[SOLVED] Is GT 730 ddr5 temporarily good for ryzen 7 build while waiting for GPU prices to drop? or should I just go with APU?

[Newwbbbbss]

Is GT 730 ddr5 temporarily good for ryzen 7 build while waiting for GPU prices to drop?
So I'm planning to build a pc with ryzen 7 5000 series and I have gt 730 2gb ddr5 gpu. I would like to wait for the prices to drop for gpu's but I'm not sure if it would be detrimental for my build to use my stock gpu. Should I just go with APU's like ryzen 7 5700G?
I am a light gamer and I usually use it for video and photo editing.
What do you recommend should I go first with Ryzen 7 3700X, 5800X, install my gt 730 ddr5 temporarily and wait for GPU prices to drop or should I just change my processor to APU's?

BTW, I'm a newbie. Thanks a lot to those who would like to help me with this.

 

[Darkbreeze]

GT 730 is probably a worse performer than most of the integrated graphics from both the latest gens of Intel and AMD. And even if it were close, the APU would be the better option because there's not much point in paying extra for a graphics card that gives you the same performance as the integrated graphics unless you simply needed cheap additional video outputs.

https://www.tomshardware.com/reviews/amd-ryzen-7-5700g-review

 

[InvalidError]

Since you already have the GPU, there isn't much harm in using it for the time being. As already pointed out, it'll suck relative to most modern IGPs but should be bearable as a temporary fix.

If you are considering the 5700G, you may want to look at the i5-12500 instead. The IGP may be weaker but the CPU is much stronger for when you eventually get whatever GPU you want. You also get 4.0/5.0x16 (depending on board) instead of 3.0x8 with the Ryzen 5700G.

 

[Darkbreeze]

Yes, the i5-12500 is definitely a stronger performer. If you plan to get a higher end gaming card later it isn't the worst advice if you can live with the difference in performance between the two integrated graphics processors in the meantime.

I suspect there isn't going to be a really discernible difference in performance between the PCIe 3.0 and 4.0 slots just like there really wasn't between the 2.0 and 3.0 slots even using cards that supported the faster standard, but the CPU performance is definitely worth considering since the prices of those CPUs are like a dollar difference.

 

[InvalidError]

I suspect there isn't going to be a really discernible difference in performance between the PCIe 3.0 and 4.0 slots just like there really wasn't between the 2.0 and 3.0 slots even using cards that supported the faster standard

I refer you to the RX6500 debacle. Not viable on PCIe 3.0 even at lowest details. The 4GB RX5500 suffers pretty badly from 3.0x8 in many scenarios too.

If AMD and Nvidia are going to screw around with truncated PCIe bus, 4.0 is going to become absolutely essential.

 

[Darkbreeze]

I had not heard about that yet, will have to do some research so I don't put my foot in my mouth. I was making assumptions based on the last time around.

https://www.pugetsystems.com/labs/a...E-Speed-on-Gaming-Performance-518/#Conclusion

 

[InvalidError]

I had not heard about that yet, will have to do some research so I don't put my foot in my mouth. I was making assumptions based on the last time around.

https://www.pugetsystems.com/labs/a...E-Speed-on-Gaming-Performance-518/#Conclusion

It is the 4GB and lower cards that are getting hit disproportionately hard by crippled PCIe bandwidth since those are the ones that need to lean on system memory the most in new-ish titles, not the crazy-high-end stuff with heaps of VRAM to spare while running their respective contemporary titles. And in the currently messed up GPU market, 4GB GPUs are pretty much the most you can get new under $400 street price.

I predicted that PCIe 4.0 would become vital to the continued viability of 4GB GPUs 3-4 years ago. We're there. Only problem is nobody has released a 4GB 4.0x16 GPU to demonstrate how right/wrong I am. Closest thing is the 4GB RX5500 which does show decent gains on 4.0x8 vs 4.0x4, closing most of the gaps against its 8GB counterpart.

4.0x16 should bring 4GB from marginal to perfectly viable at a much lower cost than 2-4GB more GDDR6.

 

[tennis2]

...instead of 3.0x8 with the Ryzen 5700G.

The 5700G has 24 lanes of PCIe3.0. x16 for the GPU, leaving x8 for storage/IO. Just wanted to clarify that since your statement could be misleading to other readers.

 

[InvalidError]

The 5700G has 24 lanes of PCIe3.0. x16 for the GPU, leaving x8 for storage/IO. Just wanted to clarify that since your statement could be misleading to other readers.

Previous APUs only had x8 and AMD doesn't bother specifying connectivity beyond "3.0" on its product page. People buying one are still screwed with AMD limiting it to 3.0 while crippling their entry-to-mid-range GPUs with x4/x8 interfaces.

 

[tennis2]

It seems even Tom's is confused:
https://www.tomshardware.com/reviews/amd-ryzen-7-5700g-review

Re-using the Renoir SoC design also means that AMD has stuck with the PCIe 3.0 interface found on all its current-gen APUs. As such, the chip has 20 lanes of PCIe 3.0 connectivity (16 for graphics, four for the chipset, and four for storage) compared to 24 lanes of PCIe 4.0 found on the Ryzen 5000 models for the desktop PC.

And their mostly copy-pasted 5600G review:
https://www.tomshardware.com/reviews/amd-ryzen-5-5600g-review

As such, the 5600G has 24 lanes of PCIe 3.0 connectivity compared to 24 lanes of PCIe 4.0 found on the Ryzen 5000 models for the desktop PC.

I trust Steve at Gamers Nexus. He said 24 lanes of PCIe3.0.

 

[InvalidError]

It seems even Tom's is confused:

To be fair, the x4 that goes to the chipset really shouldn't be counted since they aren't user-accessible, so it really is 16+4 usable. Otherwise, you should count Intel's CPUs as 28 lanes since they have x8 for DMI.