icon 字幕
正在載入字幕...

Best Budget Home AI gpu of 2026!! It destroys the Instinct Mi50.. without ROCm: AMD Radeon Pro v620

youtube 翻譯 youtube 中文翻譯 youtube 字幕 youtube 中文字幕 youtube 翻譯成中文 youtube 視頻翻譯 youtube translate to chinese translate youtube to chinese youtube transcript to chinese translate youtube video to chinese

YouTube transcript, YouTube translate

32/32

A quick preview of the first subtitles so you know what the video covers.

Are you in the market for a GPU with 32 GB of RAM that hardly anybody has heard about and the prices aren't through the roof? $750,000 $2,000 yet? If so, this video is for you. Hey, howdy y'all. Welcome back to the show. It's your man David with Country Boy Computers. If you're in the market for a 32 GB GPU that costs $350 and outperforms an MI50, stick around. This video is for you. The action starts right now. So, I've already let the cat out of the bag a little bit with last week's video, but that was in Windows. Today, we're going to be running Linux. First, let's test the V620 in Linux in Vulcan and see what she can do. All right, real quick. I want to show you what I'm working with here. Show you what I'm working with. System about our test bench is again the ASUS Prime X299 Deluxe 2. We've got Ubuntu 26.04, 4. I have Rockm 7.1 installed. 64 gigs of RAM, three different uh NVMEs totaling 7 tab here. So on the left is my website where I ran both the MI50s and the V620s in Windows using Vulcan only. And today in Linux, I'm going to first test Vulcan and then I'll test HIP. So, I've got the V620s loaded. And let me show you this. We'll get Mission Center, open, and moved over. And all right, we've got Quinn 3.6 27B at a Q8 loaded up. And here we go. We got flash attention off. That's going to come into play here in just a minute. 309 313. All right, there we go. We finished up with 356 tokens prompt processing and in Windows under the Vulcan we got 303. So quite a big improvement there.

設定

100%

翻譯目標語言

🔊 音訊播放
正在播放翻譯音訊