• PC Building V4 - "ok SSDs got cheap, now do RAM next"
    999 replies, posted
Yes, generally, but I wouldn't bet money on Nvidia following their classic release playbook because there are credible rumors that they may release the 2070 Ti at Turing's launch alongside the 2080 and 2070.
I keep forgetting what Kryonaut is, you're right.
The GPU I'm looking at getting has an 8-pin power connector, and AFAIK this machine's PSU only has a couple 6-pin connectors for PCIe. I'm guessing that using the 6-pin plug in the 8-pin socket wouldn't work? They still seem to be keyed correctly to fit eachother, which is a little confusing.
It's technically work, since the two extra pins on the PCIe 8P are basically just ground pins. But generally it's an indicator your PSU isn't spec'd to deliver the wattage an 8P can draw (150W vs 75W) - so It's advisable to avoid.
It might work but it's not really a good idea, best case scenario the GPU runs at lower clocks because it isn't getting enough power, worst case scenario it tries to pull it via the PCI-E connector and frys the MB. They have to be keyed like that as most PSUs come with a 6+2 connector so it works on both 6 and 8 pin power plugs. The OP has a list of good quality PSUs that won't explode on you randomly. Don't cheap out on the part that can take the whole system down with it when it breaks.
If that's the case, I guess an adapter like this would be a bad idea too? https://www.amazon.com/dp/B01HJBV50Q/ref=psdc_3236443011_t3_B0776PMH16
What PSU exactly do you have? How much current can it deliver on the 12V rails? That adapter could be a bad idea if the PSU manufacturer cheaped out on wires that aren't specced for larger currents, so they could overheat or even start burning up. The 2x 6-pin --> 8 pin is probably a safer option overall.
no actual information to add but I love your username c:
What kinda muddies things for me though, is that I've not only been told my current GPU is a 150w card, but that 150w cards with 6-pin are a thing thanks to up to 75w supplied by PCIe itself, combined with the 6-pin's 75w to provide 150 total, which is what the new GPU is supposedly rated at as well. I know bus powered GPUs exist, so PCIe has to be providing SOMETHING safely, I just don't know the specifics.
So I've been poking around with the idea of upgrading, as my CPU and GPU are starting to show their age (i52600 and GTX770), but the system performs nicely all things considering, but I guess my question is how much more of a difference in loading times is there between SATAIII SSDs and NVME SSDs? Because Windows 7 currently loads fairly quick (it takes less time from when the logo first appears to the desktop appearing than it does from the time you power it on to when Windows starts loading, mostly due to certain BIOS settings I can't disable for some stupid reason).
Manufacturer-provided TDP values are notorious for not being very exact or meaningful. It's meant to represent the amount of heat generated under a typical load scenario, but the peak power draw can often be even higher than that. So they probably put an 8 pin on it just to be safe and to provide some extra headroom for overclocking. If you were running a 570 up till now, then you should probably be fine with a 1070, judging by reviews they seem to use a similar amount of power.
If you can find the length of the wire, gauge and current you can push more power through it in Nvidia bios flashers. Perhaps AMD has something as well, me and my brother did it for his GPU just to see how high we could get it. No issues 1 year onward.
It's not a noticeable improvement at all.
I have a quick question related to RAM. Right now I have 16GB - two 8GB sticks - of this RAM I was thinking of buying some new RAM, but I also had the idea of instead just buying two more sticks of this same RAM since it's not that bad and I'm pretty sure for the same price as a new 16GB kit at say, a higher MHz like 3000, I could get two more 8GB sticks of what I already have. I checked the number of the sticks I have and matched them up online so I know I'd get the exact same sticks. This is my current motherboard and I'm not sure how I would find out if it would be better to get more of the same RAM, or buying different RAM altogether and staying at 16GB. I have two monitors and multi-task a lot so I figure bumping up my RAM even more to 32GB might help smooth things out a bit. Not sure if that would even help much though.
I own a lot of SSD's NMVE is barely perceptible in daily use. If you're working with programs that do a shit ton of IOPS or need to move large files fast maybe. But Even on bootup, the CPU is doing plenty of work for the drive to wait on.
With Nvidia's next gen imminent, I can only speculate that an update to Vega will follow it, since that's the game AMD has been playing lately
Vega still can go 12 or 7nm, that's all I'd expect from a refresh. Might let it get higher clocks due to better power., since historically Vega benefits from undervolting.
That's pretty much exactly what I meant, a small revision like that for a "Vega 2" just to stay in the news
Oh well, I'd say it's guaranteed then. That's all they got.
There are rumors that Navi could just be another GCN refinement. If true, then RIP RTG, we had some wonderful time together.
I wouldn't expect any consumer refreshes at all, and if we did it wouldn't be of Vega. Vega is simply expensive as hell for AMD to make, so just like fury it's going to have it's run, and then disappear into only fireproof/WX/Instinct products. Vega is getting a 7nm version, but it's more than a refresh, not only does a new node required different degin decisions, they're also going with 4 stacks of HBM, so that requires a thicker memory bus, not something exactly trivial for a refresh.
Somehow I feel like they'd put out some kind of "Vega X" super-enthusiast card just to swipe some publicity after Nvidia's announcement They wouldn't even have to make a lot of them, they'd just have to announce it, show off some specs and a demo, sell early Titan numbers, and it will have kept them relevant in people's minds long enough to do some real work
I mean I'd like it (I'd probably buy one); but I don't think it'd justify the driver validation and upkeep.
Since when has the Radeon division made sensible business decisions Aside from using freesync instead of inventing their own thing
Is the Seasonic M12-ii EVO worth it over the S12-ii ?
Only worth it if you care about ATX power being detachable.
Turns out the M12-ii EVO might actually wind up cheaper for me . Alright then!
If I do wind up going with a founders edition card, how much of a benefit would there be to cracking it open and re-pasting/re-padding it? Last I heard the FE cards don't use great TIM, which is a problem for a blower card.
Let me just get out my crystal ball here, I'll be back with thermal benchmarks in a jiffy
I mean for a 1070 FE. Apologies if it was a little unclear.
Sorry, you need to Log In to post a reply to this thread.