Xbox Scarlett APU Die Shot Analysis: Die Size Estimated at 401 mm2

  • 181 results
  • 1
  • 2
  • 3
  • 4
Avatar image for tormentos
tormentos

29703

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#101 tormentos
Member since 2003 • 29703 Posts

@BlackShirt20 said:

@lundy86_4: You sure? You just say the word and I’ll talk to a mod.

Let me get this right you DUSTED this low post ALT account from the 7th bowels of hell to complain with mods about poster who see HOLES in your weak arguments?

Wow man new low for lemmings,if you don't want to see peoples arguments against yours,go open a private forum of your own and post there.

@ronvalencia said:

1. Your argument that demishes Microsoft claim on both Polaris and Vega IPs are patently false.

2. Without Vega RPM usage, Vega CU acts like Polaris CU i.e. it's like any other GCN with degraded IPC when compared to Hawaii based R9-390X. The large gap between PS4 Pro and X1X indicates Hawaii GCN+ like IPC vs PS4 Pro's Polaris/Vega IPC. R9-390X OC (6 TFLOPS) can deliver similar large gap (>49 percent) against RX-470 (4.9 TFLOPS).

R9-390X OC has issues with heavy blends and alpha effects which X1X GPU has mitigated.

3. ROPS with hyper fast multi-MB cache is one of secret sauce for NVIDIA's Maxwell/Pascal GPUs. NVIDIA updated L2 cache design for Turing by doubling the storage size i.e. 2MB (GP104) to 4MB (TU104), 3MB (GP102) to 6MB (TU102). On GPUs, NVIDIA has higher design credibility when compared to AMD's RTG.

MS never come clean with the supposed Vega Features,sony did the xbox one X was POLARIS with a so call Vega feature that MS never confirmed what was.

Even the Pro was more Vega than the xbox one X.

And i see you iognored my post the PS4 and xbox one both are sea island and the PS4 commanded more than 100% gap in several games vs the xbox one,it mean nothing,just like the xbox one X command bigger gaps than the 45% gap in power show,the PS4 did the same with the xbox one SINCE DAY 1.

In fact resolution gate was just that.

Avatar image for Pedro
Pedro

37209

Forum Posts

0

Wiki Points

0

Followers

Reviews: 65

User Lists: 0

#102 Pedro
Member since 2002 • 37209 Posts

Lol at Ron nonsensical no answer rants to simple questions. Classic.

Avatar image for lundy86_4
lundy86_4

54956

Forum Posts

0

Wiki Points

0

Followers

Reviews: 2

User Lists: 0

#103 lundy86_4
Member since 2003 • 54956 Posts

@BlackShirt20: Sure. Go nuts lol.

Avatar image for drlostrib
DrLostRib

5216

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 5

#104 DrLostRib
Member since 2017 • 5216 Posts

@ronvalencia said:
@fedor said:

If you continue Scorpio doesn't have Vega debate, I'll apply precedent applied by Juub1990. Removing you from this topic.

lol, wtf does that even mean

Avatar image for Pedro
Pedro

37209

Forum Posts

0

Wiki Points

0

Followers

Reviews: 65

User Lists: 0

#105 Pedro
Member since 2002 • 37209 Posts

@drlostrib: Fedor goes into time out.🤣

Avatar image for goldenelementxl
GoldenElementXL

3886

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#106 GoldenElementXL
Member since 2016 • 3886 Posts

We need the next gen specs to be released ASAP. This shit is exhausting

Avatar image for EG101
EG101

2025

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#107 EG101
Member since 2007 • 2025 Posts

If the rumored Specs are true then a difference of 2.8 TF's is in fact the largest Gap ever between same gen consoles.

Using Percentages to make a point is incorrect. 2.8TF's is Factually, Objectively the largest gap between consoles aiming for the same consumers.

The thing is with Ray Tracing slowing down the GPUs so much those 2.8 TF's are going to be more important than ever....

Avatar image for xantufrog
xantufrog

12229

Forum Posts

0

Wiki Points

0

Followers

Reviews: 2

User Lists: 5

#108  Edited By xantufrog  Moderator
Member since 2013 • 12229 Posts

@EG101: that's a strange way of looking at it. I have two vintage vespas, 6 vs 12 horsepower - one double, with a sizeable difference in the rider's experience. Nowadays I might also compare cars of 180 vs 220 - this is a larger gap in raw numbers, yes, but it's not double the horsepower and the differencd in experience is nowhere near comparable.

This is the same. As technology evolves, the numbers will keep growing, pushing the scale which makes it more likely one could make your argument and yet it might be meaningless. The % matters.

130 TFLOPS vs 120... a whopping 10 TFLOPS gap! That nobody will care about because the ratio is minute

Avatar image for Pedro
Pedro

37209

Forum Posts

0

Wiki Points

0

Followers

Reviews: 65

User Lists: 0

#109  Edited By Pedro
Member since 2002 • 37209 Posts

@xantufrog said:

@EG101: that's a strange way of looking at it. I have two vintage vespas, 6 vs 12 horsepower - one double, with a sizeable difference in the rider's experience. Nowadays I might also compare cars of 180 vs 220 - this is a larger gap in raw numbers, yes, but it's not double the horsepower and the differencd in experience is nowhere near comparable.

This is the same. As technology evolves, the numbers will keep growing, pushing the scale which makes it more likely one could make your argument and yet it might be meaningless. The % matters.

130 TFLOPS vs 120... a whopping 10 TFLOPS gap! That nobody will care about because the ratio is minute

Are you saying that all this performance of TFLOPS means nothing? ;)

Avatar image for ronvalencia
ronvalencia

28555

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#110  Edited By ronvalencia
Member since 2008 • 28555 Posts

@tormentos said:

@ronvalencia said:

1. Your argument that demishes Microsoft claim on both Polaris and Vega IPs are patently false.

2. Without Vega RPM usage, Vega CU acts like Polaris CU i.e. it's like any other GCN with degraded IPC when compared to Hawaii based R9-390X. The large gap between PS4 Pro and X1X indicates Hawaii GCN+ like IPC vs PS4 Pro's Polaris/Vega IPC. R9-390X OC (6 TFLOPS) can deliver similar large gap (>49 percent) against RX-470 (4.9 TFLOPS).

R9-390X OC has issues with heavy blends and alpha effects which X1X GPU has mitigated.

3. ROPS with hyper fast multi-MB cache is one of secret sauce for NVIDIA's Maxwell/Pascal GPUs. NVIDIA updated L2 cache design for Turing by doubling the storage size i.e. 2MB (GP104) to 4MB (TU104), 3MB (GP102) to 6MB (TU102). On GPUs, NVIDIA has higher design credibility when compared to AMD's RTG.

MS never come clean with the supposed Vega Features,sony did the xbox one X was POLARIS with a so call Vega feature that MS never confirmed what was.

Even the Pro was more Vega than the xbox one X.

Notice the differences

X1X GPU

Geometry and TMU has access 2MB L2 cache

ROPS has access to 2MB render cache

Total: 4MB below L1 cache.

-----

RX Vega 56/64

Geometry, TMU and ROPS has access 4MB L2 cache

Total: 4MB below L1 cache.

------

Polaris 10/20/30

Geometry and TMU has access 2MB L2 cache

ROPS has access to tiny KB render cache (no update from previous GCN generation)

Total: ~2MB below L1 cache

PS4 Pro GPU acts like typical Polaris with 4.2 TFLOPS like lesser RX-470.

---

Facts from MS

  • Claimed both Polaris and Vega IP.
  • Vega RPM feature was NOT selected.

@tormentos said:

And i see you iognored my post the PS4 and xbox one both are sea island and the PS4 commanded more than 100% gap in several games vs the xbox one,it mean nothing,just like the xbox one X command bigger gaps than the 45% gap in power show,the PS4 did the same with the xbox one SINCE DAY 1.

In fact resolution gate was just that.

That's a false narrative i.e. that's fake news about my arguments.

XBO has unstable performance due to 68 GB/s DDR3-2133 and 32 MB ESRAM divide and my argument with W5000 (~1.3 TFLOPS) is for best case which shows being insufficient to rival 7850 (1.76TFLOPS) and slightly higher rated R7-265 (1.89 TFLOPS).

7790 doesn't reflect XBO's transparent split render buffer mode which automatically load balances between 68 GB/s DDR3-2133 and 32 MB ESRAM memory pools for 1080p frame buffers.

Creating 1080p framebuffers doesn't solve CU bound issues. I did NOT dismiss XBo vs PS4 gap, I'm telling you XBO can get closer to PS4 in accordance to TFLOPS difference when programmed correctly.

There's TMU read/write workaround for XBO's 16 ROPS read/write issue with lesser data types.

32 ROPS needs memory bandwdith similar to X1X's 326 GB/s level LOL. Beyond X1X's 326 GB/s memory bandwdith, 48 or 64 ROPS switch would be the right choice.

The main reason for Hawaii's 64 ROPS are due to near brain dead two Pitcairns glued together design.

RGBA10F(quad packed FP10) : 853 Mhz x 16 * 5 bytes = 68.2 GB/s (semi-custom ROPS with FP10). XBO's 68 GB/s DDR3 memory bandwdith was targeting FP10!

X1X's ROPS

RGBA8 (quad packed INT8): 1172 Mhz x 32 * 4 bytes = 150 GB/s (X1X's GPU ROPS has 2MB render cache for micro-tile render)

RGBA16F (quad packed FP16): 1172 Mhz x 32 * 8 bytes = 300 GB/s (X1X's GPU ROPS has 2MB render cache for micro-tile render), very common usage from game console optimization guides.

RGBA32F(quad packed FP32) : 1172 Mhz x 32 * 16 bytes = 600 GB/s (X1X's GPU ROPS has 2MB render cache for micro-tile render)

MS claimed XBO has semi-custom support for X360's FP10 format (e.g. RGBAF10) which doesn't exist on normal PC GCN ROPS.

RGBA16F is common across multi-platforms.

DCC can reduce memory bandwidth consumption and may not work well with transparencies.

PS4's 32 ROPS will be memory bandwdith bound. Cows 16 ROPS vs 32 ROPS arguments are flawed!

My argument to debunk Cows 16 ROPS vs 32 ROPS arguments doesn't solve 1.3TFLOPS vs 1.84 TFLOPS problem.

Your argument detracts from this topic's Scarlet hardware. Why the f**k are you are discussing the obsolete XBO, when I'm NOT even interested buying this POS?

For RDR2 example

XBO renders at 1536 x 864 with 1.3TFLOPS which is close to expected TFLOPS difference from PS4. XBO has 70 percent of PS4's TFLOPS. XBO's 1536 x 864 resolution is ~64 percent of PS4's 1920 x1080 resolution.

PS4 renders at 1920 x1080 with 1.8 TFLOPS

PS4 Pro renders at 1920 x 2160 (half 4K pixels) with 4.2 TFLOPS

PS4 Pro render RDR2 in accordance TFLOPS difference from PS4!

X1X renders at 3,840‬ x 2160 (full 4K pixels) with 6 TFLOPS or 8 TFLOPS effective GCN v1.1

RDR2 scales it's resolution largely based on GCN v1.1 TFLOPS power until X1X's results.

Avatar image for ronvalencia
ronvalencia

28555

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#111  Edited By ronvalencia
Member since 2008 • 28555 Posts

@fedor said:

@ronvalencia: "2 unconfirmed spec lists and an Nvidia GPU who's architecture produces far better performance than RDNA? What is your case exactly?"

I notice you also avoided answering any questions or providing any sources to your claims, probably because you simply have no clue what you're talking about. How does it feel being ignorant?"

This is how it started, actually. My post to someone else, (that's completely accurate) you then jumped in with your typical tangent.

"You derailed my topic."

Lmao!!! What an incredible reach. It's hilarious one little quote of mine (which you still haven't disproved, you can't) got you all up in arms. In fact you posted an article proving me right and have been having a meltdown ever since. No giant wall of text will change that, Ron. My post to Black shirt was 100% true, you should have just kept quiet and not looked like a lunatic.

You "derailed" your own thread.

My point between you and Blackshirt is NVIDIA GPUs such as TU102 and TU104 has classic GPU hardware superiority which are mostly involves with preparing and feeding data to the CUs.

Adding TFLOPS are nearly nothing without scaling I/O properly. I'm telling you the main reasons why AMD's higher TFLOPS has diminishing returns from Hawaii GCN. NAVI 10 effectively restored Hawaii's IPC vs raster ratio.

TU104 has six GPCs which NAVI 10's quad geometry/rasterization units can't match!

Bakground context

NVIDIA six GPC framework enables NVIDIA to scale TFLOPS with minimal raster bottlenecks while AMD runs into quad raster engine bottlenecks since Fury X.

TU102's higher TFLOPS are showing it's diminishing returns relative to TU104.

Expect 8 GPCs with Ampere RTX 3080 Ti i.e. scale TU104's six GPCs with 256 bit bus into eight GPCs with 384 bit bus. This is not including extra TFLOPS and higher clock speeds from 7nm.

NAVI 10's classic GPU hardware is about TU106's three GPCs at slightly higher clock speeds.

Raster engine deals with converting floating point geometry into integer pixels which is one of the core idea for being a GPU separated from DSP (e.g. CELL's SPE).

----------------------

My point, IF XSX GPU has three shader engines and 12 geometry inputs (vs NAVI 10's duel shader engines and 8 geometry inputs) then XSX would be targeting RTX 2080/2080 Super level.

The ability for XSX riving RTX 2080 Super is dependant on X1X GPU's classic GPU hardware scaling i.e. adding higher TFLOPS doesn't complete the solution.

Avatar image for ronvalencia
ronvalencia

28555

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#112  Edited By ronvalencia
Member since 2008 • 28555 Posts

@EG101 said:

If the rumored Specs are true then a difference of 2.8 TF's is in fact the largest Gap ever between same gen consoles.

Using Percentages to make a point is incorrect. 2.8TF's is Factually, Objectively the largest gap between consoles aiming for the same consumers.

The thing is with Ray Tracing slowing down the GPUs so much those 2.8 TF's are going to be more important than ever....

It depends on XSX GPU's configuration

Configuration 1

Three Shader Engines with 12 geometry inputs (backface culled to 6 geometry output)

60 CU and 96 ROPS from three NAVI 10 like Shader Engine 20 CU and 32 ROPS config.

Configuration 2

Duel Shader Engines with 8 geometry inputs (backface culled to 4 geometry output, like NAVI 10)

30 CU and 64 ROPS from duel Shader Engine 30 CU and 32 ROPS config.

Configuration 2 would have TFLOPS diminishing returns.

Avatar image for tormentos
tormentos

29703

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#113 tormentos
Member since 2003 • 29703 Posts

@ronvalencia:

Ron go cry about elsewhere and take your excuses there,were there 720p games on Xbox launch that were 1080p on PS4?

Yes they were and after launch as well,that is all I care the end results the PS4 had gaps bigger than 100% when the power gap was 40%.

So yeah it's normal.

Avatar image for tormentos
tormentos

29703

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#114 tormentos
Member since 2003 • 29703 Posts

@EG101:

This is bullshit and the reason some of you are a joke,you people try to do this same shit with the Xbox one X,trying to blow up the 1.8tf difference using the flop number rather than the REAL % difference,fact is the gap is the smallest between this gen machines,performance is relative by % the number can seem big because you are comparing it to the older weaker machines.

Example you people use to say the gap between the Xbox one X and pro was a complete PS4 of difference vs the gap of the PS4 vs the Xbox one on 2013 which was like 500+ gflops,but % speaking the gap was almost the same,and in fact in both cases the Xbox one X double the pro in resolution,and the PS4 did the same with the Xbox one since launch in 2013.

The gap was basically the same regardless of the 1.8TF number been more than 3 times bigger than the gap between the first 2 consoles.

Is lol worthy at best see you people try to use the whole TF number instead of the actual % The gap would be smaller if this end been true.

Avatar image for deactivated-5e20304971dd5
deactivated-5e20304971dd5

67

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 5

#115 deactivated-5e20304971dd5
Member since 2020 • 67 Posts

@tormentos said:

@BlackShirt20:

Is funny how 20% is now HUGE to lemmings on 2013 40% was nothing.

But considering that on the 360 generation many lemmings make a huge deal over a 720p vs 640p resolution difference it doesn't really surprise me at all.

Lemmings 40% is nothing when the PS commands,when the Xbox command 20% is HUGE.🤷🤦

How bad is your math... It's 40% (PS4 v. XBO) back then and 30% now (PS5 v. XSX), not 20%. Not that it matters, we don't officially know PS5's power nor that of the XSX. We'll just have to wait for the reveal.

Avatar image for BlackShirt20
BlackShirt20

1357

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#116  Edited By BlackShirt20
Member since 2005 • 1357 Posts

@tormentos: I am gonna take this slow on you. The difference in power is a lot more than 20%. Don’t get my statement about 20% being a huge difference confused. That wasn’t what I was saying at all.

I was explaining to Mr.PC that his assumption was wrong. He claimed that the Xbox Series X is only 10% more powerful than the PS5 which when he ran benchmarks resulted in only a few extra FPS. 3-6 frames to be exact.

I explained to him his math was wrong. He than said the difference was only about 20% and he admitted later his math was off. That is when I said that the difference between the consoles is huge based off the 20%.

Now, the difference in horse power is actually 30%. I was only pointing out to the PC homer that even though his math is still off, 20% extra power will result in a lot more than just a few FPS like he claimed. It would mean higher settings, at better resolution at more stable frames.

I understand you’re upset about the power difference. That is why you’re making excuses and being defensive. Pointing out the differences from past generations. It’s also why you keep repeating yourself about wanting to wait for official specs. Because you’re hoping the power difference is a lot closer. But as more and more leaks come to light and creditable sources like DF and tech power keep confirming, it’s getting harder for you to hold on to that hope.

So here it goes. Even though the PS5 is weaker, doesn’t mean it’s not a good system. It is. It’s a very good system. I will be getting one. It just won’t be as good as Xbox. That’s all I am saying. But I never made the claim that the Xbox Series X is only 20%. I just used Mr.PC’s math and analogy and called him out for saying the a 20% difference in power isn’t much of a difference. The power difference is a lot bigger than 20% and will be clearly noticeable.

That is all.

Avatar image for fedor
Fedor

6015

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#117 Fedor
Member since 2015 • 6015 Posts

@BlackShirt20: The new PS5 leak rumors have it now at 10.25 tflop and an identical CPU clock to the Series X. Now these, like any leaks are unconfirmed and potentially untrue. Though since you believe every leak is 100% fact, you must now accept that the PS5 is stronger than you originally anticipated and that the gap in performance has dwindled significantly.

Avatar image for BlackShirt20
BlackShirt20

1357

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#118  Edited By BlackShirt20
Member since 2005 • 1357 Posts

@fedor: That is an old leak (months old). Not new. Trust me, I have watched all the PS5 you tubers. It’s the old sources saying they stand by there initial specs. Even so. That would be 20% which would still be noticeable. As I said, if it is as little as 20% that is still enough to hold frames without dropping resolution and settings.

Xbox Series X will be the the standard that the PS5 is compared to. Just like the Xbox One X is the standard that the PS4 Pro is compared to. Just like the original PS4 was the standard that the original Xbox One was compared too.

Xbox Series X will be the premier console.

Avatar image for fedor
Fedor

6015

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#119  Edited By Fedor
Member since 2015 • 6015 Posts

@BlackShirt20: It is not an old leak, it came out during CES. YouTube is also not a source. Once again, since you believe leaks are 100% true you must now accept that the PS5 is 10.25 tflops with an identical processor clock and core/thread count. It also makes it a 15% difference, not 20%.

Avatar image for BlackShirt20
BlackShirt20

1357

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#120  Edited By BlackShirt20
Member since 2005 • 1357 Posts

@fedor: I trust DF and TechPower. Which if you saw DF video the actual TF number is 12.19 (12.2), with TechPower also confirming the math that DF did. That’s still 20%.

However, I don’t care. I’m just pointing out to Tomato that Xbox Series X will be the best. Sorry Bud.

Avatar image for fedor
Fedor

6015

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#121 Fedor
Member since 2015 • 6015 Posts

@BlackShirt20: They didn't do math, sorry bud. They just reiterated leaks. Leaks that they specifically state in the DF video are complete speculation. But nevertheless, you believe leaks are 100% true so therefore you now believe PS5 is 10.25 tflop.

Avatar image for BlackShirt20
BlackShirt20

1357

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#122  Edited By BlackShirt20
Member since 2005 • 1357 Posts

@fedor: Sorry bud, but you didn’t say anything relevant to my argument.

Avatar image for fedor
Fedor

6015

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#123 Fedor
Member since 2015 • 6015 Posts

@BlackShirt20: I did actually. I just destroyed your claim that they "did the math".

Avatar image for BlackShirt20
BlackShirt20

1357

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#124 BlackShirt20
Member since 2005 • 1357 Posts

@fedor: DF did do the math. Next false narrative.

Avatar image for fedor
Fedor

6015

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#125  Edited By Fedor
Member since 2015 • 6015 Posts

@BlackShirt20: They didn't, I welcome you to time stamp the video though showing them doing it though. The most they do is say it's possible that it's 12 tflop.

https://wccftech.com/xbox-scarlett-apu-die-shot-analysis-die-size-estimated/

"Scorpio had 2560 SPs with 2 CUs disabled and while the Xbox Scarlett APU is clearly bigger, it is not possible to speculate based just on the die size because of the different process node involved (7nm vs 16nm) but we can safely say that you are looking at at least 50% more power in the same die space. Since the Scorpio APU can output roughly 6 TFLOPs of power, we can guesstimate a range of at least 9 TFLOPs for the Scarlett APU ( up to 12 TFLOPS is within the realm of possibility)."

More PS5 leaks

https://wccftech.com/playstation-5-9-2-teraflops-not-accurate/

Avatar image for ronvalencia
ronvalencia

28555

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#126  Edited By ronvalencia
Member since 2008 • 28555 Posts

https://seekingalpha.com/article/4316089-amd-shines-intel-struggles-ces

The mobile APU die size is about 150mm2 per estimates from AnandTech, which makes the laptop APU a high-yielding part

AMD's 8 cores Ryzen 4000 APU has about 150 mm2 which includes 8 cores Zen 2 with 8 MB L3 cache, Northbridge (PCI-E 4.0 I/O, two 64 bit DDR4 memory controllers) and improved Vega IGP.

Vega IGP consumes about half of APU die area.

Potential XSX configuration

8 core Zen 2 with 8 L3 cache with northbridge ~= 75 mm2

WIth 400 mm2 chip budget minus 75 mm2 yields 325 mm2 size GPU.

Game consoles doesn't need PC APU's PEG (PCI-Express Graphics I/O) support, hence Northbridge can be smaller.

Similar speculation can be applied for PS5's APU.

Avatar image for BlackShirt20
BlackShirt20

1357

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#127 BlackShirt20
Member since 2005 • 1357 Posts

@fedor:

https://youtu.be/0PqMj6aSYH0

https://youtu.be/v2M01ph1VHM

It’s in one of those Frodo. I have better things to do than rewatch 30min of football. But the first one breaks down the math while the die video “pretty much backs up what they mentioned in the first video which seems to confirm 12TF (12.19 to be exact)”.

Feel free to continue to cry.

Avatar image for ronvalencia
ronvalencia

28555

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#129  Edited By ronvalencia
Member since 2008 • 28555 Posts

@tormentos said:

@ronvalencia:

Ron go cry about elsewhere and take your excuses there,were there 720p games on Xbox launch that were 1080p on PS4?

Yes they were and after launch as well,that is all I care the end results the PS4 had gaps bigger than 100% when the power gap was 40%.

So yeah it's normal.

I'm not crying. I'm stating background context for XBO's unstable results.

For the record, I have R9-290 Pro (4.8 TFLOPS, later died) and R9-290X factory OC (1040 Mhz 5.86 TFLOPS, later reflashed to R9-390X at 1050 Mhz 5.9 TFLOPS LOL) GPUs in year 2013. Year 2012 are my 7970 1Ghz factory overclock (similar to 7970 GE) and 7950-BE (reflashed to 900 Mhz base clock BIOS).

I'm here to shootdown Cow's flawed arguments since year 2008 while my W5000 example has placed a limit on XBO's middleware fixes.

Avatar image for fedor
Fedor

6015

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#130  Edited By Fedor
Member since 2015 • 6015 Posts

@BlackShirt20: I've already watched them... Thats why I challenged you to time stamp it, because you can't. They go no further than saying it's in the realm of possibility, it's impossible to "do the math" past pure speculation on what could be completely inaccurate data.

Avatar image for tormentos
tormentos

29703

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#131 tormentos
Member since 2003 • 29703 Posts

@BlackShirt20:

10,20,30 the gap will be smaller than this gen,and that is taken for good information who leakers claim is wrong.

At least for the PS5.

Avatar image for tormentos
tormentos

29703

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#132 tormentos
Member since 2003 • 29703 Posts

@BlackShirt20:

So leaks for the Xbox series X are true,even when MS it self hasn't confirm 12TF.

But PS5 leaks = old..lol

Let me tell you this.

Leakers on Resetera agree both machines are a spit away one from the other.

The PS5 can't be 9.2TF with 36CU because there is no way that SOC will be that hot and high clocked.

Add to this that AMD used fake I'd on their test to throw of leaks.

In fact several of the leakers claimed the PS5 was even a little faster.

And the gibhub leak has no HBRT period when it was confirmed already.

I don't know which will be more powerful,but you sure look sold out on the Xbox series X been stronger based on an unconfirmed leak,not even MS explained the 2X comment they make,and they already claimed 4X the power of the Xbox one X without clarifying on the post that it was CPU based and not GPI based.

Avatar image for tormentos
tormentos

29703

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#133 tormentos
Member since 2003 • 29703 Posts

@ronvalencia:

No you are making excuses for the Xbox one,like you ALWAYS DID Everytime the PS4 beat the Xbox one.

Like on doom again we're the PS4 pulled a huge lead in both frames and resolution.

Plainly put any game on PS4 that had better frames on Xbox was basically screw up on PS4.

Avatar image for ronvalencia
ronvalencia

28555

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#134 ronvalencia
Member since 2008 • 28555 Posts

@fedor said:

@BlackShirt20: They didn't do math, sorry bud. They just reiterated leaks. Leaks that they specifically state in the DF video are complete speculation. But nevertheless, you believe leaks are 100% true so therefore you now believe PS5 is 10.25 tflop.

Known RDNA Shader Engine deployment can scale between 20 CU to 24 CU. Keep it within RDNA Shader Engine framework.

Avatar image for BlackShirt20
BlackShirt20

1357

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#135  Edited By BlackShirt20
Member since 2005 • 1357 Posts

@tormentos: It’s still a noticeable gap. I know you don’t believe that. But it is and will be a noticeable difference.

Let’s for the sake of argument say there is 2 new consoles coming out from companies not named MS or Sony. We will call them Console Alpha and Console Bravo. Let’s also say that the difference in power is 5% (It will be a lot more than that. But let’s pretend they are separated by just 5% to make you happy).

Console Alpha is 5% more powerful than Console Bravo. We are on the eve of release day and the top game for both console is Call of Duty Korean War. Both games run at the same 4K setting. However upon review Console Bravo is prone to a few frame drops in demanding moments. Even if it’s 3-5 frames at a time. While Console Alpha holds steady at 60...... Which console would you concede is better?

Avatar image for ronvalencia
ronvalencia

28555

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#136  Edited By ronvalencia
Member since 2008 • 28555 Posts

@tormentos said:

@ronvalencia:

No you are making excuses for the Xbox one,like you ALWAYS DID Everytime the PS4 beat the Xbox one.

Like on doom again we're the PS4 pulled a huge lead in both frames and resolution.

Plainly put any game on PS4 that had better frames on Xbox was basically screw up on PS4.

1. That's a FALSE narrative, you repeatedly missed my W5000 vs 7850 (and R7-265) arguments.

R7-265 ~= PS4

You can't handle center position between XBO and PS4.

2. You didn't factor in developers skill and non-transparent nature with DDR3 and 32MB ESRAM split.

MS later provided transparent 1080p framebuffer split render with automatic load balancing between DDR3 and 32 MB ESRAM memory pool.

For example, Killer Instinct Season 1 on XBO released as 720p. Killer Instinct Season 2 has 900p.

You repeatedly missed my W5000 vs 7850 (and R7-265) arguments.

3. Cow's 32 ROPS vs 16 ROPS arguments are flawed. XBO's problems are somewhere else.

Avatar image for tormentos
tormentos

29703

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#137 tormentos
Member since 2003 • 29703 Posts

@BlackShirt20Again you are giving for true the rumors that the Xbox one X is 12TF and that the PS5 is 9.2.

Hell the Xbox series X can't be 12TF with 56CU at 1700mhz,the Xbox one X ded kit has 4 CU more than the Xbox one X,which means the Xbox series X will probably have 52CU and will require even higher clock rate which on a big Soc as that will be troublesome.

Avatar image for BlackShirt20
BlackShirt20

1357

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#138 BlackShirt20
Member since 2005 • 1357 Posts

@tormentos: Oh boy. Are you gonna bring up ray tracing now? The last time you dodged one of my comments you brought up ray tracing. PS5 probably doesn’t have ray tracing hardware.

Avatar image for deactivated-5e20304971dd5
deactivated-5e20304971dd5

67

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 5

#139 deactivated-5e20304971dd5
Member since 2020 • 67 Posts

@tormentos: if your argument is: "You are following the rumors that XSX is 12.2 and PS5 is 9.2.", then why even have a discussion... Even if the PS5 has 1TF extra power there's still a noticeable difference. And if you don't want to discuss this as long as they are rumors, well then why are you discussing it...

Avatar image for Pedro
Pedro

37209

Forum Posts

0

Wiki Points

0

Followers

Reviews: 65

User Lists: 0

#140  Edited By Pedro
Member since 2002 • 37209 Posts

The difference between the two systems would be negligible and I am open to being proven wrong when the systems releases later this year. As far as I am concern, anyone hyping the Series X 2 TFLOPS advantage is being silly.

Avatar image for ronvalencia
ronvalencia

28555

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#141  Edited By ronvalencia
Member since 2008 • 28555 Posts

@tormentos said:

@BlackShirt20Again you are giving for true the rumors that the Xbox one X is 12TF and that the PS5 is 9.2.

Hell the Xbox series X can't be 12TF with 56CU at 1700mhz,the Xbox one X ded kit has 4 CU more than the Xbox one X,which means the Xbox series X will probably have 52CU and will require even higher clock rate which on a big Soc as that will be troublesome.

XSX GPU can be 60 CU with the following configuration

Config 1

Three RDNA Shader Engines. Each RDNA Shader Engine has 20 CU. Two CUs are closely glued together as DCU (dual CU).

Config 2

Two RDNA Shader Engines. Each RDNA Shader Engine has 30 CU. Two CUs are closely glued together as DCU (dual CU).

Backgound information on RDNA GPU design framework

Two CUs are binded together as DCU (Dual CU).

Config 1 has the benefit with scaling RBE (Render Balckend) and Geometry Prim Units scaling with CU count.

Config 2 has diminishing returns with additional CUs. Fury X/Vega raster graphics IPC degradation (which is not a problem for server TFLOPS workloads).

Known RDNA deployment

NAVI 10's Shader Engine has 20 CU. NAVI 10 has two Shader Engines

NAVI 14's single Shader Engine has 24 CU.

RX-5600 XT has NAVI 10 with 192 bits bus which is first time AMD was able to follow NVIDIA's Pascal/Turing 192 bit bus design, hence extra BOM cost flexibility for AMD.

https://hothardware.com/news/microsoft-xbox-series-x-holiday-2020-zen-2-amd-rdna

Getting down to brass tacks, Microsoft says that it is using a "custom designed" AMD Zen 2-based processor (although early reports have indicated that it will be based on an 8-core/16-thread Ryzen 3000 design) and "next generation" RDNA graphics architecture

Microsoft claims "next generation" RDNA graphics architecture which points to "RDNA 2".

RDNA 2 includes 7nm+ EUV with 20 percent increase density.

https://www.techpowerup.com/254656/amd-zen3-to-leverage-7nm-euv-for-20-transistor-density-increase?cp=2

7 nm+ EUV (extreme ultraviolet) silicon fabrication node at TSMC, which promises a significant 20 percent increase in transistor densities compared to the 7 nm DUV (deep ultraviolet) node

325 mm2 with "7nm+ EUV" roughly equalvent to 390 mm2 size GPU under 1st gen 7nm DUV node

Microsoft confirm their Scarlet APU has "RDNA 2".

Avatar image for ten_pints
Ten_Pints

3954

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 5

#142 Ten_Pints
Member since 2014 • 3954 Posts

Is this a dick measuring contest?

Avatar image for xantufrog
xantufrog

12229

Forum Posts

0

Wiki Points

0

Followers

Reviews: 2

User Lists: 5

#143  Edited By xantufrog  Moderator
Member since 2013 • 12229 Posts

@ten_pints: I think they THINK it is

Avatar image for 04dcarraher
04dcarraher

23319

Forum Posts

0

Wiki Points

0

Followers

Reviews: 2

User Lists: 0

#144 04dcarraher
Member since 2004 • 23319 Posts

@Pedro said:

The difference between the two systems would be negligible and I am open to being proven wrong when the systems releases later this year. As far as I am concern, anyone hyping the Series X 2 TFLOPS advantage is being silly.

The differences will be negligible when it comes to the core architecture that both will be using aka "RDNA". Even when we look at AMD's RX 5700xt vs 5700, that's a 1.8 TFLOP difference (nearly 2 TFLOP) and the difference is like 20% at best in framerate's at 1440p and 4k same settings.

All they have to do is switch a few settings from ultra to high, or high to medium etc etc.... or use things like more dynamic resolution solutions than the other. Its something that nobody is going to outright see the differences without having to use side by side comparisons.

Its going to be other aspects of the consoles that determine if the other has a bigger gap in performance and or graphics. memory amounts and bandwidth, cpu clock rates, amount of cores/threads used for features, and which RT implementation each company is going to use..... custom 3rd party solution? an AMD hybrid solution?, or a dedicated hardware solution from AMD?.

Avatar image for Pedro
Pedro

37209

Forum Posts

0

Wiki Points

0

Followers

Reviews: 65

User Lists: 0

#145 Pedro
Member since 2002 • 37209 Posts

@ten_pints said:

Is this a dick measuring contest?

Yep! The odd part is they measuring dicks they haven't seen. LOL

Avatar image for ronvalencia
ronvalencia

28555

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#146  Edited By ronvalencia
Member since 2008 • 28555 Posts

@04dcarraher said:
@Pedro said:

The difference between the two systems would be negligible and I am open to being proven wrong when the systems releases later this year. As far as I am concern, anyone hyping the Series X 2 TFLOPS advantage is being silly.

The differences will be negligible when it comes to the core architecture that both will be using aka "RDNA". Even when we look at AMD's RX 5700xt vs 5700, that's a 1.8 TFLOP difference (nearly 2 TFLOP) and the difference is like 20% at best in framerate's at 1440p and 4k same settings.

All they have to do is switch a few settings from ultra to high, or high to medium etc etc.... or use things like more dynamic resolution solutions than the other. Its something that nobody is going to outright see the differences without having to use side by side comparisons.

Its going to be other aspects of the consoles that determine if the other has a bigger gap in performance and or graphics. memory amounts and bandwidth, cpu clock rates, amount of cores/threads used for features, and which RT implementation each company is going to use..... custom 3rd party solution? an AMD hybrid solution?, or a dedicated hardware solution from AMD?.

It depends on GPU design difference i.e. memory bandwidth, CU TFLOPS, classic GPU raster hardware and L2 cache/render cache.

Xbox One GPU family would need semi-custom changes to be carried over to XSX e.g. FP10 support from Xbox 360 (not on PC GCN or PC NAVI), perhaps 8MB L2 cache due to X1X GPU's 2MB L2 cache + 2 MB render cache semi-custom design.

5700 has the same classic GPU hardware as 5700 XT i.e. same geometry input, same RBE (render backends), same L2 cache storage, same L1 cache storage.

5700 has the same memory bandwidth as 5700 XT i.e. GDDR6-14000

5700 has ~7.49 TFLOPS at 1650 Mhz game boost

5700 XT has 8.9 TFLOPS at 1755 Mhz game boost

17.8 percent difference in raw TFLOPS.

8 percent difference in classic GPU hardware.

Average difference between the combined CU's ALU and classic GPU hardware is 12.9 percent

https://www.techpowerup.com/review/amd-radeon-rx-5700/28.html

Effective difference is 1440p's 14% to 4K's 13% difference

E3 2019's 320 bit GDDR6-14000 has 560 GB/s which may not show the full memory bus config.

560 GB/s vs NAVI 10's 448 GB/s difference is 25 percent.

PS4 Pro's 224 GB/s vs X1X difference is 46 percent.

https://www.techpowerup.com/gpu-specs/xbox-series-x-gpu.c3482 has 672 GB/s from 384 bit bus.

672 GB/s vs 448 GB/s difference is 50 percent higher.

XSX's ~325 mm2 size GPU is no match against the rumored PC's NAVI 21 with twice RX 5700 XT power and 505 mm2

Microsoft already confirms "next gneration RDNA" for XSX.

Avatar image for Pedro
Pedro

37209

Forum Posts

0

Wiki Points

0

Followers

Reviews: 65

User Lists: 0

#147 Pedro
Member since 2002 • 37209 Posts

@ronvalencia: You spew all of that garble and nothing changed. The difference between the two consoles will still be negligible.

Avatar image for BlackShirt20
BlackShirt20

1357

Forum Posts

0

Wiki Points

0

Followers

Reviews: 1

User Lists: 0

#148 BlackShirt20
Member since 2005 • 1357 Posts

@Pedro: LOL ok. Hold on to that.

Avatar image for tormentos
tormentos

29703

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#149 tormentos
Member since 2003 • 29703 Posts

@ronvalencia said:

XSX GPU can be 60 CU with the following configuration

Config 1

Three RDNA Shader Engines. Each RDNA Shader Engine has 20 CU. Two CUs are closely glued together as DCU (dual CU).

Config 2

Two RDNA Shader Engines. Each RDNA Shader Engine has 30 CU. Two CUs are closely glued together as DCU (dual CU).

Backgound information on RDNA GPU design framework

Two CUs are binded together as DCU (Dual CU).

Config 1 has the benefit with scaling RBE (Render Balckend) and Geometry Prim Units scaling with CU count.

Config 2 has diminishing returns with additional CUs. Fury X/Vega raster graphics IPC degradation (which is not a problem for server TFLOPS workloads).

Known RDNA deployment

NAVI 10's Shader Engine has 20 CU. NAVI 10 has two Shader Engines

NAVI 14's single Shader Engine has 24 CU.

RX-5600 XT has NAVI 10 with 192 bits bus which is first time AMD was able to follow NVIDIA's Pascal/Turing 192 bit bus design, hence extra BOM cost flexibility for AMD.

https://hothardware.com/news/microsoft-xbox-series-x-holiday-2020-zen-2-amd-rdna

Getting down to brass tacks, Microsoft says that it is using a "custom designed" AMD Zen 2-based processor (although early reports have indicated that it will be based on an 8-core/16-thread Ryzen 3000 design) and "next generation" RDNA graphics architecture

Microsoft claims "next generation" RDNA graphics architecture which points to "RDNA 2".

RDNA 2 includes 7nm+ EUV with 20 percent increase density.

https://www.techpowerup.com/254656/amd-zen3-to-leverage-7nm-euv-for-20-transistor-density-increase?cp=2

7 nm+ EUV (extreme ultraviolet) silicon fabrication node at TSMC, which promises a significant 20 percent increase in transistor densities compared to the 7 nm DUV (deep ultraviolet) node

325 mm2 with "7nm+ EUV" roughly equalvent to 390 mm2 size GPU under 1st gen 7nm DUV node

Microsoft confirm their Scarlet APU has "RDNA 2".

And hereeee we goooo with your usual speculation did MS CLAIMED RDNA 2?

OPENLY?

From what i read on resetera from leaks is RDNA 1 with RDNA 2 features the same is say about the PS5.

And is not new we know the Pro was polaris with vega features is not new.

This leaks are about the development kit not about retail units,which mean if the dev kit has 56CU the final hardware will have less,just like the XBO X has 4 less than the devkit.

So it can't be 60CU.

@BlackShirt20 said:

@tormentos: Oh boy. Are you gonna bring up ray tracing now? The last time you dodged one of my comments you brought up ray tracing. PS5 probably doesn’t have ray tracing hardware.

You are the one who dudged my argumen dude..But then again you have not much to fight with.

Avatar image for EG101
EG101

2025

Forum Posts

0

Wiki Points

0

Followers

Reviews: 0

User Lists: 0

#150 EG101
Member since 2007 • 2025 Posts

@xantufrog said:

@EG101: that's a strange way of looking at it. I have two vintage vespas, 6 vs 12 horsepower - one double, with a sizeable difference in the rider's experience. Nowadays I might also compare cars of 180 vs 220 - this is a larger gap in raw numbers, yes, but it's not double the horsepower and the differencd in experience is nowhere near comparable.

This is the same. As technology evolves, the numbers will keep growing, pushing the scale which makes it more likely one could make your argument and yet it might be meaningless. The % matters.

130 TFLOPS vs 120... a whopping 10 TFLOPS gap! That nobody will care about because the ratio is minute

I don't think your analogy works here.

Game graphics scale linearly with how much HP that engine (CPU & GPU) has while horse power on a car have a lot of other variables that affect performance.

For example things like, torque, gear ratios, traction, suspension, etc, etc all affect the performance of your vehicle.