Name: 這款RTX 2070實際售價為MSRP!但它值得嗎？ (This RTX 2070 ACTUALLY costs MSRP! But is it worth it?)
Uploaded: 2021-01-14T10:06:15.000Z
Duration: 11 min 31 s
Description: 【看影片學英語】數萬部 YouTube 影片，搭配英漢字典即點即查，輕鬆掌握單字發音與用法，長久累積看電影不必再看字幕。

Jay's two cents here and welcome to my home office, where if you're a long term viewer of the channel, you remember I made a lot of my videos here.

程度副詞

Hundreds of videos were made in this room and because the winds kicked up in Southern California and unfortunately, they kicked up trees right out of the ground and broke power lines and is just a mess here in Southern California.

So I am here at the Home Studio Home office, basically making a studio again with a very temporary setup.

I'm gonna be bringing in my review here of the art e X 2070 Fermin video, a highly anticipated launch and one that I think most of us are very nervous about, because obviously the 2018 20 t I lunch went without with that much enthusiasm, I guess you could say it was a terrible price to performance ratio.

So let's see how the 2070 stacks up sort of recap.

If you guys were not paying attention to the R T X launch back in September, it was, um yeah, it's launched almost a month ago by the time you guys are watching this video.

I don't know many reviewers who gave it a very positive review, because although the graphics card is a huge leap forward in graphics technology or GP technology, obviously we've got the 10 secours in there.

We've got real time Ray tracing, which, by the way, at the time making this video.

We still have no titles that truly take advantage of Ray tracing.

We've got the final fantasy 15 demo deal SS demo, which is still not Ray.

Tracing is just using the tour touring cores for anti a leasing processing.

And we've still got the Star Wars Reflections demo, which is not a game that is an extremely Taylor demo designed to make rage.

Real time Ray tracing look amazing, which is not indicative of real world performance.

So I don't even show it because it's really just it's not indicative of what the actual user experience is gonna be like, but the price to performance ratio on 2080 and 28 e t.

I made it and extremely lackluster launch.

Sure, it was faster upwards of 25 30 40% depending on the title, but at an increased cost of up to 70% compared to the 10 80 t I, where we're used to seeing generational improvements in the 25 to 30% range at the same price.

Unfortunately, when you compare the very comparable price of a 2080 to attend a t.

T I, you saw that it was anywhere from 0 to 9% which was not great.

So that's where the 2070 comes in a card that I was extremely interested in taking a look at because I wanted to see how would a video is going to fit this in?

Let's look at a M D for a second here because A M D doesn't have a very crowded stack.

We have the 5 65 75 80 Rx cards, obviously, and then we've got Vega, 56 in Vegas, 64.

We've got basically five cards that make up the entire product stacked if you look it in video.

On the other hand, we have got not including the GT 10 30 because that's not really a gaming card at all.

That's just a card to give you video for, like, home theater or just browsing the Internet.

50 10 50 10 t I The 10 63 g the 10 66 g We've got the 10.

70 10 70 10 t I that 10 80 10 80 I tight next Pascal Titan V.

Although it's not a gaming card playing, people are so gaining on them.

That's 13 cards that you could buy today.

The other thing that made the r t x 2080 and 28 e t.

I launch extremely disappointing was the nearly impossible.

M S r p price point to reach, Remember in video is also in direct sales.

They had the Founders edition card, which is basically $100 premium on top of the M S r p even J gig by a souse p n Y m s.

I mean all these companies that are also making a video graphics cards and buying cores from in video are competing against their supplier.

So it's a very interesting market that we saw starting with Pascal so in video is charging $100 premium on their cards because they are basically bending them and keeping the best ones for themselves and slightly over clocking them.

The babies are having to compete with that, and they are not basically capable of making cards m s, r p and still make money.

If they did that, they would be out of business before you knew it, because there would be no profit.

There is very little market whatsoever when it comes to graphics cards, and so it's very difficult for them to do that.

So the 2070 we're looking at today is kind of a big deal because it is an M S r.

J that being the arty X 2070 black edition.

Now, one of the ways they kept the cost down on this card is it doesn't have any frills.

It doesn't have a back plate, it doesn't have ah, triple slot cooler.

Actually, it's Ah, the same style fan as you would find on the exit Ultra.

It only has one power plug, so it's not really built for high end over clocking.

It's going to be stuck to the power limits.

They're available to it through a single PC.

I express a pen as well as the power being delivered through the PC I express, so it's gonna have a lower power target on there.

It's not going to over clock very far, but for the sake of this test and comparing it to all the cars that we've already benchmark recently here with these launches, everything here and see in this video is out of the box settings anyway.

So the 2070 Founders edition card boosts up to 1710 megahertz.

Where the reference cards are advertised is boosting up to 1620.

But actually saw on this card with the temperatures kept really, really low with stock fan curve, because again, not a lot of heat coming out of this corps.

It is seeing up to 1860 megahertz and then dropping down into the mid seventeen's when it's under load and the temperatures have sort of equalized and saturated.

But keep in mind pretty much all of the artifacts reviews you're seeing, though, and it comes to temperatures 2080 and 20 t I included.

The tensor cores are pretty much sitting there idle, not doing anything, and those will generate heat.

Once title start calling for those course, Do something specifically deal SS being the main thing that's gonna be asking for the tens, of course, to do something.

So as soon as those start becoming more active and in play, the temperatures on all these cards are probably gonna go up pretty significantly now.

The V G 8 2070 Black was sitting at 62 c max temp in all of my testing.

That's fire strike, fire strike Extreme fire strike Ultra times by extreme, of course, the game titles that we've tested.

And I found that to be actually pretty good.

Now that is obviously in an open air test bench.

But the reason why we do that is because we don't want the case to introduce any sort of impact on the cooling when we're testing cooler capability on the graphics card.

So we allow it an open air environment so we can see what the max efficiency cooling capability is of the cards.

Obviously, you're cooling eyes going to vary depending on your case in ambient temperatures and your environment.

Now, speaking of testing, let's go and take a look at those benchmarks.

You could see how it stacks up to the 10 70 10 70 t I Vegas 64 a 10 80 all sitting within $100 of the price of this card, and you'll see that this is actually what we were kind of expecting to see with the 2080 and 28.

So this is what we were kind of expecting to see.

What the R T X launch cards had actually become available at M S, R.

Compete against their past generational brethren, smack them around a little bit and do it at the same price point.

That's what we were expecting to see Now a lot of people might look at this and say, Well, then why wouldn't you just run out and buy a GTX 10 80 or GTX, 10 70 t I and call it a day.

Well, that might make a whole lot of sense, but the problem is, we're still waiting for features that are not available yet.

These cards are technically trying to future proof themselves by waiting for features to come online as developers implement them.

So this is still an investment into the future.

Whether or not that's ah sound investment is still waiting to be seen.

Demos are not gonna tell us the big picture.

That's why I include synthetic benchmarks in my testing as well as actual game titles, because sometimes you'll see a certain card may take the lead in synthetics.

But as soon as you actually put a more real world test on there or start looking at titles that actually leverage a deep you in different ways and synthetics, which are very consistent and very optimized, then you'll see that the picture definitely changes.

The Vegas 64 does really well in synthetics, but then, in real world testing, it definitely trades blows very differently than it does in synthetics.

That's one of the reasons why we test that now we're gonna be taking a look at 27 custom cards that well as well.

But I found the MSR peak are to be sort of the unicorn, which is usually never ever found.