dudedude
12 Dec 16#28
Uh, what?
£150 for a comparable GPU to the one in the PS 4 Pro
The PS 4 Pro costs £350
So not comparable in price.
So:
£150 for the GPU
£20 for 4gb RAM
£40 for the 1tb drive
£60 for an AMD 5350 and motherboard
£50 for Rosewill case + 400W PSU
Total: £320.
What was your point exactly?
vulcanproject
12 Dec 16#27
I think £50 for that Jaguar CPU is generous.
No, you get a more powerful GPU for this. To put into a PC. Which will cost more than a PS4 Pro. Because it does ten times more things. It's a PC not a games console. This has nothing at all to do with games console. It's a PC graphics card to build a more powerful machine
Glad you managed to figure that out for yourself after obviously stumbling onto the wrong page.
vulcanproject
12 Dec 16#26
Teraflops is a measure of raw floating point performance. Maths basically. Not of overall graphics performance comparing two totally different GPU architectures, Nvidia Pascal v AMD Polaris in this case.
A GTX1060 generally beats a desktop Radeon RX 480 on overall graphics performance, to the tune of 10-15 percent. PS4 Pro's GPU is based on it, but significantly downclocked (about 80 percent of the clock speed) and therefore slower than a RX 480.
So there you go.
smellyonion
12 Dec 16#25
You just proved my point, here you get a GPU...not a cpu, no system ram, no tb hardrive.
smellyonion
12 Dec 16#24
Raw power. Terraflops. Have a look then get back to me.
smellyonion
10 Dec 161#19
This has the same power of the ps4 pro, interesting to think of it like this. Makes the pro amazing value for money.
vulcanproject to smellyonion
12 Dec 16#22
No it doesn't. Why do you think the pro's GPU is as fast as this?
dudedude to smellyonion
12 Dec 16#23
Not true.The PS 4 Pro's GPU is similar to a £150 Radeon RX 470. And the CPU is about £50 worth of AMD FX chip, it's a terrible CPU.
tomwoodhouse
11 Dec 16#21
OOS
Rhythmeister
10 Dec 16#20
Yet another dual slot supposedly mini ITX card unusable in the Shuttle SN78SH7 :disappointed:
CAL23
10 Dec 16#15
Product name says 3GB but description says 6GB. Which is it?
Axeboy to CAL23
10 Dec 16#18
Read my post above, likely the description is incorrect, likely to be 3gb
So not a good deal if it is
vulcanproject
10 Dec 16#17
SLI is rubbish though and it's always the games that most need it that don't work with it or don't work well.
Don't SLI until you get to high end and can't do better without multiple cards anyway, it's a waste of time and money.
wanderer
10 Dec 16#16
oos
Axeboy
10 Dec 16#14
Really depends on needs, the 6gb 1060 is about twice the performance of the 1050ti
Anyways, this isn't the 6gb, its the 3gb so don't buy it :smiley:
GwanGy
10 Dec 16#13
p this is good n all , but when you have the 1050ti going for slightly over half the price ?? (£120-130) Not quite in the same league I know, but £90 /£100 better? If you (could) sli two 1050ti you'd probably get better perf than 1x1060.. my 2$c
dreamager
10 Dec 16#12
Hmm doesn't Oculus need 2 HDMI?
QuickProfits
10 Dec 16#11
3GB Ram
Axeboy
10 Dec 162#9
This seems to be the 3gb version guys
The product code is usually the most reliable part of the spec: GV-N1060IX-3GD
That's the 3gb model. I think Amazon have made a spec mistake.
edit: the picture also seems to be 3gb, check the bottom left
Also, I don't think Gigabyte do a non-overclocked version of this card in 6gb format, only 3gb
donbarney to Axeboy
10 Dec 16#10
I think you are right mate, there is no 6gb version of this card that is not the oc version
I have the OC 6gb version, it is a very good card can push it quite far with an overclock and its not too loud
scholesy27
10 Dec 16#8
ok thank you i will keep that in mind
googleboogle
10 Dec 16#7
Currently the best performance for price. This and the 480
scholesy27
10 Dec 16#5
would this be worth the upgrade from a 4gb 680? or am i just better off saving of 1080?
MBeeching to scholesy27
10 Dec 16#6
It's much faster than a GTX 680, I would also consider the 1070 though it's double the price.
hotmep
10 Dec 16#3
This or an RX 480 8GB?
thelagmonster to hotmep
10 Dec 16#4
Personally I went 480 (and got for a lot less than this), but really just depends which you can get for the best price. This is still a good price.
booboy2
10 Dec 16#2
There is also this one which is oc for a few pounds more...
![Fresh Breakfast Meat Box (Using code) @ Campbells Meat [Haggis / Lorne Sausage / Bacon / Black Pudding & More]](/images/topics/208684_1.jpg)
![[Xbox One] Q. u. b. e: Director's Cut on Deals with Gold](/images/topics/208716_1.jpg)
Opening post
- Integrated with 6GB GDDR5 192bit memory
- 90mm cooler with 3D active fan
- One-click Super Overclocking
- 17cm Compact Card Size
Custom-designed Cooling System
Equipped with 2 pure copper heat-pipes and 90 mm unique-blade fan with 3D active functionality, and heat pipes direct touch technology, the cooler can dissipate heat effectively from the GPU while keeping the fan at lower speed and noise.
Cooling System-Fan
Unique Blade Fan Design
The airflow is spilt by the triangle fan edge, and guided smoothly through the 3D stripe curve on the fan surface, effectively enhancing the air flow by 23% over traditional fans.
3D Active Fan
The semi-passive fans will remain off when the GPU is under a set loading or temperature for low power gaming. It allows gamers to enjoy gameplay in complete silence when the system is running light or idle.
Cooling System- Heat Pipes
Composite heat-pipes
The composite heat-pipes combines both thermal conductivity and phase transition for efficiently managing the transfer of heat between two solid interfaces which increases 29% of cooling capacity.
Heat Pipe Direct Touch
The pure copper heat pipes are shaped to maximize the direct contact area to the GPU.
Compact Card Size
Measuring merely 17cm in length, the card, perfectly compatible for building most of the PCs from ATX to mini-ITX.
One-click Super overclocking
With a simple click on XTREME engine utility, gamers can easily tune the card to meet their various gaming requirements without any overclocking knowledge, while saving the hassle of manual adjustment.
Ultra Durable Graphic Components
Engineered with the highest-grade chokes and capacitors, this graphic card delivers outstanding performance and durable system lifespan.
Other Characteristics:
CUDA: Yes
CUDA cores: 1152
Colour of product: Black
Cooling type: Active
DVI-D ports quantity: 2
Data transfer rate: 8 Gbit/s
Depth: 169 mm
DirectX version: 12.0
Discrete graphics adapter memory: 6 GB
DisplayPort version: 1.4
DisplayPorts quantity: 1
Dual Link DVI: Yes
Form factor: ATX
Graphics adapter memory type: GDDR5
Graphics processor: GeForce GTX 1060
Graphics processor family: NVIDIA
HDMI ports quantity: 1
HDMI version: 2.0b
Height: 37 mm
Interface type: PCI Express x16 3.0
Maximum digital resolution: 7680 x 4320 pixels
Maximum displays per videocard: 4
Maximum resolution: 7680 x 4320 pixels
Memory bandwidth (max): 192 GB/s
Memory bus: 192 bit
Memory clock speed: 8008 MHz
Minimum system power supply: 400 W
NVIDIA G-SYNC: Yes
NVIDIA GameWorks VR: Yes
Number of slots: 2
OpenGL version: 4.5
Power consumption (typical): 120 W
Processor boost clock speed: 1708 MHz
Processor frequency: 1506 MHz
Supplementary power connectors: 1x 6-pin
Width: 131 mm
Windows operating systems supported: Yes
Likely to be an Error
- Axeboy
Latest comments (52)
"You also seem to fail to understand how memory allocation and bandwidth works. PS4 Pro still only has 5.5GB TOTAL usable memory,which has to be split video/system and it also has to share that bandwidth with the CPU and everything else. It still has major overheads." :smile:
I am actually shocked.
You don't even know what is meant when I said video/system resources sharing. The console has 5.5GB of memory available for developers, and no, that doesn't mean it has 5.5GB just for video memory. Because you still have to run system processes related to the game in that space such as the game assets......
I wasn't talking about OS or UI overheads. The developers get 5.5GB for their whole game. That's it. They have to split that resource. Maybe they can use 3GB or so for video, and 2.5GB for game assets. System processes for their game etc. Developers can alter that ratio a bit each way, but not massively. Which is partly why PS4 Pro is unsuited to native 4K. It barely has enough memory for one thing.
It is honestly mind boggling how little you know about how any of this works and yet you continue to amaze with your ignorance.
Aaaaaand you're STILL talking utter poop about a console in a deal for a PC video card. Honestly you should be banned and probably reported for doing this in this thread.
Don't talk nonsense. Xbox One runs on a superset of Direct3D, it's obviously not that inefficient on PC either because a GTX1060 STILL beats PS4 Pro on games at the same resolutions it can do more frames or higher quality effects.
Peak theoretical FLOPS is not a very useful measurement of graphics performance for a GPU. Especially compared between different architectures as you have been using it. I already said this. You fail to understand even now!!! It's a measurement of compute performance, NOT overall graphics rendering performance.
PS4 Pro does only have access to 5.5GB of memory for developers. Suck it and see. Unlike you I actually read and know about hardware. Mark Cerny (the guy who oversaw the hardware) outlined this fact himself. Unified memory IS shared. Get a dictionary and look up the word unified. It ISN'T the optimal way for the most powerful systems because the CPU is constrained to poor latency memory, slower memory access. It's just cheaper and easier than having a separate chip with it's own dedicated memory like in a PC which is the best way.
This is hilarious. You're seriously out of your depth here. You don't know about anything you are talking about.
Clueless! Bet you regret talking crap and being called out on don't you? Learn your lesson and stop doing it.
Youve been incorrect on several occasions and you keep going on.
I used Teraflops as my metric, you use 2 x desktop GPUs on a inefficient desktop api that have nothing to do with consoles vs desktop.
As I keep saying, consoles will use those flops, the pc never will. Amd cards are under performing on desktops despite having more objective performance..simples.
I know it sucks to think that your 200 pound graphics will perform very similar to a games console and you will make horrendous claims like ps4 hae access to 5,5 gb for the games and system or that 8gb gddr5 was the cheap choice :smile: in 2013.....unified is NOT shared memory, don't confuse the two. Unified is better for gaming unfortunately... Have a read, do some research and talk when you know what you're talking about https://en.m.wikipedia.org/wiki/Heterogeneous_System_Architecture.
Use some logic and have a deep think.
You also don't know what you are talking about. AMD's desktop performance is fine...please stop talking utter nonsense.
Yes there is a reason why PS4/Xbone have gone AMD, and that is because AMD are the only company that can offer a complete SoC with an x86 based CPU and a modern gaming level GPU of the performance required. Nobody else can do that, and nobody else can build it for the price AMD bid.
Yes 8GB of GDDR5 is cheap- compared to having more complicated separate chips, extra buses to feed them, separate RAM modules, more space on the motherboard for it, better power supply because consumption would be higher, more expensive/advanced cooling design to support it all etc etc...
Once you decide to have an SoC with unified memory (even GDDR5) because it's cheaper, simpler, smaller and uses less power, then it comes out costing less and makes it easier to manufacture. NOT because it's the fastest way to build a gaming system.
It's still a compromise. It's still not the best way for maximum performance.
You know literally nothing about the computer industry. You are still talking about consoles in a thread about PC graphics cards. GTX1060 is STILL faster than PS4's Pro GPU.
Next?
There's a reason all consoles have gone amd.
So wrong, oh well. I tried. 8gb gddr5 is cheap.... One of the most expensive components in the ps4. Like I said, you run last of us on a 7800GTX and get back to me.
You've been corrected on so many levels.
Even with better coding to the metal, PS4 Pro is STILL not as fast as this.
Unified memory in consoles exists because it's simple and cheap. Not because it is the best solution for ultimate performance. Not when you have to apportion it, share the memory AND the bandwidth between all aspects of the system. The idea that a single unified bank of memory is anything but a compromise to save money (and is somehow better than large split pools dedicated to their tasks) is hilarious.
Physics calculations done only on the CPU are so 2005. Most games have their own methods for physics but modern APIs can do most of it on the GPU because it is much better for that kind of thing, as GPUs are already hugely parallelised. Besides this, PC CPUs are so preposterously faster than PS4's CPU that memory is neither the point nor the bottleneck.
You really don't know what you are talking about do you? Please stop. Staahhhppp!
Unified memory is a huge advantage, both cpu and GPU and manipulate the same memory pool as the same time. Things like physics, that require the superior calculations of the CPU can be written over the same memory live.Something no pc can do. Yes it's worse for non gaming applications but it's a console and for gaming, it is a far superior set up than separate ddr3 and gddr5. No transfer time when the GPU needs cpu work like the pc. Also, 5.5gb, is the amount reserved, just for games...almost double this card. System is side lined.
Which is what I outlined to you. If you buy a console to do very specific compute based scientific calculations then you might have a point.
You don't.
You also seem to fail to understand how memory allocation and bandwidth works. PS4 Pro still only has 5.5GB TOTAL usable memory, which has to be split video/system and it also has to share that bandwidth with the CPU and everything else. It still has major overheads. Whereas a PC graphics card has all that memory and bandwidth entirely for video, and the CPU has all it's own memory bandwidth and a bunch load of it's own lower latency memory superior for CPU performance. You are so far out on these points I can't even be bothered to correct the rest of your errors.
This is the only thing you need to know though:
GTX1060 has superior graphics rendering performance to PS4 Pro's GPU, end of story. Besides, PC has a few more than about 10 games that you can actually play 'at larger resolutions.' unlike PS4 Pro. It has nearly everything from the past 15 years :stuck_out_tongue:
In this case, I chose Teraflops and thats the end of it.
Wiki definition:
In computing, FLOPS or flops (an acronym for FLoating-point Operations Per Second) is a measure of computer performance, useful in fields of scientific calculations that make heavy use of floating-point calculations. For such cases it is a more accurate measure than the generic instructions per second.
I brought up api because you referred to how the 1060 beats the 480 in desktop benchmarks yet has lower objective perfomance (FLOPS). Therefore, I was referring to how poorly optimized AMD cards to Direct X, PS4 does not use Direct X as its language so will not have this bottleneck that AMD cards have on desktops.
Then, we havent begun to tallk about 3gb GDDR5 @ 192 gbps vs the pros 8gb GDDR @ 218gbps, the 1060p wont scale at larger resolutions like the pro.
You said that Pro had the same power of this GPU and then quoted teraflops to back it up, nothing about what API the two platforms use. So I corrected you by saying no, Pro's GPU is not as fast as a GTX1060. It isn't.
Coding to the metal on consoles always helps their performance but Pro still won't be faster than a GTX1060 in a decent PC build.
I'll repeat, this isn't a thread for you to talk about a games console which is essentially built as a toy and limited to that. This is a GPU for a PC which can be used to do a ridiculous array of extra things, and in 18 months when you feel like even more GPU performance you can take this out, sell it and trade up for something twice as fast again should you so wish.
Not that it matters, ps4 pro vs PC isn't really the right comparison anyways :smiley:
Again, the pro features a card similar to those currently at £200 on the market.
You can build a much faster PC for the cost of a PS4 pro if you want to shop around. I5-2500k £50 on ebay for example.
RX 460 2GB £100 - terrible compared with the GTX 1050
RX 460 4GB £110 - the 4GB is not worth it compared with the extra speed of the 1050
GTX 1050 2GB £110
GTX 1050 TI (4GB) £130
RX 470 (4GB) £175 - even though it's not much slower than the RX480, the price difference is too small at the moment. It's been sold as low as £150 though, at which price it's a no-brainer.
RX 480 4GB £188 I tend to prefer the GTX 1060 3GB, because it's generally faster today, and lower power.
GTX 1060 (SE) 3GB £189
RX 480 8GB £200
GTX 1060 6GB £220
GTX 1070 £380
GTX 1080 £500
The PS4 is around the RX 470 level. Which means it's only mid-range with the likes of the GTX 1070, 1080, and soon 1080Ti far ahead.
Pro does not use direct x api. Gains will be better than any comparable card on the pro.
£150 for a comparable GPU to the one in the PS 4 Pro
The PS 4 Pro costs £350
So not comparable in price.
So:
£150 for the GPU
£20 for 4gb RAM
£40 for the 1tb drive
£60 for an AMD 5350 and motherboard
£50 for Rosewill case + 400W PSU
Total: £320.
What was your point exactly?
No, you get a more powerful GPU for this. To put into a PC. Which will cost more than a PS4 Pro. Because it does ten times more things. It's a PC not a games console. This has nothing at all to do with games console. It's a PC graphics card to build a more powerful machine
Glad you managed to figure that out for yourself after obviously stumbling onto the wrong page.
A GTX1060 generally beats a desktop Radeon RX 480 on overall graphics performance, to the tune of 10-15 percent. PS4 Pro's GPU is based on it, but significantly downclocked (about 80 percent of the clock speed) and therefore slower than a RX 480.
So there you go.
So not a good deal if it is
Don't SLI until you get to high end and can't do better without multiple cards anyway, it's a waste of time and money.
Anyways, this isn't the 6gb, its the 3gb so don't buy it :smiley:
The product code is usually the most reliable part of the spec: GV-N1060IX-3GD
That's the 3gb model. I think Amazon have made a spec mistake.
edit: the picture also seems to be 3gb, check the bottom left
Also, I don't think Gigabyte do a non-overclocked version of this card in 6gb format, only 3gb
I have the OC 6gb version, it is a very good card can push it quite far with an overclock and its not too loud
http://www.hotukdeals.com/deals/nvidia-gtx-1060-6gb-oc-edition-gv-n1060ixoc-6gd-218-92-delivered-laptopsdirect-2565219
http://uk.camelcamelcamel.com/Gigabyte-GeForce-1060-Mini-GDDR5/product/B01M4OVLAZ