By using this site, you agree to our Privacy Policy and our Terms of Use. Close

Forums - PC - Carzy Zarx’s PC Gaming Emporium - Catch Up on All the Latest PC Gaming Related News

Cyran said:
Trumpstyle said:
Hmm they didn't reveal the ipc gain? Now we don't know the performance for PS5, where it will land.

If compute Units/streaming processes any indication

PS5 = 36cu

Xbox series x = 52cu

6800 = 60cu

6800xt - 72cu

6900 = 80cu

Just a little back of the stamp calculation:

XBO: (100-10%)/60*52=78% GPU performance of the 6800

10% because of the ~10% lower clock speed compared to the game clock of the 6800

PS5: (100+5%)/60*36=63% GPU performance compared to the 6800

5% due to slightly higher clock speeds. I just made 5% because I'm not sure what the gameclock actually will be on the PS5, but could be sightly more or less.

In either case, the raw performance levels (if the GPUs were taken over 1 to 1) wold be more akin of some hypothetical 6700 and 6600 GPUs, so mid-range cards like with the PS4 (XBO was almost entry-level) had during it's debut, which was similar to the 7850



Around the Network

Anyway comparing specs with consoles is a bit silly. These are $500 consoles with mid range specs, not $1300+ power users gaming PC's.



vivster said:
Captain_Yuri said:
Size comparison

Loud or terrible to overclock? Take your pick.

Seriously, smaller size is a weird thing to flex when it's about your cooler.

Well, not everybody has a case big enough to fit a 3080 or a 3090. Also, with 3 big fans, I doubt it's going to be very noisy. And if Noise is your concern, just wait for a model from Powercolor or Sapphire, and you're good.



Barkley said:
Bofferbrauer2 said:
The RX 6800 (non XT) is a bit too expensive compared to the 3070. But I guess the availability is king right now, and it beats the 3070 by a fair bit.

16gb vs 8gb memory though, makes it quite a close call imo.

It makes it certainly more future-proof, the 8GB will be a limiting factor in a series of games within 1-3 years.



hinch said:
I'm guessing next year for RDNA 3 cards on TSMC/Samsung's 5nm. Speculation that they're going chiplet design. Verses a possible 7nm Ampere refresh be an interesting fight for sure.

I think RDNA 3 is for 2022, not next year.

But we'll have RDNA 2 vs Ampere both at 7nm in 2021.

Captain_Yuri said:


Round 2 bois. Please be functional!

Congratulations! I hope it gives you many hours of fun.

Bofferbrauer2 said:
Cyran said:

If compute Units/streaming processes any indication

PS5 = 36cu

Xbox series x = 52cu

6800 = 60cu

6800xt - 72cu

6900 = 80cu

Just a little back of the stamp calculation:

XBO: (100-10%)/60*52=78% GPU performance of the 6800

10% because of the ~10% lower clock speed compared to the game clock of the 6800

PS5: (100+5%)/60*36=63% GPU performance compared to the 6800

5% due to slightly higher clock speeds. I just made 5% because I'm not sure what the gameclock actually will be on the PS5, but could be sightly more or less.

In either case, the raw performance levels (if the GPUs were taken over 1 to 1) wold be more akin of some hypothetical 6700 and 6600 GPUs, so mid-range cards like with the PS4 (XBO was almost entry-level) had during it's debut, which was similar to the 7850

MSoft has been fast saying that their console(s) is the only one with full RDNA 2 GPUs, so PS5 could be slightly less powerful than that.



Please excuse my bad English.

Former gaming PC: i5-4670k@stock (for now), 16Gb RAM 1600 MHz and a GTX 1070

Current gaming PC: R5-7600, 32GB RAM 6000MT/s (CL30) and a RX 9060XT 16GB

Steam / Live / NNID : jonxiquet    Add me if you want, but I'm a single player gamer.

Around the Network

I'm pleasantly surprised, although considering recent rumours, it's not entirely surprising either. The RTX 3060 might not be such an obvious choice after all.



JEMC said:
Bofferbrauer2 said:

Just a little back of the stamp calculation:

XBO: (100-10%)/60*52=78% GPU performance of the 6800

10% because of the ~10% lower clock speed compared to the game clock of the 6800

PS5: (100+5%)/60*36=63% GPU performance compared to the 6800

5% due to slightly higher clock speeds. I just made 5% because I'm not sure what the gameclock actually will be on the PS5, but could be sightly more or less.

In either case, the raw performance levels (if the GPUs were taken over 1 to 1) wold be more akin of some hypothetical 6700 and 6600 GPUs, so mid-range cards like with the PS4 (XBO was almost entry-level) had during it's debut, which was similar to the 7850

MSoft has been fast saying that their console(s) is the only one with full RDNA 2 GPUs, so PS5 could be slightly less powerful than that.

We'll know more after launch with how multiplats run on them. But if true, then the PS5 might fall back even more than this calculation already made it look. Certainly not the best situation for Sony right now, that's for sure.



Seems I was on the money with the sizing of the infinity cache.




www.youtube.com/@Pemalite

My guess would be the day before launch?

Also, watch this.

https://www.youtube.com/watch?v=0xUioPsLRKA

Looks like AMD really tricked a bit with their slides. The performance graphs are only valid when they are paired with a Ryzen 5000. It's great that they can leverage additional power and every gamer should probably get a Ryzen 5000 as soon as possible, but it also means the slides don't really tell the full picture.

The 6900XT slide is especially damning since it barely touches the 3090 even with smart access and RAGE overclock.

Much better Linus video. He is smart guy when he wants to be. 



Cyran said:
Trumpstyle said:
Hmm they didn't reveal the ipc gain? Now we don't know the performance for PS5, where it will land.

If compute Units/streaming processes any indication

PS5 = 36cu

Xbox series x = 52cu

6800 = 60cu

6800xt - 72cu

6900 = 80cu

Is Ps5 really that weak? That's like 2 wii u's duck tape together. No?