Just two cents here.

And I haven't done a video like this in about six years where I'm gonna specifically talk about power draw.

It's an argument we've seen over and over and over get blah, blah, blah company because yada yada company is going to cost you more in the long run in terms of power consumption, we're gonna talk today specifically about various graphics cards and see pews, especially with the way the CPU market is heating up.

No pun intended.

Actually, things were kinda getting cooler.

All these for A and B Intel is gonna get hot.

But that's what we're gonna talk about, how much money it may cost you to get one brand over the other and whether or not that should even be a factor in your buying decisions.

I wasn't buying purchases.

You usually purchased your bias with its aluminum vessel, solid construction and adjustable base.

The vast 35 inch, ultra wide 100 Hertz V.

A panel from Master up offers performance at an affordable price.

Learn more about this drop by following the link in the description below.

Okay, so here's what we're gonna do it.

There's a lot of different ways that you can do this.

And the unfortunate thing about this test versus the one I did six ish years ago was the fact that we had a lot more simple product stacked versus today.

For instance, when I did this last time, I believe I was running Z 77 3700 k, which Ivy Bridges, you know, was kind of like the top dog.

And then you had your eye sevens which was hyper threading and your eye flies which were not by threes, which we're not a couple skews of each.

And then we had FX, which was four course six court acorn O S and t know hyper threading very basic, very easy to compare.

The problem now is the on the CPD side of things.

The battle heating up.

Ever since risen launched with 1800 x and down, the competition has just been getting fierce with more skews filling in the gaps villian, the price caps, the performance gaps.

Intel doing the same thing until actually having annual launches once you get once again, some of them being less than a year apart has created some serious confusion in the marketplace in terms of what the direct comparisons should be.

It's something we struggled with as we've done.

See Pew comparisons as we've done gaming comparisons, it's re coming to the point where we don't know exactly where to line them up.

Do we do it core for core?

Because when we do that, we have.

The CPI's might do this, but the pricing structure more less does that, so it becomes very difficult to compare.

So that is why, for the sake of this video, we are kind of doing core for core because that's the best way that we can determine what apples to apples comparison would be in terms of power.

Draw eight core 16 thread risen 2700 x versus a 99 100 K, which is also a core 16 thread, albeit much faster.

Clock in Turmoil Club for GPU.

On the other hand, it's ah, it's a little bit easier to sort of line those up because the right ready on seven, which is the best card that you can get right now from AMG for mainstream and even productivity, you know, being that this is the top price point, the problem is the top price point for a M D does not meet the top price point for in video.

That's lots more or less against 2080 28 e.

T.

I costs approximately $400 Maur and then obviously gives you more performance that goes along with that.

So even though I'm a 28 e t I sitting on the table, we're going to use 2080 versus Radion seven and then a 20 i for shits and giggles.

And then I would explain to you how we're going to the CPU.

We're not talking about thread Ripper.

We're not talking about ex to 99 or any of those, because what I just explained in terms of how difficult it is to match everything up wasn't even including the extreme platform.

The nice thing about AMG moving forward with Rice and three a resident to 3000 Siri's is the fact that there are more skews available for mainstream meaning, you know, having and then taken into account the cost of motherboards and extra ram channels and all that sort of stuff.

So here's what you need to figure out is before you get to determine how much something's gonna cost you to run.

You need to know how much power uses and how much you're using it.

So, for instance, I reached out on Twitter because I thought, Man, I've got about 520,000 followers there.

That's a huge data set if even a fraction of those people respond.

So I put out a tweet saying, Hey, guys, on average, how much do you gain per day?

How many hours?

Um, as expected.

It was a pretty wide range, but I saw a lot of I don't game at all anymore.

I'm too busy, which unfortunately, doesn't really help.

But they don't know the purpose of this video.

So that was still a fair answer to some people gaming more than 10 hours per day, which definitely could rack up the amount of cost in terms of wattage pricing in that sort of stuff.

So this is gonna be an interesting data set when I found was the media and tend to be between 3 to 4 hours.

So for the sake of calculations today, we're using three and 1/2 hours of gaming time for our tests.

The other thing that gets a little bit convoluted here is trying to determine C P Power draw because of its constant fluctuating load with the GPU.

Kind of tends to do nothing unless you're gaming.

GP was a lot easier to calculate.

So that's why we're gonna go ahead and start there now.

Using G grew three d, we I am gonna put our link to our sources as well as putting on screen.

They're charging stuff so you can see these are this is data sets that we've been getting from the Internet.

We looked at a bunch of different sources and not just our own, because we want to see kind of what the median is that way.

We're not just going off one data set, not to mention Guru three D puts a list of their specs for the system the same system on everything, so it keeps it very consistent.

So for graphics and with them being guru three d, any kind of sense, right that they would have graphic starts the radio on seven at three and 1/2 hours per day, using an average of 15 cents per kilowatt hour.

And yes, some of you might look at that and go.

Whoa, that's expensive.

If you live here in Southern California in the summer time, you could go.

Wow, I wish I was only paying 15 cents per kilowatt hour.

So once again, we had to kind of find a middle ground to call, you know, fair.

So at three and half hours per day, the radio on seven is gonna cost you approximately $4.78 per months to run or $57 and 34 cents per year, compared to a 2080.

It's gonna cost you $3.68 per month to run and $44.10 per year, or a difference of $13.24 per year.

That's kind of a wash if you think about it, because the radio on seven although the H P.

M to and the Vega architecture and and all that stuff has been known to be a bit power hungry because it uses a lot of has a lot of heat, needs a lot of cooling and that heats generated obviously by power draw one of the things I like to do that when it comes to cost calculations is kind of average them over five years because I find that five years is kind of the the time frame when a lot of people will upgrade their systems.

Yes, I know many of you are running much older systems in that.

And, yes, our people out there that do annual upgrade so they don't care about the power costs because they're never gonna make it back anyway.

So at $13 times five years, obviously you're talking about $65 in cost difference between the radio on seven versus the 2080.

No, if you spent that extra $65 on a graphics card, guess what?

You wouldn't have gone and you were higher than a 2080 because the next bump up is $400.

So at this point, that's why we're saying it's kind of a wash.

It makes no difference because directly comparable in price, very comparable in performance in over five years.

At a $65 power difference, I still would call that one pretty name close.

What about CPS?

Well, the si pues this is interesting, and he's always been known as the space heater.

If you're cold in the winter time, get yourself a name.

DCP, you know, keep you nice and toasty.

That's always been the joke, right?

The MD heater?

Well, the 9900 k.

We had a hard time finding a lot of, I guess, comparable data sets when it comes to load and all that sort of stuff because we found a very wide range of testing in terms of power draws.

Some people were normalizing my removing idol wattage and giving us just load wattage.

Some people were measuring from the wall.

Some people were over clocking.

So what we found here was we're using text spot here and their particular charts for blender and your blender, gooseberry and handbrake.

So obviously CPU trance coding in hand, brake and then CPU rendering in blender Blender does his GP as well, but obviously doing C P test.

So the 9900 k in hand brake pulled 255 watts system water.

Just total system wanted to put out there three and 1/2 hours a day would cost you $4.7 a month, or $48.90 per year to run the 2700 ex.

Ironically, considerably less power draw is pulling 209 watts from the wall for this whole total system, or $3.34 a month, $40.8 a year or a difference of $7 and some change on a map per year.

Now here's the important part.

BC Pews are not comparable in terms of pricing like it was with the GPU, the 2700 X.

And, yes, I know this sounds like a name be sponsored product video.

It's not.

This is J.

J.

Sponsored this my stuff.

We're talking about this because this is an argument that I'm tired of seeing people say, without any sort of experience in the subject matter that cost to run.

The system can make a break the difference in terms of which, where the value is because, as you can see, a difference of $7 some change per year between 2700 X in the 1900 k does not make up for the 40% extra costs of a 90 Niner kay verse 2700 x.

This is all I'm doing there.

This is the point stays videos to get you thinking this way.

There are certain things that are okay to consider.

There are certain things that are not okay to consider when it comes to the true value of a product.

Had the has been the same price, even the value would still technically be with the 2700 X if you're using that argument.

But we all know that 9900 k is approximately 10 to 15% faster, depending on the workload.

In fact, in sin a bench, it was a 2038 score versus an 18 08 or approximately 12.7% difference.

So as you can see right here if you kept both systems for five years at $7 a year difference, we're talking a difference of $35 power cost of a nd or intel.

Wow!

See?

See how in green we were to believe am these faster.

And I just did that myself because that was the case.

It used to be the case.

I p.

C was terrible.

Power drop was terrible on FX.

That all changed with reason.

Now the other thing I want to talk about here is that's kind of not a fair test.

What I mean by that is most people are not running there.

See fuse at 100% load for three and 1/2 hours a day.

Because we've shown on this channel many times.

Gaming load on CPU and less bottlenecked is not 100%.

It tends to sit around 40 to 60% depending on the game.

Games that are very GPU bound are gonna be a lot less intrusive on the CPU load, which means even lower load loads on the CPU, which means an even narrower gap when it comes to power drops.

That's good news for Intel.

Not so much for Andy, but at the end of the day we said a 40% cost difference between the two.

The point of this conversation is to tell you because you're not 100% load.

It doesn't matter what the power draw on CP was gonna be unless we're comparing something like an I three to a W 31 75 X, which obviously would not be a very fair comparison because we'd be talking more than 1000 watts on the CPU versus, like 50 watts on a night.

Three.

So at the end of the day, if you keep your system for five years, usually the cost savings is not enough to even bump you up into the next year.

That's been the case as long as we've had this CP battle going on.

But that's all getting ready to change because one of the things we're still waiting to see with Zen to or or or technically third gen reason.

I guess I don't have what you want to call it, is improved.

Power drill improved I P C improved power draw.

And a lot of that comes from the fact that it's running on seven Enemy to process and not 14 plus plus been fit like Intel still is.

Even when Intel gets its 10 centimeter going on desktop, we could still see a better power draw coming from a MD vs Intel.

The smaller the process, the less power doll, which is something that is, you know, kind of been Murphy's law and all that stuff for a long time.

I see Murphy's law Is that Is that Murphy's law?

Uh, Maura's Law?

I didn't start with him, but J.

What's the whole point of this video?

The point of this video is just to say by what works for you.

Seriously, The most ignorant comment a person can make is the power draw argument.

It means no one's anyone that says that it means they haven't done the math.

It means they haven't actually sat down and thought about it.

So, honestly, guys, at the end of the day, we have to wait and see what's gonna happen with the next architecture from a M D.

If we see these eyepieces come up and on seven and a meter, it's matching and exceeding possibly the 1999 100 k as just arbitrarily throwing it out there.

The values already with Andy in terms of price of the CPU, it's already there in terms of power draw.

And you know what?

Are you willing to pay 40% more for 10 to 15% more performance?

Plenty of people do.

There's nothing wrong with it.

It's your choice.

And that's the whole point of today's video.

Ignore the power drill conversation and go right with the performance conversation and get where the performance and the dollar amount makes sense for your wallet.

All right, guys, it's just one of those topics that inside to do a talking head on because of the amount of comments I've seen over the brutal battle that is taking place amongst the Fanboys regarding a M.

D and Intel.

Rightfully so.

I mean, it's it's okay to like what you like, but I think what's happened here is aimed.