• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

MSI GeForce RTX 5080 Vanguard SOC

W1zzard

Administrator
Staff member
Joined
May 14, 2004
Messages
28,968 (3.75/day)
Processor Ryzen 7 5700X
Memory 48 GB
Video Card(s) RTX 4080
Storage 2x HDD RAID 1, 3x M.2 NVMe
Display(s) 30" 2560x1600 + 19" 1280x1024
Software Windows 10 64-bit
Vanguard is MSI's new addition to the company's GeForce RTX 50 product stack. It sits slightly below Suprim and offers a similar premium design with a strong cooler and factory OC. In our testing, the Vanguard's thermal solution is actually the strongest of the 10 RTX 5080 cards that we've tested so far.

Show full review
 
How does the Vanguard have the best cooler when the page 39 data shows that the Suprim is quieter with lower temps? The only thing thermally in favor for the Vanguard is the slightly lower idle temp which does not really matter. Unless the numbers in that table are not correct.


MSI RTX 5080 Suprim SOC
60°C​
62°C​
25.5 dBA​
1133 RPM​
MSI RTX 5080 Vanguard SOC (Quiet BIOS)
64°C​
68°C​
30.2 dBA​
1276 RPM​
 
How does the Vanguard have the best cooler when the page 39 data shows that the Suprim is quieter with lower temps?
Compare power draw = heat output of the cards in gaming. The cooler test normalizes that
 
im attaching less and less value to those TPU awards.
 
No PNY review?

I thought you guys partnered with them.
 
I would love to see the review (mostly the cooler performance) of the Gaming Trio, the white one looks fantastic and a bit cheaper too
 
No PNY review?

I thought you guys partnered with them.
He's just put up 10 RTX 5080 reviews in 2 days (more than any other site afaik) give him a break, im sure there in the pipe line somewhere.
 
Your reviews (with big chunks of the conclusion page copy-pasted) are based on the price being $1000. From your own tables, every card except Gainward's is expected to hit at over $1200. It's quite clear that the FE cards are there just to skew the product's value perception, with minimal real availability. Why are you falling for this?
 
Your reviews (with big chunks of the conclusion page copy-pasted) are based on the price being $1000. From your own tables, every card except Gainward's is expected to hit at over $1200. It's quite clear that the FE cards are there just to skew the product's value perception, with minimal real availability. Why are you falling for this?

The copy/paste conclusion is used because the cards all use the same chip. The features are all identical. It would be an experiment in inefficiency to somehow create a 'new' description for every new card based on the same chip. Also, if he didn't repeat the info, and you only read one review, you'd wonder where the 'full' content was. If the conclusion simply read <see FE review>, that would seem excessively lazy. I look at the heat/noise sections, as that's really where things differ. OC'ing too.

As far as price is concerned. It's what W1zzard is given to use. I checked UK online yesterday and saw pre-order prices of £979 up to the crazy stupid Asus pricing. Besides, he offers a speculative, generalised current pricing model on the price/perf page.

This is a site that reaches a global audience - it's impractical to expect the prices (that will fluctuate across markets) to be dead-on. Then again, if you look at the price performance section, you'll see that he actually does incorporate changing values - once they are known in real terms. In pre-release, he uses the MSRP. As is standard.
 
Frame Gen is another gimmick.
Its a negative for the PC Gaming Industry.
 
Somehow the 7900 XTX looks more desirable to me today than it did 2 years ago.
 
It seems like you can easily get 300-400 MHz extra out of all 5080s at stock voltage. That's pretty crazy. It should also mean that the undervolting potential is amazing. I hope it'll be similar with the 5070 Ti, which is supposed to have even lower stock clocks, although those cards will obviously have lower quality silicon.
 
The copy/paste conclusion is used because the cards all use the same chip. The features are all identical. It would be an experiment in inefficiency to somehow create a 'new' description for every new card based on the same chip...

And that's absolutely fair for metrics that are identical or very close, but not when it comes to parts like

Priced at $1000, the GeForce RTX 5080 actually releases cheaper than the RTX 4080, which launched at a $1200 MSRP. The refreshed GeForce RTX 4080 Super brought that down to $1000 though...

For $1000, there is no reason you should buy RTX 4080 or RTX 4080 Super now. AMD's Radeon RX 7900 XTX is $820, or 18% cheaper, but it's also 15% slower in raster, and 38% slower in RT...

when used in the conclusion section for a product that costs north of $1200. It just doesn't match reality anymore. We are not talking "dead-on", it's 20% difference in price.
 
Somehow the 7900 XTX looks more desirable to me today than it did 2 years ago.
There is only one 7900 XTX model in stock at my local MC. It seems AMD is not making them or making very little. It's $900. There is a $1,150 5080 in stock. For $250 more I get almost playable 4k Wukong at 39fps (v 31fps), Stalker 2 at 60fps(v 46fps), and SH 2 at 48fps(v 31fps). Needless to say much bigger RT improvement and DLSS. Also full speed DP2.1. If you have $900, you have $1,150. For each his own though. I'm not a hardcore fan. I just buy the best I can afford. I'm currently on a RX6950 XT and no choice but to get a 5080.
 
I left this comment on the YT video comparing 10 models, I'll leave it here too:

could you make a "cooler performance per liter" showing how "space-efficient" (or inefficient lol) these cards are?

i.e. basically a noise / temperature / size (and power?, but just divide) normalized comparison



i.e. by dividing the Watt TDP by the liters/volume the 5080 FE is 211 "Watts per liter", the Gainward Phoenix is perhaps >30% behind,

but then normalize these numbers with the noise normalized temps
(and then you could do it with other cards, i.e. the 5090 FE is 50% better before adjusting for cooler performance, but perhaps there is a large 5090 that is "on par" with the 5080 FE
 
On those reviews I would like to see included/mentioned if fuse are used or not on the pcb, 12V from PCIE or connector from psu.
 
Back
Top