TR Forums

Just watched the review of the Apple watch 4. Meh. I was expecting a few days battery life per charge at least.

Too much going on for no real purpose other than the heart stuff. Pretty full color graphics? Cool... until you look at the battery life of a whopping 18 hours, and can only do 6 in fitness mode. No marathoners need apply. Fancy graphics really belong on the phone, the watch should be more of a sensor hub IMHO. Useless fluff. Why do people get so excited about something that you can't even track sleeping if you wish after a full day of use? It is just expensive wrist candy.

I actually like to go backpacking for a week at a time on occasion, or go out for a weekend someplace for an adventure. Not having to charge my watch in that time is kind of nice.

So I understand the limitations of the crew running the show at the Tech Report. It is a side-gig. I would do some reviews of wearables myself if I wasn't terribly busy, and frankly, a terrible writer. It would be fun if TR users would donate/loan their toys a week or so to add to the reviews, but that would require a lot of trust to get them back in working order without the top popped off to see what makes them tick.

Here's a TR review idea:

"Alexa, cook me a Hot Pocket"

Oh, I can take care of that one right now.

Just 700w, too small to microwave my personal-family-sized Stouffer Mac and Cheese. 0/10, would not recommend.

Yeah that's fair. I've got a 1200W unit that hangs over our oven, I like it quite a bit. Alexa integration wouldn't get me to give it up.

I think Engadget said it best:

Just... why?

Captain Ned wrote:

Oh, and this too:

Nice rack.

I'd read a review of the 3D printer SecretSquirrel posted in another thread. Based on responses I think I'm not the only one with interest.

Captain Ned wrote:

End User wrote:

I take it you don’t have an electric car.

Not when this is possible:

Er, electric cars can operate in such conditions:

https://teslamotorsclub.com/tmc/threads ... ing.64591/

Captain Ned wrote:

Oh, and this too:

You have a stuffed cat?

End User wrote:

Er, electric cars can operate in such conditions:

Nowhere near as well as a Subaru (like the one under 3 feet of snow in that pic). Plus, I don't lose 1/3 (or more) of my range just because it's cold AND I can fully refuel in less than 5 minutes. I'll buy an electric car once they all carry a Mr. Fusion unit.

End User wrote:

You have a stuffed cat?

Only when the mice are running (spring & fall: old house, no way to permanently eliminate mice). That's her favorite surveillance post. 15 YO and still a lethal mouser.

DancinJack wrote:

I don't necessarily disagree with some of the things you think would be good to cover, but the idea that the RTX series of cards and their reviews are alienating people is....I just don't agree

The excess focus on 4K has alienated me for a long time. I'm still obviously a reader but it's hard to even care about a lot of their GPU articles.

LostCat wrote:

The excess focus on 4K has alienated me for a long time. I'm still obviously a reader but it's hard to even care about a lot of their GPU articles.

Given the RTX numbers at 4K, are we really going to learn much by retesting at 1440p??

Captain Ned wrote:

Given the RTX numbers at 4K, are we really going to learn much by retesting at 1440p??

I wouldn't think most people prefer to make assumptions vs having hard data. *shrug*

This is the only site I've noticed only doing 4K numbers, and with respect to their time and resources I've always hated that.

LostCat wrote:

This is the only site I've noticed only doing 4K numbers, and with respect to their time and resources I've always hated that.

I'm fairly certain this isn't the case, but maybe you just don't read some of the other sites that only did 4k testing.

Captain Ned wrote:

LostCat wrote:

The excess focus on 4K has alienated me for a long time. I'm still obviously a reader but it's hard to even care about a lot of their GPU articles.

Given the RTX numbers at 4K, are we really going to learn much by retesting at 1440p??

"It's faster at 1440p than it is at 4K". Which isn't really something we'd learn, I guess. IMO 4K gaming hasn't quite arrived, even with the 20x0 series. When I can run a game with V Sync enabled and not have it dip below 60fps, I'll be the first to sign up. Until then, I'm happy for everyone else to fund continued research.

DancinJack wrote:

I'm fairly certain this isn't the case, but maybe you just don't read some of the other sites that only did 4k testing.

I'm not saying it's the case for the 2000 series reviews, because honestly I only looked at a few sites. I have little interest in any generation starting at $500.

Every other card series reviews I've seen TR were the only ones I've seen obsessing about 4K.

LostCat wrote:

I'm not saying it's the case for the 2000 series reviews, because honestly I only looked at a few sites. I have little interest in any generation starting at $500.

Every other card series reviews I've seen TR were the only ones I've seen obsessing about 4K.

Yeah, the RTX series reviews have definitely focused on 4K mostly, but I don't know anyone would expect otherwise. As for your claim about the rest of reviews here at TR focusing on 4K. I don't get that one either.

https://techreport.com/review/32391/amd ... reviewed/3 - From what I can tell everything here is 1440p.
https://techreport.com/review/32766/nvi ... reviewed/3 - This review has a mix of resolutions, but from what I can tell 1440p is the most used too.
https://techreport.com/review/31754/amd ... reviewed/3 - Another that focused on 1440p mostly.

Those are just the last few major discrete cards that were released (sans the RTX series). So i'm just not sure where you're getting your data.

Here's the problem with testing at 2560x1440: there are at least two different ways you could go about it. Are you going to use the rendering power you're freeing up by lowering resolution to increase frame rate at the same settings as 4K testing (which may not include taxing AA methods) or go the opposite direction and crank AA to make for better image quality to compensate for the resolution drop?

We test the same settings in favor of frame rate and whoops, the image-quality people are angry because they're not seeing the settings they would use. We test for image quality and whoops, the frame-rate people are angry, because there's no (major/massive) difference in the average frame rates we report between 2560x1440 and 4K. Did we get CPU-bound in either case, as the RTX 2080 Ti threatens to do in some titles (and, in fact, did, in my preliminary testing for Forza Motorsport 7?) Whoops, suddenly the relative performance of the various graphics cards under test is no longer expressed cleanly. Also, sorry, there's no way we can do both of these cases to satisfy both parties because we only have five days or less to review seven to 10 graphics cards and the workload quickly scales beyond what we have time for.

4K testing does an excellent job of showing what we need it to show in reviews of powerful graphics cards like these, which are the biggest possible differences in relative power that every step up in price buys you under a demanding but still-real-world-applicable workload, all without bringing other system components into the picture. Given a fixed amount of time that we can pour into an article, we have to go for the thing that provides us the most bang-for-the-buck, and 4K gaming is it.

DancinJack wrote:

Those are just the last few major discrete cards that were released (sans the RTX series). So i'm just not sure where you're getting your data.

Eh. I guess it's been a while and I didn't remember them all.

Jeff Kampman wrote:

4K testing does an excellent job of showing what we need it to show in reviews of powerful graphics cards like these, which are the biggest possible differences in relative power that every step up in price buys you under a demanding but still-real-world-applicable workload, all without bringing other system components into the picture. Given a fixed amount of time that we can pour into an article, we have to go for the thing that provides us the most bang-for-the-buck, and 4K gaming is it.

I respect your restraints on time and resources, and I understand these cards aren't for me anyway.

It's just going to be a very long time before I'm interested in 4K gaming or results on PC. As in, maybe 2020 onward.

If my memory serves me correctly, back in the day, the cards used to be tested at multiple resolutions. If there's only time to test 1 resolution day 1, that's fine. Why not have a follow-up article to test the resolution that 1x more people use? (https://store.steampowered.com/hwsurvey.

In addition to testing something other than 4K gaming, testing the potential to use the card for protein folding, crypto-currency, Seti, simulations, etc. would be fun. There are plenty of times I would like to put my expensive hardware to good use when I cannot be there burning time playing games.

What bothers me is I do quite a lot of circuit simulation at work, yet card acceleration for SPICE is still very much a niche.
http://ngspice.sourceforge.net/cuspice.html
The cores of these video cards are enormous matrix solvers, and should absolutely CRUSH through complex simulations.
I tend to use LTSpice (Free from Linear Tech) and I should look into if I can use CUSpice as an engine under the hood. LTSpice is highly optimized for power supply design specifically though, so I doubt it would be easy.

Captain Ned wrote:

Oh, and this too:

That picture goes rather too well with your current signature!

Captain Ned wrote:

LostCat wrote:

The excess focus on 4K has alienated me for a long time. I'm still obviously a reader but it's hard to even care about a lot of their GPU articles.

Given the RTX numbers at 4K, are we really going to learn much by retesting at 1440p??

Aside from finding out when the GPU begins to become CPU-bounded. There really isn't that much you'll get of it.

Krogoth wrote:

Aside from finding out when the GPU begins to become CPU-bounded. There really isn't that much you'll get of it.

Looks like we shouldn't really have that problem much anymore.
https://wccftech.com/shadow-of-the-tomb ... r-geforce/
I'll never agree with that assessment, either way. But whatever works. If it means I stop reading GPU reviews, so be it.