Overclock.net › Forums › Industry News › Rumors and Unconfirmed Articles › [KitGuru] AMD Radeon RX Vega caught hiding in plain sight
New Posts  All Forums:Forum Nav:

[KitGuru] AMD Radeon RX Vega caught hiding in plain sight - Page 10

post #91 of 664
Quote:
Originally Posted by iLeakStuff View Post

If Vega comes out equal to GTX 1080, AMD have failed bigtime.
It will be 2 freaking years after GTX 1080 was released and they only now come out with something competitive to that?

Slap on HBM2 and hope that is enough for sales?
AMD should sell their GPU IPs to a company that wants to compete for the sake of their customers if true.

This is BS. Better be fake

google says the 1080 FE release date was May 27,2016....where are you getting 2 years? I get it that AMD is behind, but geez, throw the brakes a bit
post #92 of 664
Quote:
Originally Posted by iLeakStuff View Post

If Vega comes out equal to GTX 1080, AMD have failed bigtime.
It will be 2 freaking years after GTX 1080 was released and they only now come out with something competitive to that?

Slap on HBM2 and hope that is enough for sales?
AMD should sell their GPU IPs to a company that wants to compete for the sake of their customers if true.

This is BS. Better be fake

Wait, didn't the 1080 just come out last year? How would that be two years?

At any rate this article is garbage (unsurprising coming from KG). I think the safest bet is to expect an absolute worst case scenario being parity with the 1080 and a best case scenario of near-1080Ti performance, with the truth lying somewhere in between.
post #93 of 664
oops
Blue Fury
(19 items)
 
  
CPUMotherboardGraphicsRAM
AMD A8 6600k asus Sapphire 290X Tri-OCX 4GB 2x4 (8)GB G-Skill Ripjaws 1600 
Hard DriveHard DriveHard DriveHard Drive
OCZVertex3 60GB Seagate 1TB WD 500GB Samsung 
Hard DriveOptical DriveCoolingOS
Samsung Evo 850 ASUS DVDRW24x Corsair H100 W10 
MonitorKeyboardPowerCase
55" 4k Hisense Razer Blackwidow Ultimate 2013 Corsair CX750 AzzA Hurrican 
MouseMouse PadAudio
Razer Deathadder 3500 Diablo3 RoS LE onboard 
  hide details  
Reply
Blue Fury
(19 items)
 
  
CPUMotherboardGraphicsRAM
AMD A8 6600k asus Sapphire 290X Tri-OCX 4GB 2x4 (8)GB G-Skill Ripjaws 1600 
Hard DriveHard DriveHard DriveHard Drive
OCZVertex3 60GB Seagate 1TB WD 500GB Samsung 
Hard DriveOptical DriveCoolingOS
Samsung Evo 850 ASUS DVDRW24x Corsair H100 W10 
MonitorKeyboardPowerCase
55" 4k Hisense Razer Blackwidow Ultimate 2013 Corsair CX750 AzzA Hurrican 
MouseMouse PadAudio
Razer Deathadder 3500 Diablo3 RoS LE onboard 
  hide details  
Reply
post #94 of 664
Quote:
Originally Posted by AlphaC View Post

Polaris at 1.4GHz+ is past the efficient range of GCN. They had to bump up voltage to > 1.1V to achieve that. However, the Techpowerup made-up statistic had 1.2GHz boost clocks

The voltage is not the only metric that went up.

From the RX480 to RX580 .. (model depending)
-TDP went from
--> 160W to 185W (bigger coolers necessary)

-power consumption went from
~160-180W (OC to 200W+ possible)
-> to ~225-250W (no OC headroom left)


The jump from 232mm² Polarsi to a assumed 400mm² Vega are 58% increased die-space. That is the number the smaller Vega will have. A bigger one is planned later. To scale this chip up with such outragious power- and heat-requirements isn't possible. HBM2 can't save this either, especially since smaller chips are known to be more efficient than bigger ones. . Vega needs a radical design.

Quote:
Originally Posted by AlphaC View Post

1.4GHz Polaris is only equal to a 2GHz Pascal part if it is an optimized situation where there is no memory compression possible and use of more cores is better. In reality you're looking at less than Pascal GTX 1060 6GB performance because the RX 580 edges outs the GTX 1060 6GB at stock (at best) and not across the board.

Above I talked about chipsize. ironically Polaris was the bigger one compared to Pascal. GP106 is ~200mm² big. Nvidia is outclassing them with a smaller chip on every metric. The extra 2GB VRAM are not the deciding factor here. It's the core itself, which makes that much troubles. The higher clocks of the RX500-series are not helping either, yet making matters worse.

Quote:
Originally Posted by AlphaC View Post

The RX 480 / RX 580 is a 2304 shader part , the GTX 1060 6GB uses 1280 CUDA cores. That's 80% more cores. Those cores use up power even when idle, so you can't expect it to scale well if the same clocks are used.

Polaris does not look good here either compared with the mid-range Pascal chip.

From Polaris's 2304 to 4096 shaders of Vega are 58% more shaders.

Nvidia's GTX 1080 has 2560 CUDA cores (shaders), which is essentially a doubled 1060. By the numbers Vega's theoretical performance is bigger than the 1080, though we all know AMD never gets their rawperformance fully on the road. Results between a 1070 and 1080 are therefore possible. At least I think them as realistic in DX11 titles.

The results of the Vega demos with DOOM were Vulkan exclusive. Meaning it was already the best case scenario for AMD here in an AMD favoring title / engine. I start to imagine Vega being all over the place like Fiji was before. In the long run this could stabilize with driver updates, yet I doubt it will be at launch.
post #95 of 664
Quote:
Originally Posted by Kuivamaa View Post


When we discuss pure performance, in terms of applied tech, price is irrelevant. If 500mm2 of brand new NCU cannot match a pascal chip of comparable size, AMD failed. Selling the end product cheap would be the result of that failure. That being said, I expect a card that will compete VERY well with GP102.
Quote:
Originally Posted by Hardware Hoshi View Post


Having the higher general costs of HBM2 in mind, this is very unlikely. Having the memory alone with a said to be 4096 shader GPU will never cost under 600-700 bucks. If Vega is remotely close to the 1080 Ti, AMD can price it similar .. maybe slightly cheaper. Offering their product would indicate they are weaker than Nvidia. AMD's CEO Lisa Su said that AMd does not want to be seen as the budget brand.

Ryzen was an exception because AMD was practically dead in the CPU section for years. There is as good as no brand reputation left and prizing the CPUs way cheaper was the only way to get back into the heads of the consumers.

(sorry this was not the answer to your post, but my quotes got missing and I am not willing to go through all comments to find it again thumb.gif)
Half the price is a myth. AMD tried this in the past to gain market share (and it failed miserably). People bought the stuff but AMD also lost lots of money, staying in the losses for years. It will be a wonder if they even get to 1080 Ti levels without another round of compromizes.

The AMD stock is nothing bug an object for high-risk gamblers. If Vega is not the mega-hit, it could theoretically drop dramatically again.
You either meant Maxwell's GTX 980 Ti or had meant with 1 year delay after Pascal GTX 1080.

Of course it would be a fail, but if we consider the scarce ressources AMD is operating with, the result would not be a surprise. Ryzen took serious amount of the companies R&D budget. The RTG division must have suffered severely from this fact. In such a situation being competitive is outright ludicrous. Keep in mind AMD was already behind against Maxwell. Against Pascal they probably try to get as much ground as they can. Hard to do if there are so many road blocks and open issues to deal with but no time and money available.
Your suggestion is actually my ideal scenario. Better sell the RTG and let someone take control who is up for the job. Being in a pseudo competion only and deliver medicre results will help nobody except maybe Nvidia.

On the other hand, this does not necesarily be a fake. The numbers are theoretcial, but for a strange reason they could match the smaller Vega. If true the smaller Vega comes 2017 and the bigger one .. let's say early 2018. With all the chaos in the AMD naming & model schemes it is totally confusing talking about their development.
You miss the point. It is not poor Volta, rather poor AMD marketing.
Their marketing campaigns have been questionable since a while. From Fiji's overclockers dream to "has a brain and a soul" nothing but nonsense came out of their sources. I wouldn't believe a single word they say. Especially since Polaris has been such a failure in technical terms.

Wait for final reviews and prices, then we can talk again.

 

Oh, I agree with you but I just wanted to mention that from the consumer's perspective it could succeed if priced well enough. Performance wise I think, looking at Polaris clocked to the edge vs NVidia's "slashed in half" midrange die on conservative clocks, the design lead is evident.

Pernod
(28 items)
 
Oleo
(14 items)
 
The Sidewinder
(18 items)
 
CPUCPUMotherboardGraphics
Intel Xeon E5-4650 Intel Xeon E5-4650 Asus Z9PE-D8 WS GeForce GTX 780 Ti 
RAMRAMHard DriveHard Drive
4 x 4GB Corsair Dominator Platinum 4 x 4GB Corsair Dominator Platinum Samsung Spinpoint F3 2TB Sandisk Ultra 64GB 
Hard DriveCoolingCoolingCooling
mushkin Triactor 240GB EK Supremacy Nickel/Plexi EK Supremacy Nickel/Plexi TFC XChanger 360 
CoolingCoolingCoolingOS
Koolance RP452X2 Swiftech MCP-655 Corsair SP120 "QE" Windows 10 Enterprise x64 
OSMonitorMonitorMonitor
openSUSE Leap 42.1 x64 Dell U2412M Dell U2412M Acer P241W-d 
KeyboardPowerCaseMouse
Ducky 1008 w/ Cherry MX Blue Coolermaster Silent Pro M2 1.5kW Silverstone Raven RV01B-W Microsoft Sidewinder X5 
AudioAudioAudioOther
Asus Xonar DX Beyerdynamic DT 880 Premium AudioEngine A2 (2.0) Wacom Intuos Pen & Touch 
CPUMotherboardGraphicsRAM
Core i7 2600K @5.2  MSI MPower Z77 GeForce GTX 470 Corsair Vengeance 2000MHz C10 
Hard DriveCoolingCoolingCooling
Samsung Spinpoint F1 XSPC Raystorm EK Spin Bay Res Clear mcp-350 (ddc1) 
CoolingCoolingOSOS
XSPC EX120 XSPC EX240 openSUSE 12.3 x64 elementaryOS 0.3 (Freya) x64 
PowerCase
Corsair TX750 750W Lian Li A05NB 
CPUMotherboardGraphicsRAM
Intel C2Q Q9550 Asus P5G41C-M LX GeForce GTX 470 Corsair XMS2 DDR2 800 
Hard DriveHard DriveHard DriveHard Drive
Seagate LP 500GB Crucial M4 128GB WD Red 2TB WD Red 4TB 
CoolingOSMonitorPower
Scythe Big Shuriken w/ Silverstone Suscool 121 Windows 10 Home x64 Sony Bravia KDL-37EX500 Antec TP650 New 
CaseOther
Lian Li V351B Silverstone Suscool 121 120mm case fans 
  hide details  
Reply
Pernod
(28 items)
 
Oleo
(14 items)
 
The Sidewinder
(18 items)
 
CPUCPUMotherboardGraphics
Intel Xeon E5-4650 Intel Xeon E5-4650 Asus Z9PE-D8 WS GeForce GTX 780 Ti 
RAMRAMHard DriveHard Drive
4 x 4GB Corsair Dominator Platinum 4 x 4GB Corsair Dominator Platinum Samsung Spinpoint F3 2TB Sandisk Ultra 64GB 
Hard DriveCoolingCoolingCooling
mushkin Triactor 240GB EK Supremacy Nickel/Plexi EK Supremacy Nickel/Plexi TFC XChanger 360 
CoolingCoolingCoolingOS
Koolance RP452X2 Swiftech MCP-655 Corsair SP120 "QE" Windows 10 Enterprise x64 
OSMonitorMonitorMonitor
openSUSE Leap 42.1 x64 Dell U2412M Dell U2412M Acer P241W-d 
KeyboardPowerCaseMouse
Ducky 1008 w/ Cherry MX Blue Coolermaster Silent Pro M2 1.5kW Silverstone Raven RV01B-W Microsoft Sidewinder X5 
AudioAudioAudioOther
Asus Xonar DX Beyerdynamic DT 880 Premium AudioEngine A2 (2.0) Wacom Intuos Pen & Touch 
CPUMotherboardGraphicsRAM
Core i7 2600K @5.2  MSI MPower Z77 GeForce GTX 470 Corsair Vengeance 2000MHz C10 
Hard DriveCoolingCoolingCooling
Samsung Spinpoint F1 XSPC Raystorm EK Spin Bay Res Clear mcp-350 (ddc1) 
CoolingCoolingOSOS
XSPC EX120 XSPC EX240 openSUSE 12.3 x64 elementaryOS 0.3 (Freya) x64 
PowerCase
Corsair TX750 750W Lian Li A05NB 
CPUMotherboardGraphicsRAM
Intel C2Q Q9550 Asus P5G41C-M LX GeForce GTX 470 Corsair XMS2 DDR2 800 
Hard DriveHard DriveHard DriveHard Drive
Seagate LP 500GB Crucial M4 128GB WD Red 2TB WD Red 4TB 
CoolingOSMonitorPower
Scythe Big Shuriken w/ Silverstone Suscool 121 Windows 10 Home x64 Sony Bravia KDL-37EX500 Antec TP650 New 
CaseOther
Lian Li V351B Silverstone Suscool 121 120mm case fans 
  hide details  
Reply
post #96 of 664
Quote:
Originally Posted by AlphaC View Post

Quote:
Originally Posted by Yttrium View Post

I dont understand how you got to the 980GB/s number for vega.

And your own numbers for 'effective bandwidth' are wrong (effective bandwidth is rather disingenous for enthousiast but I'll get to that ) your own link states that the 1080 has 1.7 greater 'effective bandwidth' compared to a 980 without any compression, something the 980 has but is not taking into account by Nvidia in this slide. the actual benefit from compression is 1.2 as stated in the slide and as stated by the reviewer. "NVIDIA pegs the effective increase in memory bandwidth from delta color compression alone at 20%. The difference is of course per-game, as the effectiveness of the tech depends on how well a game sticks to patterns"

which means the 1080's 380GB/s has 380*1.2 = 456GB/s

Then you talk about the 1080 Ti and a factor of 2.5 for compression, If you could elaborate more on this that would be great.

PS, on the same page it mentions bandwith over GPU power is getting lower over generations, how would it be bad if Vega follows that trend?
The GTX 1080 has 256-bit memory interface operating at a 10Gbps = 320GB/sec for reference cards. 320GB/s x 2.5 = 800 GB/s when there is DCC involved (such as graphics).

Corrected. thumb.gif
Quote:
Originally Posted by AlphaC View Post


Why don't you read the article again about delta color compression...

The 2.5X of Pascal is based on no compression. It's 1.2x the benefit of Maxwell (GTX 980/GTX 980 Ti).

The GTX 1080 has 1.4X the raw bandwidth (i.e. no compression) from going to 10GBps instead of 7 Gbps GDDR5x.

The GTX 1080 Ti uses 11Gbps GDDR5X on a 352-bit memory bus = 484 GB/s , with 2.5X color compression = 1,210GB/s

To obtain 1,210GB/s with VEGA assuming Polaris level 1.3X DCC , you need ~ 930GB/s.


This DCC doesn't apply to compute workloads.


I'm starting to see where you're coming from, Pascal supposedly has 20% better compression over Maxwell, which reportedly has an average of 25% compression in games. Then you assume vega will/needs to hit the same "effective bandwidth" as the 1080 ti and calculate the raw bandwidth from tahiti's compression numbers.

However this does not amount to 2.5 times better "effective bandwidth"

I am starting to doubt wether it truly is 20% over non-compressed memory or 20% over maxwell's bandwidth numbers. But I have good reason to suspect the former. - keep in mind, Because compression differs on a per-game basis and that it HAS to be lossless.



It makes no sense to bench 5 games on maxwell, and post the average, results of memory compression and then bench pascal on 6 completely different games and THEN subtract the average of last generation's compression benefit. The reason the article sounds so confusing is because it resembles the way the whitepaper was written.
Warning: Spoiler! (Click to show)
Quote:
To put all of this in numbers, NVIDIA pegs the effective increase in memory bandwidth from delta color compression alone at 20%. The difference is of course per-game, as the effectiveness of the tech depends on how well a game sticks to patterns (and if you ever create a game with random noise, you may drive an engineer or two insane), but 20% is a baseline number for the average. Meanwhile for anyone keeping track of the numbers over Maxwell 2, this is a bit less than the gains with NVIDIA’s last generation architecture, where the company claimed the average gain was 25%.

The net impact then, as NVIDIA likes to promote it, is a 70% increase in the total effective memory bandwidth. This comes from the earlier 40% (technically 42.9%) actual memory bandwidth gains in the move from 7Gbps GDDR5 to 10Gbps GDDR5X, coupled with the 20% effective memory bandwidth increase from delta compression.
Quote:
A
s a result of the improvements in memory compression,
the GeForce GTX 1080
is able to significantly
reduce the
number of bytes that have to be fetched from memory per frame.
This reduction in bytes
fetched translates to roughly 20% additional effective bandwidth, and when combined
with G
eForce
GTX 1080’s
10
Gbps GDDR5X memory, this
ultimately provides the GTX 1080
a 1.7x effective memory
bandwidth increase over GeForce GTX 980
.


Source: http://international.download.nvidia.com/geforce-com/international/pdfs/GeForce_GTX_1080_Whitepaper_FINAL.pdf

Warning: Spoiler! (Click to show)
Quote:
Originally Posted by AlphaC View Post


If you recall Tahiti HD 7970 / R9 280X transitioned from 384-bit to 256-bit on Tonga (R9 285) by using delta color compression.


http://www.anandtech.com/show/8460/amd-radeon-r9-285-review/3
AMD likes to use "Up to 40% better efficiency*"

*up to, meaning, best case scenario.

Maybe one day the press will be able to test these kinds of claims with better accuracy or leave this out.


In the end, memory compression is irrelivant for the consumer but relevant for the enthousiast if used correctly.
Edited by Yttrium - 4/30/17 at 1:57am
post #97 of 664
Quote:
Originally Posted by Hardware Hoshi View Post

Quote:
Originally Posted by AlphaC View Post

Polaris at 1.4GHz+ is past the efficient range of GCN. They had to bump up voltage to > 1.1V to achieve that. However, the Techpowerup made-up statistic had 1.2GHz boost clocks

The voltage is not the only metric that went up.

From the RX480 to RX580 .. (model depending)
-TDP went from
--> 160W to 185W (bigger coolers necessary)

-power consumption went from
~160-180W (OC to 200W+ possible)
-> to ~225-250W (no OC headroom left)


The jump from 232mm² Polarsi to a assumed 400mm² Vega are 58% increased die-space. That is the number the smaller Vega will have. A bigger one is planned later. To scale this chip up with such outragious power- and heat-requirements isn't possible. HBM2 can't save this either, especially since smaller chips are known to be more efficient than bigger ones. . Vega needs a radical design.

Quote:
Originally Posted by AlphaC View Post

1.4GHz Polaris is only equal to a 2GHz Pascal part if it is an optimized situation where there is no memory compression possible and use of more cores is better. In reality you're looking at less than Pascal GTX 1060 6GB performance because the RX 580 edges outs the GTX 1060 6GB at stock (at best) and not across the board.

Above I talked about chipsize. ironically Polaris was the bigger one compared to Pascal. GP106 is ~200mm² big. Nvidia is outclassing them with a smaller chip on every metric. The extra 2GB VRAM are not the deciding factor here. It's the core itself, which makes that much troubles. The higher clocks of the RX500-series are not helping either, yet making matters worse.

Quote:
Originally Posted by AlphaC View Post

The RX 480 / RX 580 is a 2304 shader part , the GTX 1060 6GB uses 1280 CUDA cores. That's 80% more cores. Those cores use up power even when idle, so you can't expect it to scale well if the same clocks are used.

Polaris does not look good here either compared with the mid-range Pascal chip.

From Polaris's 2304 to 4096 shaders of Vega are 58% more shaders.

Nvidia's GTX 1080 has 2560 CUDA cores (shaders), which is essentially a doubled 1060. By the numbers Vega's theoretical performance is bigger than the 1080, though we all know AMD never gets their rawperformance fully on the road. Results between a 1070 and 1080 are therefore possible. At least I think them as realistic in DX11 titles.

The results of the Vega demos with DOOM were Vulkan exclusive. Meaning it was already the best case scenario for AMD here in an AMD favoring title / engine. I start to imagine Vega being all over the place like Fiji was before. In the long run this could stabilize with driver updates, yet I doubt it will be at launch.


Of course it's possible to scale up because the power requirements won't be as you say. People keep forgetting that Vega will have a tile based renderer. If they do it right, it will be just like when Nvidia went from Kepler to Maxwell, where we got a new generation of GPUs on the same node.

Also, the 2 GB of VRAM on the current Polaris is connected to a 256-bit memory bus and controller, whereas the GTX 1060 has a 192-bit memory bus and controller and thus the Polaris chip is necessarily a bit bigger on that account. That's also to say that GP106 is not half of GP104. Along with the 192-bit bus, it also has 48 ROPs.

Results for Vega between a 1070 and a 1080 would only require migrating a Fury X to the new node and clocking it at 1.2 Ghz. That would mean that AMD would have done zero work on the architecture and we would have been waiting for nothing. That goes against everything that we already know about Vega.
 
Metro 2033 review
Metro 2033
CPUMotherboardGraphicsRAM
Core i7-3820 Asus Sabertooth X79 MSI GTX 750 Ti TF Gaming 16 GB Corsair DDR3 1866 Mhz Dominator 
Hard DriveOptical DriveCoolingOS
Samsung SSD 830 128GB + WD Caviar Black 1TB Sony Optiarc DVD-RW Corsair A70 + Noiseblocker M12-P Windows 7 Home Premium 64-bit 
MonitorKeyboardPowerCase
BenQ RL2455HM Cooler Master Octane Corsair AX750 Professional Modular 80 Plus Gold Cooler Master HAF 912 Plus 
Mouse
Cooler Master Octane 
  hide details  
Reply
 
Metro 2033 review
Metro 2033
CPUMotherboardGraphicsRAM
Core i7-3820 Asus Sabertooth X79 MSI GTX 750 Ti TF Gaming 16 GB Corsair DDR3 1866 Mhz Dominator 
Hard DriveOptical DriveCoolingOS
Samsung SSD 830 128GB + WD Caviar Black 1TB Sony Optiarc DVD-RW Corsair A70 + Noiseblocker M12-P Windows 7 Home Premium 64-bit 
MonitorKeyboardPowerCase
BenQ RL2455HM Cooler Master Octane Corsair AX750 Professional Modular 80 Plus Gold Cooler Master HAF 912 Plus 
Mouse
Cooler Master Octane 
  hide details  
Reply
post #98 of 664
We can't estimate the performance from what it seems a completely new architecture. It should be the same jump of efficient of Kepler -> Maxwell because of the tiled based rasterize.

If it just a scaled up Polaris they could release the card ages ago.

I am waiting minimum 1080ti performance otherwise it is a fail.

AMD the last years could fight on the high end (7970/290x) and even fiji failed because of the maxwell oc abilities.
Workstation
(4 items)
 
Workstation
(4 items)
 
 
CPUMotherboardGraphicsMonitor
Xeon E5-2690 Supermicro 2011 Amd W9100/Quadro K5200 Dell ultrasharp 4k 
CPUMotherboardGraphicsMonitor
Xeon E5-2690 Supermicro 2011 Amd W9100/Quadro K5200 Dell ultrasharp 4k 
  hide details  
Reply
Workstation
(4 items)
 
Workstation
(4 items)
 
 
CPUMotherboardGraphicsMonitor
Xeon E5-2690 Supermicro 2011 Amd W9100/Quadro K5200 Dell ultrasharp 4k 
CPUMotherboardGraphicsMonitor
Xeon E5-2690 Supermicro 2011 Amd W9100/Quadro K5200 Dell ultrasharp 4k 
  hide details  
Reply
post #99 of 664
but... why will you rebel?
post #100 of 664
This could be small Vega, http://www.3dmark.com/spy/1544741
Doppelganger
(13 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7 5820K X99-UD4-CF R9 Fury G.Skill 
RAMHard DriveOSMonitor
G.Skill Intel P600 NVME Windows 10 Pro AOC G2460PF 144Hz 1ms Freesync Monitor 
KeyboardPowerCaseMouse
AULA Mechanical Blue Switch Keyboard EVGA 700w Phanteks P400 Carbon Fiber Custom Mad Catz R.A.T. 3 
Mouse Pad
Steel Series Cusion 
  hide details  
Reply
Doppelganger
(13 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7 5820K X99-UD4-CF R9 Fury G.Skill 
RAMHard DriveOSMonitor
G.Skill Intel P600 NVME Windows 10 Pro AOC G2460PF 144Hz 1ms Freesync Monitor 
KeyboardPowerCaseMouse
AULA Mechanical Blue Switch Keyboard EVGA 700w Phanteks P400 Carbon Fiber Custom Mad Catz R.A.T. 3 
Mouse Pad
Steel Series Cusion 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Rumors and Unconfirmed Articles
Overclock.net › Forums › Industry News › Rumors and Unconfirmed Articles › [KitGuru] AMD Radeon RX Vega caught hiding in plain sight