In the context of UE, you’ll most likely always be GPU bound.
It looks like both have the same amount of cores (12) but I’m not sure how to account for the ‘Performance / Efficiency’ cores in the Intel though. The only multi-threaded (as in operations that get faster with more cores/threads) you’d be doing in UE is compiling shaders and (maybe) compiling the engine.
the AMD 5900 only supports DDR4 RAM, not sure if there is a big reason to want DDR5 in the context of UE though.
AMD is releasing Zen 4 / new AM5 socket next month, It might be good to take that into consideration.
It won’t matter how good you’re PC is, the code hardly works anyway, no point spending big money like i did just so it breaks with versions that are not updated or fixed, unless you like fixing everyone else’s projects then i don’t see why upgrading for this editor makes sense.
I always use Intel processors less conflicts overall, but when you have windows cloud sync to deal with and all these cloud functions they will destroy you’re good setups and keep bad setups so stuff fails regardless of how great you’re PC is, they just break all the programming on it so it runs bad no matter what.
Hi Thx for your reply. Indeed the GPU is the main concern, but I discover that CPU is also a concern afterall. Specially for the build shader part, what ca take a long long time with a old CPU.
From a i9 9xxx to AMD 1700 (from 2017!) , it goes from 5mn to 30mn!!
And yes, AMD release the zen 4 in a week. So I will wait to that!
Yes CPU speed and SSD speed are important, to put it this way my old quad core 8gb that’s 15+ years old takes about 5 hours to move 500gb from an external SSD compared to the new system with faster SDD and 20 logical cores compared to 4 on the old quad which takes around 2 hours to move those files because it then slows down to the external hd speed because my main SSD has a higher throughput.
If you look at the difference in current SSD tech the onboard SSD Samsung 970 Evo Plus NVME M.2 SDD has a 3200mb/s throughput compared to the Samsung 870QVO SATA 2.5 which plugs into your normal power cables & connects to the mainboard via a cable only has a speed of 530mb/s so you can see the speed difference just in SSD & the way they connect to the mainboard using what technology.
You’ll run things much faster on a high-speed SSD they are worth the slots on the mainboard compared to old plugin Sata drives which are much slower due to the cabling, it lessens the throughput because the data has to travel over the cable which will always slow the rate of data flow compared to a direct connection of the NVME M.2 which obviously has a much faster rate of transferring voltage.
Faster CPU’s help it all read quicker due to having more access to memory for the processors to perform tasks. Think of it this way with CPU’s and the power range you have groups of smart people & dumb people the smart people are 4.5GHZ and the dumb people are 2.5GHZ each physical core has two brains so let’s put that in human terms each side of your brain left & right to make the whole brain is the logical cores of the one physical core, you have 10 people rated at 4.5GHZ compared to 10 people rated at 2.5GHZ the 10 people with the higher GHZ will work faster.
Comes down to memory access, faster cores will have more access to memory to store needed functions to access files quicker with things like paging which stores files for quicker access, much like shaders cache when first loading so it saves time reloading the same cached shader.