Destop-level gaming performance is certainly what many gamers who have sworn off the clunky tower want. The Mastercase EG200 accommodates full-length desktop graphics cards, supplies them with power and connects them to the laptop via Thunderbolt 3. Besides that, there are interesting extras like an additional hard drive bay, more USB ports or an integrated laptop holder. We check how a dGPU in the case performs against full-fledged desktop PCs.
https://www.notebookcheck.net/Cooler-Master-MasterCase-EG200-in-hands-on-test-Does-an-external-GPU-via-Thunderbolt-case-make-sense.537232.0.html
such a major cpu bottleneck over there 😂
"BuT qUaD cOrE iS eNoUgH1!1!1!"
Should be running the test with newer intel chips too. The older model Skylake and Cometlake chipsets are still using separate thunderbolt controllers, which will limit their responsiveness.
It'd be interesting to run another set of the same tests using icelake or tigerlake chipsets to see if there's any improvement on having integrated controllers, even if the lower wattage chips will probably be outclassed in raw CPU power.
I hoped to see info on noise levels (unless i missed it)... I currently have a Sonnet eGPU box which lacks any USB ports (which would be handy for a 'one thunderbolt cable' solution... but its whisper quiet and sips power (I'm off grid solar so idle watts to me too'. This might be a good upgrade but not if its fan roars all the time or is a power hog.
you really need to pair this with a laptop that has an embedded TB controller instead of external. At least a 28W PL1 i7-1165G7. There's a huge bottleneck when the controller is external to the CPU.
https://www.youtube.com/watch?v=ooKlVR1z_EM
The above comments have a point, it's the controller difference that should perform better with laptops supporting TB4. This is why last week I sent the editors a request if it's possible to test eGUP with TB4 when connected to an midrange laptop with the latest TigerLake CPU like the new Dell detachable 7320 that also supports TB4.