We are changing how we test GPUs FOREVER!

Рет қаралды 5,529

Күн бұрын

We have finally decided to upgrade our GPU testing hardware for future videos. We are also in the process of changing how we test everything as well to make our results a bit more compelling for both Windows and Linux users.
Hyte Y70 Touch Build: • Hyte Y70 Touch - Feel ...
#gpu

Пікірлер: 22

@tonyo4437 8 ай бұрын

I wouldn't mind a video of the new studio area. Keep up the great work. I'm glad to see you guys taking our input to create even better content.

@KlaireGearSeekers 8 ай бұрын

We would, but we don't have a new studio! Still in the same place ❤❤

@mirkodiciano4647 8 ай бұрын

looking forward to the new videos especially older cards and the thunderbolt testing sounds awesome

@jwaye89 8 ай бұрын

Banger of a video fams. Keen for the new content. Keep up the hard work

@oilpatch432 8 ай бұрын

Love it. Thanks Nic and Klaire!!

@siq41 8 ай бұрын

Enjoy your videos. You always give useful information. Please test with different cpus.

@lightninguru26 8 ай бұрын

i appreciate the attention to Linux. Thank you.

@JarrodsTech 8 ай бұрын

Pad instead of paste is smart, consistent results and don't have to clean that shit up every time. My test system doesn't have the GPU release button either and I get mad every time lmao, can't reach it with these days giant GPUs smh 😑

@mkatakm 8 ай бұрын

I heard you loud and clearly so from now on I am testing GPUs after inserting them into a PCI-e slot.

@Weezedog 8 ай бұрын

What I really want to see is a reviewer test GPUs in different physical orientations, such as I/O out the top (TT Tower 100, Jonsbo VR1 style) which is known to be a a problem with certain heat pipes and gravity vs temps in standard horizontal orientation. No one tests this, but some cards get a massive 20C+ increase in temps and high fan speeds when oriented that way, while other cards work fine and there’s no way for the average consumer to know what cards work and which ones don’t.

@techaddressed 8 ай бұрын

Wasn't expecting to hear your test bench had a Linux install. #oneofus

@GearSeekers 8 ай бұрын

We always test with Linux ;)

@se3000 8 ай бұрын

Please add 1% and/or 0.1% lows! And please make a video benching CS2 with different GPUs, would be especially interested how Nvidia and AMD perform with vulkan on Linux.

@queegfivehundred8197 8 ай бұрын

FYI the armory crate installer is a BIOS option that's enabled by default, so in my opinion it's something that should be checked and disabled when setting up a new ASUS motherboard or updating/resetting its BIOS (or ASUS could do the right thing and set it to disabled by default, but I won't hold my breath for that).

@GearSeekers 8 ай бұрын

Yes I am aware. The issue is most of the time when it's disabled it will not install the LAN drivers as mentioned in the video ;)

@queegfivehundred8197 8 ай бұрын

@@GearSeekers Ah, okay, I must have missed that bit :) I'm old school enough that when I setup my new machine (the first with an Asus motherboard since the A7V133) I did so offline and only connected the wifi once the chipset drivers, VGA drivers, network drivers etc. were manually installed from USB.

@OlavAlexanderMjelde 8 ай бұрын

I'm using 8000 Mhz RAM also in X670E Gene, maybe it's the same kit. It was called: F5-8000J4048F24GX2-TZ5RK I am not sure I gain any FPS in gaming with it, since it's got a higher CL than slower RAM. But I just had to test it on AM5. First tried it with agesa 1080, then I'm on AGESA 1100 beta. Both work fine, but the measured latency is a tad "high" compared to slower kits. Didnt try to tighten timings. But I kindof see better CPU scores but mainly same FPS scores. Running it on AMD 7950X which I slightly overclocked in the PBO settings, along with a 15 curve undervolt and load line calibration set to 5. I also set it to boost +150 MHz in PBO and I set optimized phases instead of auto. Memory phases to extreme is set by Asus on default for these XMP proriles. So far so good, Agesa 1100 beta actually boost quicker than the AM4 platform I had. Agesa 1080 did boot a bit slow when training, but was "ok". The 1100 beta however is much quicker in booting. The bios that came with the mainboard I dont know what was, but it booted very slow.