I know there is always yelling of calling people shills left and right. However, I think it is justified to suggest bias when you compare across multiple brands for one market (consumer graphics), to only one brand of a different market (workstation graphics) without mentioning that one of the consumer brands you compared has a workstation product.
The video in question:
Graphics products compared:
| Market: |
Consumer |
Workstation |
| Intel |
Yes |
|
| Nvidia |
Yes |
Yes |
| AMD |
Yes |
No |
For the uninitiated, Intel does not make any kind of workstation graphics product.
All they had to do was say: “AMD also sells workstation cards, but was not willing to give us any to test with” and I would have no issue. Instead we get this gem:
With all of that said, Team Red shouldn’t act too high and mighty here, because big and hot Vega 64’s impressive lead melts away immediately when you pull out the big guns; or rather the really small guns. This is the entry level and power-sipping Quadro P1000 that stomped all over all but the highest-end consumer cards.
Emphasis is not mine.
I should note that the Vega Frontier Edition was included in the comparison, but was not mentioned by name, unlike P2000 and Titan V.
The omission of EPYC from the CPU comparison is less glaring because its market competitor, Xeon, is barely mentioned.
I wondered the same thing. And I noticed the comment section was mostly yelling about this too. I’m not sure if he even reached out to AMD for any Radeon Pro cards, because it would have been nice to see some of those included in the benchmarks.
I read through all of that just to find out it was that dork you were talking about?
I thought putting the video at the end would dissuade anyone just wanting to yell “shill”; but it does make it hard to see what video the post is about. I’ll edit it.
I was sure that they had some ‘w’ add cards unless they had to send them back or something.
Not sure what I am supposed to find surprising, That Linus is actively doing and saying things that hurt my head, or that you just figured this out.
TL;DR of the video: Downright misleading
Solidworks… What about Catia? IronCad? MatLab (does simulation)? AutoCad?
There are speed benchmarks. But what if all the 64 bit calculations are inaccurate? FEA with inaccurate numbers is probably fine (it is not).
The “oddities” to the long time watchers of the GPU market stick out as expected. Different GPUs do different jobs.
As others mentioned, the lack of Xeons and Epyc makes this “guide” a very strange guide. The lack of FirePro and even Tesla GPUs makes the video worthless.
Feels like a marketing video for Solidworks and Nvidia. And someone did not know how long FEA takes on more complex parts than a steel beam.
No worries. I just can’t stand the guy and haven’t been able to force myself to sit through an entire video of his.
Edit - It’s actually my fault. I should know a quote with emphasis on that many words came from someone I probably can’t tolerate.
Most of his stuff seems fine to me, if a bit click-baity. The reviews seem fine; the tours of some company/place are usually interesting, even though they are presented with an advertisers’ enthusiasm.
More than half the time there is something interesting, but I completely understand how the ad/hype/clickbaity-ness can be irritating enough to ruin most of the videos.