39:20 You should really paint the inside tabs (both your diy cable and the PCB adapter) white or yellow like MSI Mag PSUs. At a glance you can see if its seated or not. This is super important as its the main failure point of 4090. I do this for mine so I don't 2nd guess myself. I also put 2 tiny dots of black hot glue at the ends to be extra safe, emphasize on tiny, you don't need a lot.
@@jackipiegg Found it. That's such a simple and clever modification that could help prevent unnecessary failures. Seeing it first hand gives me some ideas to explore. Appreciate you for bringing it to my attention.
Greetings and good day. Unfortunately I cannot (and would not) tell you specifically what GPU upgrade to make, since there isn't enough info available to do so. However, if I were in your shoes, here's the scenarios I'd investigate to try and definitively answer the question. What dimension am I targeting with the upgrade? General computing? Gaming? Python workloads on GPU? If general computing, I wouldn't target a GPU upgrade, I'd target a CPU upgrade. If GPU processing for python workloads, I'd validate that the data to be processed would be accelerated by more (count of and power of) CUDA cores. If gaming, the variables are more complex to understand. I'd perform some tests in the games you actually play to see if you are bottlenecked by your current setup, and what component you are bottlenecking. While I believe bottlenecking is made out to be a larger problem than it actually is, its also foolish to deny its existence. Other important questions to consider are... - What resolution are you targeting? There is a tipping point where, generally speaking, the GPU matters more than the CPU and vice versa. This is not explicitly true for every situation though...just a guideline. - What framerates are you targeting? - What is the hz rating of the monitor you are using? In a perfect world you'd want your pc to spit out as many frames as your monitor can display. - Does your monitor have some sort of frame syncing technology (gsync/freesync)? And of course your power supply isn't monumentally capable, so regardless of what you decided to upgrade, you'd possibly need to consider that upgrade as well to cover the overhead of more powerful components. This is not an exhaustive list of considerations, but with the information you supplied, these would at least get you started on answering the question. Hope that helps. Take care.
39:20
You should really paint the inside tabs (both your diy cable and the PCB adapter) white or yellow like MSI Mag PSUs. At a glance you can see if its seated or not. This is super important as its the main failure point of 4090.
I do this for mine so I don't 2nd guess myself. I also put 2 tiny dots of black hot glue at the ends to be extra safe, emphasize on tiny, you don't need a lot.
Hadn't thought of that but its a very good suggestion. Thanks for the idea.
@@htfiBEEP
My first inspiration was from PCWorld's MSI Mag video. Was surprised how insecure that connector is and that video shows it really well.
@@jackipiegg I'd be interested in checking that PCWorld video out. Do you happen to have a link saved?
@@htfiBEEP
Just search "pcworld msi mag"
sorry couldnt give links, youtube will delete the comment
@@jackipiegg Found it. That's such a simple and clever modification that could help prevent unnecessary failures. Seeing it first hand gives me some ideas to explore. Appreciate you for bringing it to my attention.
Crimpin' ain't easy.
Neither is decrypting Yautja, but you'll have my admiration if you are able to do so.
What is best gpu to upgrade from gtx 1660s i have i7 9700 16 gb ddr4 ram and 600w psu
Greetings and good day. Unfortunately I cannot (and would not) tell you specifically what GPU upgrade to make, since there isn't enough info available to do so. However, if I were in your shoes, here's the scenarios I'd investigate to try and definitively answer the question.
What dimension am I targeting with the upgrade? General computing? Gaming? Python workloads on GPU?
If general computing, I wouldn't target a GPU upgrade, I'd target a CPU upgrade.
If GPU processing for python workloads, I'd validate that the data to be processed would be accelerated by more (count of and power of) CUDA cores.
If gaming, the variables are more complex to understand.
I'd perform some tests in the games you actually play to see if you are bottlenecked by your current setup, and what component you are bottlenecking. While I believe bottlenecking is made out to be a larger problem than it actually is, its also foolish to deny its existence.
Other important questions to consider are...
- What resolution are you targeting? There is a tipping point where, generally speaking, the GPU matters more than the CPU and vice versa. This is not explicitly true for every situation though...just a guideline.
- What framerates are you targeting?
- What is the hz rating of the monitor you are using? In a perfect world you'd want your pc to spit out as many frames as your monitor can display.
- Does your monitor have some sort of frame syncing technology (gsync/freesync)?
And of course your power supply isn't monumentally capable, so regardless of what you decided to upgrade, you'd possibly need to consider that upgrade as well to cover the overhead of more powerful components.
This is not an exhaustive list of considerations, but with the information you supplied, these would at least get you started on answering the question.
Hope that helps. Take care.