Microsoft reveals detailed specs for its Xbox Series X Console - RDNA 2 confirmed!

It's looking very good, now offer the ability to run a Keyboard and Mouse in any game, the ability to use a Hotas like the X56 or Warthog etc and also the ability to have your audio output via a 3.5mm or phono connection from the rear of the Xbox and for me it will be a true pc replacement for just gaming.


Without that, I will probably still get one but the PC will still be my main platform of choice.
 
Didn't really learn much more than what was already known tbh. HDMI 2.1 and VRS were the newest things. Everything else was a given

Now it's Sony's turn. Should be interesting
 
Last edited:
Didn't really learn much more than what was already known tbh. HDMI 2.1 and VRS were the newest things. Everything else was a given

Now it's Sony's turn. Should be interesting

Yes, a lot of this is just confirmation for existing rumours, but confirmation is confirmation. The 12 TFLOPS number has been flying around for ages.

The big question is, when will AMD release an RDNA 2 GPU to the public? The Xbox Series X has AMD's strongest (announced) gaming graphics chip.
 
Well TFLOPS isn't a great unit to measure performance. So to me it's meaningless. AMD has had multiple slower cards with far more than Nvidia cards in TFLOPS yet doesn't translate to gaming performance. Its such a broad "measurement" and is just a simple equation. Not even a test.
The formula is:
Cores x Clock Speed in Hertz x Floating Point Operations per clock cycle / One Trillion.

While xbox is enticing so far, I could just get xbox game pass and play all the games on Windows then just play PS on PS. Not much sense getting an xbox imo as of now.
 
Last edited:
Well TFLOPS isn't a great unit to measure performance. So to me it's meaningless. AMD has had multiple slower cards with far more than Nvidia cards in TFLOPS yet doesn't translate to gaming performance. Its such a broad "measurement" and is just a simple equation. Not even a test.
The formula is:
Cores x Clock Speed in Hertz x Floating Point Operations per clock cycle / One Trillion.

While xbox is enticing so far, I could just get xbox game pass and play all the games on Windows then just play PS on PS. Not much sense getting an xbox imo as of now.

TFLOPS is only useful for comparing graphics cards with the same, or similar architectures. RDNA 1 had an average IPC boost of 25% over GCN in their internal tests, which makes this comparison kinda silly.

Even so, a 2x boost in TFLOPS over Xbox One X is big. Add the IPC boost and its better. Than add features like VRS and rapid packed math and developers will be able to get a lot more than that.

Even if RDNA 2 added no IPC boost and just feature upgrades, the Xbox Series X would be 2.5x stronger than Xbox One X (25% IPC boost + 2x TFLOPS) in terms of average compute performance. Add VRS etc and we have a huge performance leap.

Hopefully, RDNA 2 will add some extra IPC into the mix, alongside the already confirmed features like VRS and raytracing. The Xbox Series X and PS5 are going to be great systems, especially now that AMD's Jaguar cores aren't holding it back CPU-wise.
 
Last edited:
Again using TFLOPs is wrong and shouldn't be considered in your 2.5x as an argument. Just as you said its silly to even make the comparison. TFLOPS isn't useful for anything. Even amongst its own architecture. What the heck is the point? It's literally only useful for ONE comparison and that comparison is useless. Just looking at the architecture details tells you the same thing.

It is the architecture that matters here, not TFLOPs. We don't have any way of knowing that until RDNA 2 releases on PC and even then it's not going to be an exact value in any terms we can find compared to old architectures as for consoles they will be slightly different.


It'll be far faster. That's obvious but media definitely isn't doing any favors here by trying to quantitate any performance metrics as it's misleading.
 
Last edited:
I do kinda agree with NBD but it all depends on how they sourced the figure, and if it's stated. FLOPS doesn't have to be a theoretical number, you can of course find it experimentally(Run a set number of floating point ops and divide that by the time taken in seconds to complete), but even then the actual or average number will usually vary significantly depending on workload, vector size, ect, you can find an average from these or use an averaged workload, but this is probably a theoretical max quoted here tbf.

But even then, the theoretical max for what precision? The traditional assumption would be FP32, but if they were being crafty, this could be an FP16 number, a 16-bit Floating point op is still a FLOP tbf, and the Xbone didn't have "rapid packed math" (Double rate FP16).
 
Last edited:
I do kinda agree with NBD but it all depends on how they sourced the figure, and if it's stated. FLOPS doesn't have to be a theoretical number, you can of course find it experimentally(Run a set number of floating point ops and divide that by the time taken in seconds to complete), but even then the actual or average number will usually vary significantly depending on workload, vector size, ect, you can find an average from these or use an averaged workload, but this is probably a theoretical max quoted here tbf.

But even then, the theoretical max for what precision? The traditional assumption would be FP32, but if they were being crafty, this could be an FP16 number, a 16-bit Floating point op is still a FLOP tbf, and the Xbone didn't have "rapid packed math" (Double rate FP16).

Exactly. It's such a vague measurement of performance as it is entirely theory-based.
 
RX 5700 XT - 9.8 TFLOPS - Summer 2019 - $500

Xbox Series X - 12 TFLOPS - Christmas 2020 - $500

Wow, sad for early adopters... I hope RDNA2 dGPUs are far more interesting.
 
RX 5700 XT - 9.8 TFLOPS - Summer 2019 - $500

Xbox Series X - 12 TFLOPS - Christmas 2020 - $500

Wow, sad for early adopters... I hope RDNA2 dGPUs are far more interesting.

See my point about misleading people? Look how fast that was.

@jcc, don't bother comparing the TFLOPs. Literally means nothing here, especially in your context.
 
To be fair, RDNA 2 is almost certainly going to have an equivalent or better level of rendering efficiency(Though of course this still depends on other external factors including memory bandwidth), at least for traditional shader based workloads, and so an increase if it's an increase in 32-bit FLOPs it does almost guarantee an increase in end performance against an RDNA1 card, assuming other aspects & workload are equal. Even then they can usually eek a little more out on a console.
 
To be fair, RDNA 2 is almost certainly going to have an equivalent or better level of rendering efficiency(Though of course this still depends on other external factors including memory bandwidth), at least for traditional shader based workloads, and so an increase if it's an increase in 32-bit FLOPs it does almost guarantee an increase in end performance against an RDNA1 card, assuming other aspects & workload are equal. Even then they can usually eek a little more out on a console.

My point was just using TFLOPs as any kind of metric is wrong and with people putting articles out there just comparing TFLOPs as if it's is equal across all architectures is wrong. The fact so many people are putting articles out like this informing the less informed (we can all agree generally console players are less tech savy and therefore more easily gullible) is awful awful journalism.

Which is exactly what was done above. Even if they are similar doesn't mean it's an equal metric for both. It's not. Yes we know RDNA 2 will be better but still it's different and 10 TFLOPS of performance from RDNA 1 != RDNA 2 performance.
 
Last edited:
sony

I don't think sony has anything to worry about The Games are where playstation stands out .

I play my ps4 pro on a 1080p monitor , i would like to see freesync support and unlocked refresh rate come to playstation and 1440p support.

Not intrested in xbox has i can play there games on my pc.


I AM NOT A FANBOY, xbox 360 was my platform back in the day
 
I don't think sony has anything to worry about The Games are where playstation stands out .

I play my ps4 pro on a 1080p monitor , i would like to see freesync support and unlocked refresh rate come to playstation and 1440p support.

Not intrested in xbox has i can play there games on my pc.


I AM NOT A FANBOY, xbox 360 was my platform back in the day

I'm not sure you can make statements like that anymore. How many times I heard that Nvidia dont have to worry about AMD, or Intel have nothing to fear about AMD... Look at the CPU decline of Intel now.

XGP has been a massive success for MS. The Series X could be a true winner, given they learn from the mistakes made by Xbox in the past with lack of exclusivities. As you say, its all about the games in the end but we know very little right now, and well PS Now was a flop.
 
RX 5700 XT - 9.8 TFLOPS - Summer 2019 - $500

Xbox Series X - 12 TFLOPS - Christmas 2020 - $500

Wow, sad for early adopters... I hope RDNA2 dGPUs are far more interesting.

RDNA2 in the XBSX is still 7nm just like RDNA1 but with that comes a refinement in production process making it far cheaper to manufacture, Plus we don't know if MS are counting FP16 or FP32, XBSX could still have 6TFLOPS of FP32 but have the ability to use FP16 making it 12TFLOPS.

Gotta wait until it comes out to have any viable opinion.
 
I'm not sure you can make statements like that anymore. How many times I heard that Nvidia dont have to worry about AMD, or Intel have nothing to fear about AMD... Look at the CPU decline of Intel now.

XGP has been a massive success for MS. The Series X could be a true winner, given they learn from the mistakes made by Xbox in the past with lack of exclusivities. As you say, its all about the games in the end but we know very little right now, and well PS Now was a flop.


I still think Sony will be fine, nintendo switch does ok and that doesn't have 12 teraflops. LOL
 
I still think Sony will be fine, nintendo switch does ok and that doesn't have 12 teraflops. LOL

Agreed, Microsoft would have to be very disruptive to deal a lot of damage to Sony. Even with this gen, Microsoft made a lot of money. Sony may have "won" this console generation, but it is hard to consider Microsoft losers.

As far as the Switch goes, the fact that it is a handheld is a big factor. When you are using a small 720p screen it doesn't need 12 TFLOPS of performance.
 
Back
Top