I don’t have any idea what the 4000 series mobile cards will look like. It seems like the desktop ones pretty much just jacked up the power draw to nutty levels to get their performance. Not really a good idea on laptops. I guess we will see if they have something more efficient in store.
They did jack up the power draw, but they are simultaneously more efficient.
They did jack up the power draw, but they are simultaneously more efficient.
The issue there is that despite its greater efficiency the 4080 is still consuming close to the same amount of power as its predecessor-- the 3080.
So laptops probably won't enjoy better battery life or lesser heat concerns, but there will be a jump in performance. However, it's not even a certainty the 4070 Ti (or whatever they label it) will actually be available in laptops one can purchase in the next 4 months. They 3070 Ti launched 4 months after the 3080, and while there were a few units out at the end of January 2022, there wasn't a mainstream supply until several months later. I'd expect the best price on laptops with the 4070 Ti available in the next 6 months will run close to $3K CAD.
Indeed, this is a very strong sale.
Ah, LOL, that's right. Yeah, I catch this stuff all the time. Merchants constantly mislabel stuff. The Gigabyte page I linked there is very clear:Well now I’m confused as to what the hell I even bought: did I buy an XE4 or an XE5? It says XE4 in the name, but the stats said it had DDR5 ram. But if you look at the end of the description it says XE5.
View attachment 955392
So what on Earth did I buy? Guess I’ll find out when it gets here. Stupid newegg.
I looked at it again and for some reason I thought the 4080 was way worse than it is. Looks like it uses slightly less power than the 3080 which is impressive.They did jack up the power draw, but they are simultaneously more efficient.
The issue there is that despite its greater efficiency the 4080 is still consuming close to the same amount of power as its predecessor-- the 3080.
So laptops probably won't enjoy better battery life or lesser heat concerns, but there will be a jump in performance. However, it's not even a certainty the 4070 Ti (or whatever they label it) will actually be available in laptops one can purchase in the next 4 months. They 3070 Ti launched 4 months after the 3080, and while there were a few units out at the end of January 2022, there wasn't a mainstream supply until several months later. I'd expect the best price on laptops with the 4070 Ti available in the next 6 months will run close to $3K CAD.
You're mistaken. You may be thinking back to right after launch when Techspot put out one of the first reviews. That review is what had me stunned because I thought 1000W PSUs would be required, but apparently not quite that much power is needed. Apparently their testing was tainted by a BIOS flaw (one seen in the past). They got beat up on Reddit for not being more careful in discerning this flaw in their testing. Several examples besides Techpowerup:
Im not thinking about thatYou're mistaken. You may be thinking back to right after launch when Techspot put out one of the first reviews. That review is what had me stunned because I thought 1000W PSUs would be required, but apparently not quite that much power is needed. Apparently their testing was tainted by a BIOS flaw (one seen in the past). They got beat up on Reddit for not being more careful in discerning this flaw in their testing. Several examples besides Techpowerup:
Im not thinking about that
I don't know whats up with the different result
Right but if you will recall at the start I was just saying that it uses an extreme amount of power. The 3090ti wasn’t a very well received card. It was incredibly power inefficient really.
Yeah I’m not going to read this long bitchy post. Obviously I’m comparing generations with similar branding. I know that doesn’t need to be explained to you but I’m sure you’ll play stupid and keep asking me more
What the hell are you talking about? Why not compare the 3070 Mobile to the 2070 Super Mobile? Why jump to comparing two cards which aren't analogous to these respective to their generations? More importantly, the reason this comment is stupid is because the 2080 Mobile and 3080 Ti Mobile-- unlike the 3070 Ti Mobile and the 4070 Mobile-- do not have identical specified TDPs. For the base standards, the 2080 Mobile carried a 150W TDP while the 3080 Ti Mobile carried a 115W TDP (the same as in the 3070 Ti and 4070). These are what I'm comparing above with the juxtaposed figures, not variants hitting the ceiling power threshold.
Moreover, the RTX 2000 Mobile series was on the 12nm die size, meaning the RTX 3000 Mobile series only shrunk the fabrication process 50%, not the 100% the RTX 4000 Mobile series did to improve over the 3000 series. This is why, on paper, one might expect better results from the 4070 once the nearly identical power ceilings are realized as with the variants discussed in the vid.
Third, what do you mean "close"? This isn't a controlled benchmark head-to-head like you'd get with desktop GPUs. The 4070 has a CPU with nearly double the total processing power driving it in his comparisons [2]. Did you even skim the video?
Finally, here is the comparison of the analogous GPUs from the previous two generations. As you can see, no, the 2070 Super Mobile (aka the 2070 Ti Mobile) definitely was NOT superior to the subsequent 3070 Mobile.
2070 Super Mobile vs. 3070 Mobile
Pixelation --------------> 88.3 GP/s vs. 124.8 GP/s (-30%)
Texturing ---------------> 220.8 GT/s vs. 249.6 GT/s (-12%)
VRAM speed ---------> 448.0 GB/s vs. 448.0 GB/s (0%)
Processing Power ---> 7.1 TFLOPs vs. 16.0 TFLOPS (-56%)
Ray-Tracing -----------> 55.2 RT/s vs. 62.4 RT/s (-12%)
VRAM Amount --------> 8GB vs. 8GB
That's why the title of his video carries, "Is this a joke?"
You said some dumb shit. Either have the good sense to slink away, or STFU, and spare me this snark.
