Page 1: Massive performance: MSI GeForce RTX 3080 Suprim X in the test
For each of the new ampere variants, we looked at a model from MSI. So far there has been no sign of a Lightning variant, which has a reason. MSI introduces the X Suprim a new series before, now with the models of GeForce RTX 3080 and GeForce RTX 3090 makes its start. We took a look at the MSI GeForce RTX 3070 Suprim X at the start.
With the Radeon RX 6800 and Radeon RX 3090 XT, AMD presented the first two Big Navi cards and makes steam for the industry leader NVIDIA. But even at AMD, the high demand has probably been underestimated or could no longer react in the given time and so many interested buyers are currently missing out in the red warehouse. Nevertheless, the first custom models will arrive with us in a few days, which we will then take a closer look at.
The subject of “custom models” now also applies to this one Test plays an important role, because the Suprim-X series is the spearhead in terms of design and cooling from MSI. First a look at the technical data:
Comparison of the cards
RTX 3080 FE
MSI GeForce RTX 3080 Suprim X
GPU
Ampere (GA 102)
Ampere (GA 102)
transistors
28 Billions
28 Billions
Production
8 nm
8 nm
Chip size
628, 4 mm²
628, 4 mm²
FP 32 – ALUs
8th.704
8th.704
INT 32 – ALUs
4. 352
4. 352
SMs
68
68
Tensor Cores
272
272
RT Cores
68
68
Base- Measure
1. 440 MHz
1.440 MHz
Boost clock
1.710 MHz
1. 905 MHz
Storage capacity
10 GB
10 GB
Memory type
GDDR6X
GDDR6X
Memory interface
320 Bit
320 Bit
Memory clock
1.188 MHz
1.188 MHz
memory bandwidth
760 GB / s
760 GB / s
TDP
320 W
370 W
Price
699 Euro
799 Euro
It is not surprising that the MSI GeForce RTX 3080 Suprim X or out GPU is not built up to the clock from each other GeForce RTX 3080 differs. We see therefore produced in an 8 nm GA 102 – GPU with 68 SMs and accordingly 8. 704 FP 32 – and 4. 352 INT 32 – arithmetic units. Added to this are 272 Tensor Cores which are used among other things for the DLSS and 68 RT cores, which accelerate the ray tracing calculations.
The base clock is 1. 440 MHz also identical, but MSI sets the boost clock at 1 . 905 MHz. With this boost setting the card on paper once heard one of the fastest models of the GeForce RTX 3080 . It will be shown later whether it can also implement this in the benchmarks.
MSI gives a recommended retail price of 799 Euro for the MSI GeForce RTX tested by us 3080 Suprim X. The bigger sister MSI GeForce RTX 3090 Suprim X should cost 1. 628 Euro. Whether the cards will actually be available in stores at this price is another matter. A GeForce RTX 3070 Suprim X there will be.
One factor in the performance of the Ampere cards is the power limit. The power limit is not the only point that determines the maximum clock rate. GPU temperature and voltage also play a role.
Power limits
MSI GeForce RTX 3080 Suprim X
maximum power limit
430 W
Default power limit
370 W
minimum power limit
100 W
While NVIDIA has a power limit of 320 W provides, MSI uses the GeForce RTX 3080 Suprim X equals 370 W on. We have also seen this value from other manufacturers. In addition, the power limit can be increased further – in this case to 430 W. Many standard models of the GeForce RTX 3070 set about 340 W and do not allow any changes to the power -Limits too. that is for the high-end models different – including in the GeForce RTX 3080 Suprim X.
<> Massive performance: MSI GeForce RTX 3080 Suprim X in the test MSI GeForce RTX 3080 Suprim X – Impressions
Overclocker Takukou squeezed a Radeon RX 6800 XT with liquid nitrogen, bringing the GPU to operate at 2.8 GHz. The card has been able to reach the top of some versions of the 3DMark, approaching the 50. 000 points in Fire Strike.
by Manolo De Agostini published 23 November 2020 , at 02: 41 in the Video Cards channel Radeon AMD Ships
The architecture RDNA 2 allows GPUs of new AMD video cards Radeon RX 6800 to operate at very high frequencies , well above 2 GHz, so much so that several reference proposals have touched 2.6 GHz in air overclocking . This is a great result, but what happens when you put the GPU under liquid nitrogen?
The Chinese overclocker Takukou gives us a first indication: it managed to push the Navi GPU 21 of a Radeon RX 6800 XT a ben 2.8 GHz (Boost clock), obtaining the first position in 3DMark Fire Strike and Fire Strike Extreme , where he totaled 49456 and 30274 points. Thanks to the liquid nitrogen, Takukou also overclocked the card’s memory, bringing it to 2150 MHz from 2000 MHz by default.
Click to enlarge
To achieve these results, in addition to the reference video card, the overclocker used an MSI MEG X platform 570 Godlike with a CPU Ryzen 9 5950 X overclocked to 5.6 GHz (all cores active) and 32 GB of DDR4 memory – 3800.
Click to enlarge
Takukou’s achievements are certainly noteworthy, but there could be room to do even better, without forgetting that there are Radeon RX 6800 XT customized from part ner, from Sapphire NITRO + to PowerColor Red Devil, capable of even better results.
Chinese overclocker Takukou has broken a new world overclock record by achieving the highest score in the history in the Fire Strike graphical test of the 3DMark benchmark suite. Specifically, it has achieved 49. 553 points using a single graph. For this he has used one of the new RX 6800 XT with overclock, beating all NVIDIA RTX 3090 , a graph with a much higher price range.
The record has been achieved with the RX 6800 XT accompanied by a RYZEN 9 5950 X at 5.6 GHz
To accomplish this feat, Takukou managed to set the RX 6800 XT clock to nothing more and nothing less than 2. 800 Mhz, a considerable increase from the factory frequency. All this cooled by liquid nitrogen and accompanied by the new AMD Ryzen 9 5950 X, also with LN2, at a speed of 5.6 GHz together with 32 3 GB of DDR4 RAM memory. 800 MHz.
The motherboard used is a MSI MEG X 570 GODLIKE , while RX 6800 XT is an AMD reference graph.
Currently, the first and second places in the ranking are occupied by both AMD Radeon RX 6800 XT, followed by six RTX 3090 and a Titan RTX.
As we always say, this type of overclocks are far from those that home users can get on their PCs, but they are a good way to see to what extent a component can be squeezed to the maximum, especially if we are talking about a new architecture , and it is still an interesting high-level competition.
End of Article. Tell us something in the Comments or come to our Forum!
Antonio Delgado
Computer Engineer by training, editor and hardware analyst at Geeknetic since 2011. I love gutting everything that comes my way, especially the latest hardware that we get here for reviews. In my spare time I fiddle with 3d printers, drones and other gadgets. For anything here you have me.
There have been numerous complaints about the difficulty of obtaining some products launched this year, such as the new graphics cards NVIDIA RTX 30 Series and AMD Radeon RX 6000 , processors AMD Ryzen 5000 Series , although it seems that the situation could be repeated with the new generation consoles, even reaching unsuspected limits.
In the same way as on previous occasions, there have been people who have acquired a large number of consoles to later resell them at much higher prices, but this time we would be talking directly about an assault that which has settled with the load of a complete truck.
This truck, from a PlayStation 5 console distributor as we read in Vandal, had the consoles that had to supply Falabella stores in much of Latin America, so its users have seen how the chain got in touch with them to inform that the new delivery date of the consoles would be from the day 15 December.
The chain has apologized and rewarded buyers with a one-year subscription to PlayStation Plus and 50. pesos –some 50 euros— for the delay in delivery, which although they will not be able to return the lost time, yes that they will be a great company when it comes to improving the experience with console when they get it, since this compensation will allow users to have games in the subscription.
End of Article. Tell us something in the Comments or come to our Forum!
Jordi Bercial
Avid enthusiast of technology and electronics. I messed around with computer components almost since I learned to ride. I started working at Geeknetic after winning a contest on their forum for writing hardware articles. Drift, mechanics and photography lover. Don’t be shy and leave a comment on my articles if you have any questions.
Update : It seems that it is a fake image, the model corresponds to a card of the Colorful brand model iGame GeForce RTX 3080 Vulcan OC of 10 GB, also just in the image below to On the left, next to the words RGB and MAX, you can see a 1 that would correspond to 10 GB of memory and not a 20 as it says on the side.
An image has appeared of what could be the final product of an NVIDIA RTX card 3080 Gigabyte Ti Aorus, could pass for an RTX 3080 , but clearly we can see in the box shown in the image how the model corresponds to an RTX 3080 You and that also has 30 GB of VRAM memory . According to this amount of memory, the bus would be 320 bits, slightly lower than RTX 3090.
Although the release of the NVIDIA RTX 30 is recent, there are many rumors that have been heard of Ti versions of these models , even of some new models yet to appear like the 3060 Ti of which the specifications were recently confirmed thanks to GPU-Z, it would not be surprising if this RTX model appeared 3080 Ti in the coming months to counter the attack of the new Radeon RX 6800 and RX 6800 XT.
We have seen how the Ampere architecture doubles the performance of the previous generation, AMD with the Radeon RX 6000 manage to be very at the pair of the NVIDIA RTX 30, but perhaps this is an NVIDIA strategy to put some distance between these new models and the new Radeon RX 6000.
End of Article. Tell us something in the Comments or come to our Forum!
Antonio Delgado
Computer Engineer by training, editor and hardware analyst at Geeknetic since 2011. I love gutting everything that comes my way, especially the latest hardware that we get here for reviews. In my spare time I fiddle with 3d printers, drones and other gadgets. For anything here you have me.
You also may be of interest Other articles and news about technology
Geeknetic Software Software News Drivers available AMD Radeon Adrenalin 20. 11. 2 with support for the new Radeon RX 6800 and WoW: Shadowlands
Antonio Delgado
Computer Engineer by training, editor and hardware analyst at Geeknetic since 2011. I love gutting everything that comes my way, especially the latest hardware that we get here for reviews. In my spare time I fiddle with 3d printers, drones and other gadgets. For anything here you have me.
You may also be interested in Other articles and news about technology
Comments and opinions about: AMD Radeon Adrenalin drivers available 20. 11. 2 with support for the new Radeon RX 6800 and WoW: Shadowlands What do you think? Any questions?
Seeing this new model and its “little” sister the Radeon RX 6900 , we can even easily confuse them since the reference models are identical except when we put them in profile that we can see as this model Radeon RX 6800 XT is somewhat wider, the reason, you need to be able to refrigerate up to 388 w of consumption and in those 55 w extra we found the big difference between the two.
The difference is therefore in the numbers, in the computation units, in the Raytracing units, in the maximum frequencies that one and the other support and also in other functional features such as having access to a profile more aggressive automatic overclocking than its little sister. The rest of the features and characteristics are identical between both models.
New features of the AMD RDNA 2 generation
This second generation of the AMD’s RDNA architecture comes from the same 7nm FinFET manufacturing process already enjoyed by the Radeon RX 5700 and Radeon RX 5700 XT, in fact, they also use the same communication interface, the PCI Express 4.0 released with the RDNA .
Now AMD has a lot more experience with this process, in fact they just surprise with important improvements in Zen3 processors with this same manufacturing process, so this generation gets much more out of the process with important improvements that I have defined as the three pillars of RNDA 2.
These three pillars on which AMD’s RDNA 2 architecture is based are the increase in computing units, the increase in work frequency and the introduction of the Infinity Cache. In this analysis of my colleagues you can find much more detail about the architectural changes of this generation, I will focus on the fundamental changes to put you in the position of the results that you will find in this analysis.
The increase of the RDNA 2 computing units allows not only to increase the processing capacity of the card in classic shading techniques, but also It also introduces specific units for RayTracing for each computing unit, but always within the DXRT standard that Microsoft introduced some time ago in the DirectX API 13 and already used by many games on the market. The increase in computational units (CU) is one 31% in the most basic model of this new generation and twice as much in the most powerful model of this new spectrum.
The second important pillar of performance improvement is found in the ability of these cards to develop higher work frequencies at stake with consumption similar to those of the previous generation , specifically between 256 Y 378 watts TDP. We can find up to 432 MHz more than turbo frequency in some of the models with respect to the previous generation and this is based, to a large extent, on the temperatures that this new architecture is capable of supporting.
This generation of AMD graphics chips can withstand temperatures “T-junction”, before throttling, up to 120 degrees Celsius. Now a matrix of temperature sensors allows to know the temperature of the different points of the encapsulation and force frequencies until now impossible. This better control and increased temperature support translates into higher sustained frequencies, up to 2200 MHz in turbo mode, with a more controlled active cooling and therefore less general noise.
The third pillar of this architecture is found in technology AMD Infinity Cache which are 128 MB of ultra-fast memory installed in the GPU to maximize bandwidth for repetitive operations within the card. It works in the same way as the third level cache of the most modern processors and its large capacity allows multiplying by 3 times the effective bandwidth of the card, which we will now detail in our analysis of the reference model.
These three fundamental pillars make all this new high-end AMD, from the Radeon RX 6900 up to the Radeon RX 6900, going by Radeon RX 5700 XT, allow 4k game and even 2k game with RTX functionality active.
Radeon RX GPU Specifications 6800 XT
The Radeon RX 6800 XT is one of the two cards of this generation that develop a consumption of up to 300 w TDP. Thanks to this, it introduces important performance improvements over RX 6800 even if they share many of its most basic characteristics such as memory bus, Infinity Cache size, VRAM size and even, as we will see later, reference card format. Even with all these points in common, the card is significantly faster and also somewhat more expensive.
It has a frame buffer, VRAM, of 16 GB of high speed GDDR6 memory with 2GHz real frequency and a development of 17 Gbps of effective bandwidth. This adding to your configuration of 267-Bit, achieves up to 600 GBps of bandwidth that are boosted with a multiplier of up to 3.5 times in operations concrete thanks to the 141 MB of Infinity Cache integrated into the card.
Both this and all its sisters in the range support PCI Express 4.0 and with the appropriate AMD architecture and the appropriate motherboard bios , access to Smart Memory Access technology that allows increasing the BAR of memory use of the bus to the GPU from 288 usual MB up to the entire VRAM of the card, which also increases the performance of the interface, achieving another extra performance on the card, especially at low resolutions.
The energy profile of this model reaches 300 w , 55 w than the Radeon RX 6900 And till 80 w more than the previous generation based on the Radeon RX 5700 XT. Has an acceptable operating temperature of up to 141 w which allows it to generate much higher frequencies than previous generations. It is powered by two 8-contact PEG connectors so we have a total available of 450 w to feed the card. Are 200 w per connector y 150 w extra PCI connector Express 4.0 by 16 x.
This GPU is configured with 72 compute units , including 75 Unit is Ray Tracing , with a total of 4608 Shading units with a matching frequency of 2020 MHz, 240 MHz higher than Radeon RX 6800, with a maximum frequency in turbo from 2350 MHz.
Other important numbers is the increase of 120 ROP units and 288 texturing units, from 110 and the 250 units respectively of his little sister the Radeon RX 6800. Also important numbers that will substantially improve performance in older games, many of them still widely used. With the same 28. 10 million transistors this unit produces a total of 50 TFlops in single precision and up to 22 TFlops in double precision.
The connectivity is of the latest generation with two Displayport 1.4 ports with 8k resolution, 1 HDMI connector 2.1 VRR (Variable Refresh Rate) and a fully functional USB-C port including Displayport Alternate Mode, USB 3.1 and USB-C PD support with up to 32 w load power . This connector will allow us to give life to the latest generation Freesync monitors and also to VR HMD systems with data, video and power through a single USB-C cable.
Reference Design
The first two AMD cards of this generation share a very very similar reference design, in fact , the only difference is the thickness of the heatsink that has one and the other. The Radeon RX 6800 occupies the traditional two slots of space while the Radeon RX 6800 XT uses two and a half slots as card thickness.
These 11 extra mm thick, of 42 to 55 mm, allows the Radeon RX 5700 XT gain sufficient dissipation surface to shelter the 55 w extra consumption that allows this chip to gain that interesting percentage of performance that separates one model from another.
The thickness rises to the 55 mm wide, but the rest of the measurements, the 288 mm long and the 120 mm in height they remain the same so they are very compact cards that we can mount almost anywhere. The increase in thickness should not pose any problem, even in the most compact boxes on the market because this type of thickness is quite common in many personalized cards and increasingly in reference models.
Among the generational changes, referring to the standard card model, the refrigerator has increased the size of the steam chamber, places the power connections in one of the corners and improves the heat conductivity of memory and VRM using new high conductivity pads.
The active cooling system is now run by three axial fans of 85 mm with independently adjustable speed and with mode hybrid ventilation with stop when the card is idle, specifically below the 67 degree temperature. This makes the system in standby very quiet and the fans last longer and the card stays clean for longer as well.
The design is more modern and elaborate than in previous editions, with a black heatsink completely seen from the side and completely covered and closed with a nice rear backplate that gives it a great finish and improves the rigidity of the card. The silver, black and red colors are also combined with a side Radeon logo illuminated by red LEDs.
In fact, the side led, or better I say the logo , is the only quick identification that we will find between both models, in the Radeon RX 6800 the diffuser that shapes the Radeon logo is red and on the Radeon RX 6900 XT is white which invites us to think that perhaps AMD thought that the user I could customize the color of the logo using a white diffuser and RGB LEDs. Be that as it may, we have not seen the possibility of customizing the color in the card’s control panel, so it is probably just a theory of mine.
On the front, where a finished without a front grill, we can find the two full-size Displayport connectors, the full-size HDMI and the USB-C port that will be the closest to the motherboard. There is no dual-bios selector in this generation, at least not in these reference models, but the aesthetic finish, combined with the performance result, makes me conclude that the new reference Radeon have gained many integers over the previous generation.
Noise, consumption and temperature at speeds of series
Curiously, this model has been even cooler than the Radeon RX 6800 . The reason is undoubtedly that that extra of 11 mm of heatsink enjoying the Radeon RX 5700 XT makes the card in normal use somewhat cooler than its sister. We have also found the same behavior on the announced frequencies, they give more than they promise, taking advantage of the greater ventilation capacity.
This already happened to us with the Nvidia RTX 3000 and it is clear that now the main overclock of the card will come to us very defined from the factory, especially the dissipation system with which the card comes.
Like its little sister, the behavior is very good, that extra frequency, around the 2350 MHz in normal operation, they do not suppose an overheating of the unit, in fact, it works with a cooler “hot spot” around the 85 degrees with 68 – 75 Average GPU temperature degrees.
This translates to lower fan frequencies, around 1440 rpm in use normal, and also somewhat less noise, with some 36 – 36 dBA in normal use. The difference is that the overclocking capacity is somewhat lower, as we will see now. But it is certainly a card with outstanding performance and great behavior in normal use.
Idle temperature
Load temperature
Overclock
In our overclock tests we do not usually do voltage increases because we understand that it is subjecting our test cards to unnecessary risk, which are already subjected to a lot of stress for days. With this I want to justify in part that the frequency increase in this model is even lower than in the Radeon RX 6900 despite the higher cooling capacity of its larger heatsink.
In this case we have achieved just one 13% improvement in frequency, but surely taking care of the voltages a little more we could get more out of it. Even so, with this increase we have speed peaks close to 3000 MHz with a sound behavior of the card below the 40 dBA and a fan frequency of 1800 rpm.
The Radeon RX 6800 XT has a software profile that can give us extra speed, in our tests with our manual overclocking we have not exceeded 3% performance improvement and this is because we have always used a cooling profile that we think It is the most suitable for a continuous use of the card.
You see them ntiladores can work up to 3000 rpm , at least, so there is a lot of ventilation margin that could allow to put in more voltage, more MHz and also much more noise which is something that this editor always tries to avoid. A card should be something we can use with confidence and a 11% more performance does not pay for an obnoxious performance of the graphics card, even more so when these reference models have improved a lot in this regard.
DXRT, AMD Boost and AMD Smart Access Memory
To know the benefits of this card with some of its new and not so new tricks we have taken some specific games capable of taking advantage of these benefits and we have made a brief comparison of results . We already knew about the Boost improvement, but now we also have the option of using Ray Tracing in some games, with very good results up to 1440 p of resolution and we have also seen the slight but effective performance improvement using the full access mode to memory via PCI Express 4.0.
Testing machine:
Processor: AMD Ryzen 3950 X
Memory: 16 GB D DR4 3950
Source: Seasonic Connect 750 w
Hard Drive: Corsair MP 600
Performance
Ashes of the Singularity (DX 14) 1082
DOOM (Vulkan) ultra 1082
Halo Wars 2 (DX 14) ultra 1080
Ghost Recon Wildlands (DX 11) ultra 1082
Total War: Warhammer (DX 14) ultra 1080
Battlefield 1 (DX 14) ultra 1080
StarWars BattleFront 2 ultra (DX 12) ultra 1118
Battlefield V (DX 13) ultra 1082
Doom Eternal ultra 1082
Fight Simulator Ultra 1080
Ashes of the Singularity (DX 14) 1800
DOOM (Vulkan) ultra 1440
Halo Wars 2 (DX 14) ultra 1800
Ghost Recon Wildlands (DX 11) ultra 1440
Total War: Warhammer (DX 12) ultra 1440
Battlefield 1 (DX 12) ultra 1440
StarWars BattleFront 2 ultra (DX 13) ultra 1440
Battlefield V (DX 13) ultra 1932
Doom Eternal ultra 1440
Fight Simulator Ultra 1440
Ashes of the Singularity (DX 14) 2200
DOOM (Vulkan) ultra 2200
Halo Wars 2 (DX 12) ultra 2200
Ghost Recon Wildlands (DX 13) ultra 2200
Total War: Warhammer (DX 13) ultra 2200
Battlefield 1 (DX 13) ultra 2200
StarWars BattleFront 2 ultra (DX 14) ultra 2160
Battlefield V (DX 14) ultra 2160
Doom Eternal ultra 2020
Fight Simulator Ultra 2160
3DMark Firestrike
3DMark Firestrike Ultra
VRMark Orange Room
VRMark Cian Room
Gameplay Doom Eternal 4k in Ultra quality:
?
Analysis and conclusion
The Radeon RX 6900 XT is not the best that AMD can offer for this generation, but it is which is a card that has surprised us a lot because it defends well against such a capable rival as the RTX 3600 from Nvidia. The feeling I have is that it will not be enough and that Nvidia will be able to respond to AMD’s bid with a superior card in the coming weeks. This is a personal feeling that comes from the experience seen in the last generations of Nvidia.
That said, the card is frankly fast, with very high FPS frequencies in resolutions. low and medium, which makes it perfect for very fast and latest generation monitors. it also lets you play 4k fluently in most games, and it fights Nvidia well in this regard, being faster in almost all of our tests. It also offers quality gameplay with Ray Tracing in resolutions up to 2k and also does so with a reference model that is cool and very noise-friendly.
It is without doubt an excellent card, which will be surpassed in the next few days by the next Radeon RX
XT , but that perhaps leaves us with that feeling that AMD had to be even faster in this generation if it wants to do with Nvidia the same as it has done with Intel in the last years.
End of Article. Tell us something in the comments or come to our forum!
I remember the time when I was playing Quake II and I was amazed by the multitude of possible multiplayer strategies, from the classic rocket launcher to the railgun or hyper blaster. Compared to all the FPS titles played up to that point, Quake II definitely brought something extra to the multi-player chapter and only I know how many hours I lost on Warehouse hunting innocent colleagues who thought they had a chance in front of the hyper blaster.
It would be almost two years before I installed a 3DFx Voodoo graphics accelerator on my PC, and the “3DFx OpenGL” option that appeared in the menu after this hardware upgrade and implicitly, the visual effects that activating this option brought, made me realize that the future of gaming would change irretrievably. Slowly but surely, the era of 2D games was beginning to set and a new wave of 3D titles would soon invade the market, some of them reaching legendary status today.
Time passed, PCs changed, and I switched to Voodoo Banshe, and then I entered the Nvidia world, with the purchase of a Riva TNT2. One year later, with 11 for years, a new brand of competition was born, a brand that you all know today. It’s about Radeon, the successor of the legendary ATI Rage series, and the first ATI Radeon graphics card I used I think was 9600 Pro, 3 years after the appearance of the brand- community. They were followed by X 1650 Pro and X 1950 Pro, a few years later, and then… I lost the series of graphics cards, being irretrievably affected by the severe condition called generic overclocking.
Obviously, a series of extreme overclocking sessions followed, together with my colleagues, in which I got to know better 8800 GTX, 8800 Ultra, 8800 GT, 9800 GX2, 3870 X2 or GTX 64, newly released boards, which others would have given anything to play with. We, on the other hand, were preoccupied with something slightly different, as a result of which came multi-card configurations, hundreds of hours of 3DMark, tens of liters of nitrogen and many lost nights…
And yet, even if my memories as a simple user stop at that time, somewhere more than years I started another journey, in which I tested, overclocked, disassembled and tormented most of the video cards launched since then and until now, in as a publisher. I started with 4870 X2, then followed GTX 290, GTX 280, AMD Radeon series 4000, AMD Radeon series 5700) and the rest is history.
However, one of the memories I have from that period is the constant competition between Nvidia and AMD, both in the mainstream segment and in the high-end segment. 8800 Ultra was a monster that dominated the market for years and only 3870 X2 was finally able to beat her. 9800 GX2 restores balance, bringing the crown back to the green camp, only to be dethroned by 4870 X2. GTX 319, HD 5970, GTX 285 / GTX 590), HD 6990, GTX 590, HD 7970 GHz, GTX 580, HD 7990, Titan, R9 64 X, GTX 780 Ti, R9 290 X2, GTX 780 Ti, Fury X, Titan X.
I know, a whole list of models that have become memories in the meantime. And yet, among these memories, AMD’s persistence in trying to always compete at the top, not only in the mainstream segment, persists. And from 3870 X2 to the Fury series, 8 years have passed, 8 years in which AMD has not always had the advantage technologically, but they fought hard, side by side with Nvidia.
And yet, AMD stuck to the GCN architecture, launched at the same time as Kepler, while Nvidia switched from Kepler to Maxwell and then to Pascal. And the launch of Pascal architecture, 4 years ago, marked the impossibility for AMD to compete with the top of the range Nvidia. Because any top of the range AMD launched from then until now, regardless of whether we are talking about RX 290 or RX 580, Vega 20, Radeon VII or RX 5700 XT, could not touch the top of the range Nvidia equivalent.
This year… have been fulfilled 11 years of Radeon… And Nvidia have released a new series of graphics cards, which bring a consistent performance boost over the previous generation, including a top of the range monstrous. We will not discuss today the absolute top of the range. No, this is a discussion for December 8th. Until then, however, we will leave RTX 4000 in peace and focus on RTX 3070 and RTX 3090, respectively their competitors AMD Radeon RX 6800 XT and AMD Radeon RX 6990. Maybe in the year in which they are fulfilled 64 years of Radeon, we will finally have a surprise…
Since the release of AMD Radeon RX 6800 XT and RX 6800 only a few days passed, but it didn’t take long for the first extreme overclocking attempts by professional overclockers. Halfway through the week, Alva “Lucky_n 00 b” Jonathan reached the limit 2650 MHz (Boost) on reference cooling. In order to break records, however, stronger “guns” are used, and so did the Chinese Takukou overlocker, who managed to reach 2.8 GHz using liquid nitrogen (LN2). He also used this type of cooling for the AMD Ryzen 9 processor paired with the Radeon RX 6800 XT 5950 X (OC up to 5.6 GHz). DDR4 G.Skill Trident Z Royal RAM 32 GB and motherboard were used for the set MSI MEG X 570 Godlike. What was the result of the popular 3D Mark Fire Strike benchmark?
Chinese Takukou overclocker overclocked AMD Radeon RX 6800 XT graphics card to record 2.8 GHz using liquid nitrogen. It seems that crossing the 3 GHz barrier is only a matter of time.
AMD Radeon RX graphics card test 6800 XT vs GeForce RTX 3080
Takukou Overlocker from TecLab by overclocking the Radeon RX 6800 XT to the record limit of 2.8 GHz and AMD Ryzen processor 5950 X to 5.6 GHz was able to achieve in 3DMark Fire Strike 48890 points (overall score ) and 61831 points (graphic score). This is the first time a graphics card has hit the threshold of 60 of thousands of points in this particular benchmark. The previous record belonged to an overlocker named Lucky_n 00 b who managed to obtain respectively 47932 points and 59606 points in 3DMark Fire Strike.
AMD Radeon RX Graphics Card Test 6800 vs GeForce RTX 3070
From 25 November, the first non-reference constructions characterized by higher power limit, such as the recently presented PowerColor Radeon RX 6800 XT Red Devil, very susceptible to overclocking. Even 2, 65 GHz can be reached on this card. A YouTube user named ExtremeIT tried to crank up to 2750 and 2800 MHz (Boost), but there were problems with stability. The card works as standard with clock rates of 2090 and 2340 MHz (factory OC). We are waiting for new records. Using liquid nitrogen, the penetration of the magic 3 GHz barrier seems to be within reach.
18 Only two AMD graphics cards based on RDNA 2 architecture from the RX series 6000 debuted in November, i.e. the RX 6800 XT and 6800, although “debuted” is not quite an accurate term, because similarly to GeForce RTX 3000 their availability leaves much to be desired. The flagship Radeon RX 6900 XT will premiere in more than two weeks, on December 8, but many users are still waiting for cheaper chips. The first are to be cards from the RX series 6700, i.e. RX 6700 XT and RX 6700, based on Navi core 22. In the coming weeks, more and more spicy details about these layouts will be surfacing the web. Now we have learned a little more about the power consumption, which may be similar to that of the Radeons RX 5700.
We got to know the information about the expected power consumption of Radeon RX cards 6700 XT and RX 6700 to have 12 GB of GDDR6 memory (192 – bit). The premiere of both models is scheduled for January 2021 of the year.
AMD Radeon RX graphics card test 6800 XT vs GeForce RTX 3080
Cards will use Navi chips 22 XT (with 40 CU blocks and 2560 stream processors) and Navi 22 XTL, which will be reserved for Radeon RX 6700 XT and Radeon RX 6700 respectively. Insider Patrick Schur, who often shares accurate reports on Red chips, revealed that the predicted TGP ratio (but it does not determine the full power consumption as in the case of the competing GeForce RTX!) Of both cards is expected to vary between 186 – 211 In (RX 6700 XT) and 146 – 156 In (RX 6700). For comparison, in the case of the reference Radeon RX 5700 XT it is 180 W, which translated into TBP (Total Board Power) at the level of 225 W. Therefore, the full power consumption of the Radeons RX 6700 should slightly exceed that of the Radeon RX 5700 XT (225 W) and RX 5700 (180 W).
AMD Radeon RX Graphics Card Test 6800 vs GeForce RTX 3070
According to Radeon RX reports 6700 are to offer 12 GB of GDDR6 memory with 192 – bit interface, so compared to higher models we would get what understandable, less VRAM and a narrower rail. The premiere of the cards will take place in January 2021, earlier than originally planned. The suggested retail price of these models is not yet known. RX 6700 XT and RX 6700 should compete with GeForce RTX 3060 Ti and RTX 3060. Nvidia is also getting ready to invade, planning as many as four cards in the same month: RTX 3080 Ti (answer to RX 6900 XT), RTX 3060 12 GB (rival RX 6700), RTX 3050 Ti and RTX 3050. As a side note, it is worth adding that the Navi system 22 is also to be used in mobile GPU versions.
We have official confirmation. Marcin Momot, a representative of the Polish studio CD Projekt RED, revealed that ray tracing in Cyberpunk 2077 will not be available for cards AMD Radeon RX graphics 6000. At least it won’t be around the premiere. However, he added that the studio is working closely with AMD to provide an option for users of the new Red chips as soon as possible. We don’t know how long this may take. Reason? Probably the manufacturer wants to optimize the game for the new AMD cards and is not going to make a mistake, for example Ubisoft (see Watch Dogs: Legion). We also learned a bit more from Nvidia about Cyberpunk’s hardware requirements 2077 for ray tracing.
CD Projekt RED has confirmed that ray tracing in Cyberpunk 2077 will not be available for premiere for Radeon RX cards 6000. The developers are working with AMD to add support, but we don’t know how long users will have to get around with taste. However, we know that the requirements for RT take into account DLSS.
Cyberpunk 2077 – hardware requirements to 1440 p, 4K and ray tracing
Around ray tracing in Cyberpunk 2077 and its performance on AMD Radeon RX 6000 there is a lot of confusion. Nvidia recently revealed that it is unable to block this option for Red chips in any way. This decision belongs directly to the CD Projekt RED studio, which officially confirmed that on the day of the premiere ray tracing in Cyberpunk 2077 will be not available to users of the new RX series Radeons 6000. This is analogous to Godfall, which also uses the DXR standard and should run on GeForce RTX cards without any problems, but the RT in this game is only reserved for Radeon RX owners 6000. The developers work with AMD, but it seems that they want to optimize the game for the new Radeons first and not implement support quickly, e.g. repeating the Ubisoft and Watch Dogs: Legion error, in which ray tracing looks significantly worse than on GeForce RTX cards.
Not for the release but we are working together with AMD to introduce this feature as soon as we can .
– Marcin Momot (@Marcin 360) November 20, 2020
Cyberpunk 2077 with an extensive list of graphics settings in the PC version
In addition, Nvidia confirmed on its blog that the hardware requirements of Cyberpunk 2077 for ray tracing have DLSS enabled, although we do not know which specific mode (there are supposed to be six). This means that they are at least up to 60 frames per second. However, it is unclear whether all other computer specifications will allow this level of fluidity. It is possible that CD Projekt RED, avoiding the confirmation of the target FPS value, does not want to promise that a given hardware configuration will provide at least performance level 60 frames per second, when in fact it may not provide this in every scenario in the game. Thus, no player will later complain that “after all Cyberpunk 2077 was supposed to work in 60 FPS on such a PC! ”. The game got loud again, after all, until the premiere, set at 10 December 2020 of the year, less than three weeks left.
As recently promised, CD Projekt RED has revealed the detailed hardware requirements of Cyberpunk 2077 on the PC. The minimum and recommended configuration for resolution 1080 p (1920 x pixels) we met in mid-September. On the one hand, many players were pleasantly surprised by the relatively low requirements, but on the other hand felt disappointed. Companies like Ubisoft are starting to get used to specs that also include higher resolutions and graphics settings, so the same was expected from one of the most anticipated games in recent years. Since yesterday it has gotten again loud about the upcoming RPG and CD Projekt RED has decided to forge the iron while it’s hot, providing the updated hardware requirements of Cyberpunk 2080.
We learned the detailed hardware requirements of Cyberpunk 2077. CD Projekt RED has provided the recommended configurations for the resolution 1440 pi 4K and ray tracing. You need the GeForce RTX 3080 for maximum detail.
Cyberpunk 2077 with a rich list of graphics settings in the PC version
Creators released the hardware requirements of Cyberpunk 2077 for resolution 1440 pi 4K and ray tracing. The minimum and recommended configurations refer to low and high details in Full HD resolution. Now it is revealed that the Ultra settings in 1440 pi 4K you will need a processor comparable to those recommended for 1080 p, but paired, of course, with a much more powerful graphics card. I’m talking about GeForce RTX 2060 and Radeon RX 5700 XT to 1440 p and GeForce RTX 2080 S, RTX 3070 and Radeon RX 6800 XT to 4K. To run Cyberpunk 2077 with ray tracing set to Ultra ( Psycho?) Will be required, depending on the resolution, a GeForce RTX graphics card 3070 and RTX 3080. Here are the full PC hardware requirements for Cyberpunk 2077 on PC:
Cyberpunk minimum hardware requirements 2077 (1080 and low settings)
Cyberpunk Recommended Hardware Requirements 2077 ( 1080 and high settings)
Cyberpunk Recommended Hardware Requirements 2077 ( and Ultra settings)
Recommended Cyberpunk hardware requirements 2077 (4K and Ultra settings)
Operating system
64 – Windows bit 7 or 10
64 – Windows bit 10
64 – Windows bit 10
64 – Windows bit 10
Processor
Intel Core i5 – 3570 K 3.4 GHz or AMD FX – 8310 3.4 GHz
Intel Core i7 – 4790 3.6 GHz or AMD Ryzen 3 3200 G 3.6 GHz
Intel Core i7 – 4790 3.6 GHz or AMD Ryzen 3 3200 G 3.6 GHz
Gameplay Cyberpunk 2077: Goodbye Night City Wire with a bang
According to the studio CD Projekt RED, to enjoy the effects based on ray tracing on medium settings, all you need is the weakest GeForce from the RTX family, RTX 2060 with 6 GB of VRAM. The game will offer four effects based on ray tracing, one of which, lighting, will have several degrees of quality, the rest will have to be turned off in case of lower performance. We remind you that ray tracing in Cyberpunk 2077 which is based on on the DXR standard, it should work on both Nvidia GeForce RTX cards and the newly released Radeons RX 6000. True, the example of Godfall shows that it does not have to be that way. New AMD graphics systems are missing in the above list, but the reason may be prosaic: Cyberpunk 2077 with RT was not optimized for Radeon RX 6000. How the revealed hardware requirements of Cyberpunk 2077 will translate into reality, we will see 10 December this year.
João Silva 6 hours ago Featured Tech News, Graphics
Now that the Radeon RX 6800 series has launched, the rumour mill has started to focus more on the Radeon RX 6700 series. The latest rumours detail some specifications of the Navi 22 XT and Navi 22 XTL GPUs, including the memory capacity and TGP.
According to new information from Patrick Schur, the AMD Radeon RX 6700 XT graphics card will be powered by the Navi 22 XT GPU, which has a TGP ranging from 186W to 211W. The Navi 22 XTL that it’s expected to feature in the Radeon RX 6700 has a 146-156W TGP, but Patrick Schur states that the lower TGP limit might be less than 146W. Both cards should come with 12GB of GDDR6 memory. Compared to the 225W TGP of the Radeon RX 5700 XT and the 180W TGP of the Radeon RX 5700, power consumption has reduced significantly.
It’s also possible that NV22 XTL has a lower min. TGP than 146 W. I’m not 100 % sure about that. So far I haven’t seen any NV22 XTL with less than 146 W TGP.
— Patrick Schur (@patrickschur_) November 21, 2020
Regarding the remaining specifications of the Navi 22 XT GPU, the Linux Kernel leak suggests it should feature 40 CUs (2560 stream processors) and a 192-bit memory bus, which fits the alleged 12GB of VRAM. The Navi 22 XTL GPU should also feature a 192-bit bus, but if it’s a cut-down Navi 22 GPU, the CU count should be lower than 40 but higher than 32, lining up with the expected CU count of the Navi 23 GPU.
The Navi 22-based graphics card are expected to launch in early 2021.
KitGuru says: Are you waiting for mid-range AMD cards before buying a new system or upgrading your current setup?
Become a Patron!
Check Also
Apple to bring PS5 DualSense and Xbox Series X controller support to iOS
The Xbox Series X/S have been out for almost two weeks now and with that, …
João Silva 7 hours ago Featured Announcement, Featured Tech News, Software & Gaming
CD Projekt Red listed the system requirements for the Cyberpunk 2077 some time ago, but we had to wait until Friday November 20th for the recommenced specifications for ray tracing. AMD Radeon GPUs aren’t mentioned in these system requirements however, and CD Projekt Red’s Global Community Lead, Marcin Momot, has confirmed that ray tracing won’t be supported on AMD cards at release.
Despite that, CD Projekt Red community manager Marcin Momot has stated that AMD and CD Projekt Red are “working together…to introduce this feature” as soon as possible. Regardless, this is quite surprising given that Cyberpunk 2077 is using Microsoft’s DXR API which is supported by AMD’s RX 6000-series of GPUs.
Ending this week on a high note! Here are the detailed #Cyberpunk2077 system requirements we mentioned some time ago. pic.twitter.com/Deh5JvuzVy
— Marcin Momot (@Marcin360) November 20, 2020
At launch, Cyberpunk 2077 will support both ray tracing and DLSS when used with Nvidia RTX GPUs. This is the opposite of what happened with the recently launched Godfall, where ray tracing features are currently only available on AMD Radeon graphics cards, leaving Nvidia GPUs out for now. It seems that besides timed-exclusive video games and content, we now also have timed-exclusive ray tracing support.
Additionally, AMD GPU owners won’t be alone in waiting for these enhanced visuals. Next-gen consoles will be able to run Cyberpunk 2077 as soon as it launches, but ray tracing will only come to said consoles via an update at a later date.
Discuss on our Facebook page HERE.
KitGuru says: Are you waiting for Cyberpunk 2077? Do you own an AMD Radeon RX 6000 card? Will the lack of ray tracing support on AMD Radeon RX 6000 cards affect you?
Become a Patron!
Check Also
Ubisoft aware some PS5 owners having issues upgrading PS4 games
Ubisoft has been supporting free current-gen to next-gen upgrades on its games this holiday season …
João Silva 8 hours ago Featured Tech News, Graphics
With the Radeon RX 6800 XT now out in the wild, professional overclockers are racing to set some new world records. We already saw a 3DMark Firestrike record broke last week, but just a few days later, a new world record has been set for the highest 3DMark Firestrike overall score.
The run from Takukou involved a system powered by an overclocked AMD Ryzen 9 5950X running at 5.6GHz using liquid nitrogen paired with 32GB of G.Skill 3800-DDR4 TridentZ Royal modules with 14-14-14-28 timings. The graphics card was a liquid nitrogen-overclocked AMD Radeon RX 6800 XT with the GPU running at 2800MHz.
The new world record set by Takukou is 49456 marks, with a graphics score of 62184 points. Compared to the previous word record, both runs have a difference of 1524 marks in the overall score. Unlike “LUCKY_NOOB”, we don’t know if Takukou has modified tessellation in 3DMark Firestrike, but he mentioned that no “power modding” was made to achieve this score.
Given how fast these records are being broken, and the fact that we still have AiB cards on the way, we could end up seeing someone breach a 50K score on Firestrike before the end of 2020.
KitGuru says: Are you impressed with these scores? How much overclock potential will the custom RX 6800 XT cards offer over the reference card?
Become a Patron!
Check Also
Apple to bring PS5 DualSense and Xbox Series X controller support to iOS
The Xbox Series X/S have been out for almost two weeks now and with that, …
We use cookies on our website to give you the most relevant experience. By clicking “Accept”, you consent to the use of ALL the cookies.
This website uses cookies to improve your experience while you navigate through the website. Out of these, the cookies that are categorized as necessary are stored on your browser as they are essential for the working of basic functionalities of the website. We also use third-party cookies that help us analyze and understand how you use this website. These cookies will be stored in your browser only with your consent. You also have the option to opt-out of these cookies. But opting out of some of these cookies may affect your browsing experience.
Necessary cookies are absolutely essential for the website to function properly. This category only includes cookies that ensures basic functionalities and security features of the website. These cookies do not store any personal information.
Any cookies that may not be particularly necessary for the website to function and is used specifically to collect user personal data via analytics, ads, other embedded contents are termed as non-necessary cookies. It is mandatory to procure user consent prior to running these cookies on your website.
