End of Year PC Build

First off if CUDA and Raytracing do not matter I would go with the RX 6700XT.
Its cheaper than the 3060/3060TI on Amazon right now. 

However if you need those two things or FSR and Re-Bar shun you then here is the current prices of the 3060's:
The 3060 will run you around $350 to $450 -unless you go MSI which will be about $500+
The 3060 TI will run you around $450 to $500  -again if you go with MSI you will be paying $500+
The Radeon RX 6700 XT will run you $370 to $470 and will beat the 3060/TI's WITHOUT using the C.R mentioned earlier. 
This shows the differences: 

Still even if not getting one anytime soon this will give you the info you need. 
There also is of course one cavate I will give according to Buildzoid the MSI RTX 3060 Ti Gaming X has a bad PCB design.
A no buy on that one if you do go 3060 TI.

Very true it doesn't have to be the GPU that is bottlenecked.
It can be any component as the previous video states that is not giving its maximum performance. 

Single core CPU & RAM bottleneck for example: 
pbs-prod.linustechtips.com/monthly_2021_02/315919272_Bottleneckorno.jpg.6f8babb165da7e4bdd06837a579abf86.jpg

Exactly. Your average consumer can't replace the rattling fans on a video card or fix coil whine either.

Hell, even I can't fix the coil whine on my EVGA 3090 FTW3.

Because it is still under warranty and I don't want to void it. I already had the original card replaced under warranty. If I had done any modification like putting superglue on the chokes, EVGA would probably not honor the warranty.

Either way, I'll most likely keep it original since it's one of the last video cards ever made by EVGA. When it's time for it to retire, I'll clean it up and put it on display instead.

It's not so much about clock speed anymore. IPC is what it's all about. And for gaming it seems 3D V-cache will start catching on.

I'll stick with Intel, thanks.

'Preciate your thorough input though!


Have a great day, Vanguardian.

To be honest, it hasn't really been clock speed ever since Intel dumped the Netburst architecture for the Core architecture. Having more IPC is also one of the reasons why AMD's Athlon, Athlon XP, and Athlon 64 were able to stay competitive despite being clocked much lower than Intel's Pentium 4 CPUs.

I'm one of those people who care about performance per watt and I'm very excited about the upcoming Ryzen 7000X3D CPUs. These new X3D-series CPUs should have lower base clocks and boost clocks than their normal X-series counterparts. So that means lower power consumption, better thermals, and much better efficiency while delivering up to 30% better gaming performance due to the gigantic L3 3D V-Cache. However, lower clocks also means they won't be as good for productivity workload.

You must know the quandary that I sometimes am right at home in being both right and wrong at the same time. If you want a 30% performance boost you most likely will have to go above the 7800X3D. The base clock is lower but they still supposedly will have the same boost clock. The TDP will be lower for the top two but higher for the final one and the L3 Cache is DOUBLELING at the least for all three X3D chips -that is right three chips. Also the new NON-X 7000's launched two days ago on the Tenth and they have the 65W I was hoping we would see.

If I was building a new PC, I'd get at least a 7950X3D to replace my 5950X.

However, I probably won't need a new PC for at least two more years. This current build is already perfect for my needs. It can get more than 100 FPS in Cyberpunk 2077 while being cool and quiet. The Gigabyte RTX 4090's heatsink is so overbuilt that the GPU and the VRAM never hit 70C. Most of the time it just chugs along at 50-60C.

Only if that 30% overclock actually translates to 30% performance increase. Most of time, it doesn't.

I don't have any experience with overclocking the 7900 XTX, but the current widely-accepted overclocking method seems to be a bit more complicated and unorthodox. First, the 7900 XTX already run very hot compared to Nvidia's RTX 40-series GPUs. Second, the AMD reference cards aren't very good for overclocking due to their power limit. Third, there are also MANY reports of defective reference 7900 XTX coolers causing overheating. We are talking about a problem that is way more widespread than Nvidia's melting 12VHPWR connectors and it might require a recall. So you're much better off with a more expensive board partner card with a 3x8-pin custom board design, a beefy custom cooler, and an increased power limit.

Unlike traditional overclocking, you don't touch the clocks at first. Instead, you increase the power limit to maximum and you slowly undervolt the GPU until it is no longer stable. Once you've reached that point, you increase the voltage a bit to keep it 100% stable. Now that you've found the lowest stable voltage, you can begin to increase the maximum GPU frequency. After you've reached a stable GPU overclock, you can then start raising the VRAM frequency. Keep in mind that higher clocks don't necessarily mean better performance. So just keep testing, benchmarking, and adjusting the clocks until you get the best benchmark numbers without crashing.

The thing is if you're spending more money on a board partner 7900 XTX, you might as well spend a bit more on something like an RTX 4080 which offers similar rasterization performance, slightly better power efficiency (10-30 watts lower at the same framerate), much better thermals, significantly better ray-tracing performance, better upscaling with the DLSS, and Tensor cores for your CUDA apps.

Remember that the RTX 4080 is already a terrible value even compared to the RTX 4090. It's mind-blowing that the $1600 4090 actually offers better performance per dollar than the $1200 4080. In conclusion, they are all terrible value for the asking price. Both the 4080 and the 7900 XTX need to be priced at least $300 lower. $900 MSRP for the 4080 and $700 MSRP for the 7900 XTX.

Why is your 4090 a bad purchase?