Not a bad chip to be ‘stuck’ with TBH.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
AMD Zen 5 rumours
- Thread starter Grim5
- Start date
Not a bad chip to be ‘stuck’ with TBH.
Associate
- Joined
- 13 Sep 2010
- Posts
- 2,314
Yeah I think I'm going to be trying to resist the urge and stick with the 5950x ...only about 10-12% from a 7kx3D at 1440p (according to TPU).
There is still a possibility of a late night YOLO though... that or splurging too much on a better GPU, I don't really need much more performance but I want it lol
There is still a possibility of a late night YOLO though... that or splurging too much on a better GPU, I don't really need much more performance but I want it lol
Caporegime
Exactly what I did a few weeks back. Just got the itch and went 5950X to 7950X3D. I can't tell much of a difference.
I’ve got the same motherboard and CPU as yourself @RSR - It would be great to see what BIOS settings you and @LtMatt have ended up with.
I’m fairly new to tuning these modern AMD CPU’s and have had no issue gaining an extra 4000+ on Cinebench over stock, with a very basic curve offset, but I would love to see what areas I might have overlooked in my efforts so far.
Last edited:
Soldato
Yeah here is to hoping that the 10950x3D (or whatever the name is going to be) gives another 10-12% and then jump and then have that compound 25% increase then. With course whatever the GPU will be like a 6090RTX or whatever.
Last edited:
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 31,028
Soldato
12% of whatever your base call 100, then 12% of that new base would be around 25% total increase of the original base. Not sure why funny? Compound increase is normal?
100 (baseline) * 1.12 for 12% = 112%
112 * 1.12% increase from the previous gen = 125.44% of the initial baseline
Edit:
Or
60fps has 12% increase = 67.2 fps
67.2fps has another 12% increase = 75.264
So 60fps with a 125.44% increase = 75.264
Hence the approx. 25% increase in performance if they managed two lots of 12% uplift.
Check Math isn't even that hard. difference in FPS between 60 and 75 (rounded cause lazy) = 15. So 60/15 = 4. 100%/4 =25%.
Edit 2: As a reminder below I completely forgot the tired & drunk factor in my calc and as a raging alcoholic I multiply all numbers I see on screen by a thousand so my increase in peformance is actually 25,000%. Thank you to @MikeB for helping me correct my calculations.
Last edited:
You are not building in the multipliers for when we are tired, drunk etc when we use the screens/pcs, obviously of course we all need 1 gagiliion per Nanosecond, by the time we reach the nirvana of 240mhz always on 4k with full atmos surround etc most on this forum will be using varifocals and walkers
Soldato
Ah fudge, I knew my maths be out for something. Cheers for the correction. I will amended accordingly.You are not building in the multipliers for when we are tired, drunk etc when we use the screens/pcs, obviously of course we all need 1 gagiliion per Nanosecond, by the time we reach the nirvana of 240mhz always on 4k with full atmos surround etc most on this forum will be using varifocals and walkers
AMD Ryzen 7 9800X3D official performance leak: 8% better at gaming, 15% in multi-threaded apps vs. 7800X3D - VideoCardz.com
Leaked AMD Ryzen 7 9800X3D marketing slide claims 8% gaming performance boost over 7800X3D This is how AMD is going to advertise its new CPU. AMD is set to launch its first Ryzen 9000 processor equipped with 3D V-Cache on November 7th. We are still two weeks away from this launch, but given the...
videocardz.com
As I like to repeat, welcome to the times where silicon tech is over and all the low hanging speed up fruits are done. That's said all the new "nm" markings are pure marketing since 16-22nm size of transistor is relatively static for years now and all they really change is just how exactly these transistors are built, which is what give us power savings etc.
It's similar story with nand chips for SSD - they can't have them any smaller (they are already perfect size for years because of quantum physics) but they change how many levels of voltage they can store and detect in each cell and how many layers they can stack on top of each other, whilst retaining sensible durability. That's how we get higher capacity, not because of making each cell smaller. All of that has limits though. Again, similar things about RAM chips and sizes of cells (capacitors) in them, that can't change much anymore either, which affects cache in CPUs too.
All in all, interesting times to see how much they can optimise these tiny elements without shrinking them anymore before we really hit the wall. I now much better understand what Nvidia CEO means by silicon tech is over - real size doesn't change for years now and they are running out of ideas how to optimise it further, whilst marketing just puts lower bs numbers each year that don't mean anything anymore.
I've been using a PC since 1989, I'm very well aware about decreasing returns, take into account that my previous CPU was an i7-3770k from 2013 and the PC before lasted 6 years too.
Frankly I think we will see a gradual switch to more specialized hardware in order to eke out more performance (not to mention marketing $$$) on one side and hopefully a stronger push towards software optimization on the other.
I'm more worried by the software industry stagnating (especially in games development) than hardware in itself, graphics have been good enough in my eyes for two decades but games capable of bringing real innovation are scarcer and scarcer, nobody is going to take real risks anymore aside from indie developers and even there we can't expect much from their limited budgets.
Quite possibly, when hardware can't do it better, horribly inefficient (in most cases) software will finally have to give and start catching up. Currently a lot of companies creating software just does it cheaply and fast, with absolutely 0 care for the performance as they can just increase system requirements. CPUs speeds rose up by hundreds of thousands times over time, software just becomes slower - it's clear where the problem lies.
I've read recently devs talking on Steam about how they optimised their game, what they had to do, how much time each stage took and how much speed up they gained. As those devs said, they started with around 100FPS on average hardware, they could've just put DLSS and FG into system req. and be done with it. But they didn't stop and bit by bit, they optimised their game so when they were done they reached over 300FPS in same settings and resolution, without DLSS and FG required to do it and they were also able to increase quality of final image and not cut it down, doing all these steps! It's clear the issue with modern games is almost always with time=monies when publishers just push hard to release product and worry later.This is also same issue - just monies. They blew out their AAA projects so horribly much that they forgot games are supposed to be fun and innovative, not just rehashing same thing in new skin over and over again for safe monies. Gladly, currently AAA games are flopping one after another for variety of reasons that could be boiled down to - they're made for nobody, people expect different thing and not the same thing in new skin. Innovate or die, is the usual thing, which publishers forgot. Gladly, in this case, market is already correcting itself, even though moving slowly.
Last edited:
How so?, MSRP of 7800X3D is £449
Soldato
Where you getting your prices from ?
It ain't lol
Last edited:
Caporegime
50%???? the 7800X3D is £440, + 50% = £660.
Associate
I guess a few weeks ago, the 50% may have been correct. £330 + 50% = ~£500