Jump to content

Alex Atkin UK

Member
  • Posts

    10,420
  • Joined

  • Last visited

Everything posted by Alex Atkin UK

  1. Someone has to pay for the server upkeep as unfortunately peer to peer became too unreliable due to various routers and NAT solutions breaking it. PCs don't need to be huge:
  2. Artificial light is orders of magnitude duller than daylight. Our eyes are designed for daylight, so of course forcing them to work in much duller conditions is not going to be great for them. I personally feel I've had much less eye strain as monitors have gotten brighter, the only problem is trying to then match the room lighting as you're basically then needing to light the room like an office. It makes far more sense to try to bring up room lighting to match the monitor than to dull everything down to the point your eyes are actually working harder to focus on a dull screen. Sure its possibly a good idea to go darker as it gets later in the day to match the day/night cycle, but for optimum focus brighter is better IMO. I know a few people who claim they don't like brightly lit rooms and they all suffer migraines. I suspect rather than them not liking bright because of migraines, I think they got it backwards as they seem to suffer migraines more when they have the rooms dimly lit, but they aren't making that connection. Its also kinda bonkers claiming blue light is bad, when the aim with monitors is to get as close to natural lighting as possible. So its more the case that a cheap poorly calibrated monitor will cause more problems than a good one. At least that's my opinion. Worth noting, I find my eyes overly sensitive to daylight and I think that is purely because I spend more time in artificial light - its the fact my eyes are used to a duller environment that causes problems. I actually feel physically better in brighter lit room, once I've adjusted.
  3. That to me suggests the motherboard is using a rechargeable CMOS battery and its not getting charged enough to last overnight. It doesn't necessarily need replacing, just needs to be left with the PSU powered on long enough to fully charge it.
  4. Everything I can find suggests the 4700U can only support 4K 30Hz or 1440p at 75Hz as it only supports HDMI 1.4. So I think its glitching here in some way.
  5. Some boards can behave like this from a completely powered down state. Have you tried NOT turning the PSU off by the switch to see if this prevents the problem?
  6. If you're not bothered about low resolution and low frame rates, then sure. But overall the Series S is not a great experience if you're used to PC.
  7. All of which happens in 100% of power switches (though you might not hear it as most the click is so loud and fast you wont hear the spark), which is why you absolutely do not do the last thing.
  8. Another problem is motherboards used to use single-use Lithium cells for the CMOS which lasted up to 10 years. A lot of modern ones assume you never turn the PSU off so use rechargeable with varying capacity, that can drain relatively quickly and may in fact wear out in a few years if you're using up charge cycles every day. Are you not pushing it hard enough? It should happen so briefly the noise would be hidden by the click of the button.
  9. I was thankful when controller support in PC games became standard. Not been able to play with keyboard and mouse for years due to immediate hand cramps in my left hand when I do.
  10. Arguably it should be, but you're right in that it seems a lot of manufacturers really cheap out on it assuming people wont be. I've definitely had one or two PSUs where the switch broke despite me very rarely turning it off that way.
  11. Generally speaking when people recommend against daisy-chaining switches they mean each switch should be connected to the main switch rather than daisy-chaining them off each other. Having a main core switch and running other switches off that is perfectly normal and the right way to do it, if its impractical to just have a single big switch and run all devices back to it. You just have to be aware that all devices on each extra switch is sharing a link back to the main switch, so if you are ever moving files between devices, you ideally want those to be on the same switch so its not causing a bottleneck that may impact the bandwidth of the other devices on that switch. The way you generally limit that being an issue is having your main switch be faster (2.5Gbit or 10Gbit) and each extra switch having 2.5Gbit or 10Gbit uplink back to that main switch. But that's mostly an issue if you have a NAS on the main switch and don't want accessing that from the other switches to slow down the Internet on other devices on those switches.
  12. That's not a good way to look at it, as wear levelling should be trying to avoid any cell getting more wear than another so once they start failing and using spares, you'll want to replace it ASAP as odds are the whole NAND is close to failure. I just wouldn't buy used anyway though. Its problematic enough with fakes on the market, there's just no way to know what you're going to get.
  13. There's no such thing as overkill, spare RAM is used to cache data which can speed up reloading assets in games that were previously loaded then discarded from RAM as they weren't used in the current scene. This can be particularly useful for large open world games with fast travel. Do you need it? No. Is it wasted to have it? Not at all. It really comes down to if the money saved by only buying as much as you need is enough to upgrade something else that would result in a bigger performance difference.
  14. Its certainly interesting, given its not even a fair fight as they are limiting the 7500F with slower RAM than anyone should be getting for that CPU, in order to make it as Apples to Apples as possible. But they do not mention RAM timings which can play a big part in offsetting slower RAM. They also do not mention if using PBO and given the 7500F is an ASUS board, they are also prone to use more aggressive defaults. So it could be the 5800X3D is running to spec and the 7500F is not. I'd also argue the 0.1% lows are important, as if the 1% lows are so high then a sudden HUGE drop even 0.1% of the time is going to be extremely annoying. I'd prefer a lower average fps over huge drops 0.1% of the time any day.
  15. Not quite any, I got an Intel Datacentre Gen 3 drive cheap and it completely shuts down under load without a heatsink, and I'm only running it on x1 lane for IOPS rather than raw speed. I guess they're designed for 1U chassis with forced airflow, though just putting a RPi heatsink on the controller completely fixed it (AFAIK NAND doesn't care about running hot, it actually helps write retention). Its true that most consumer drives work okay, but they might slow down. I have a Crucial P1 and P3 which get mighty toasty even with the motherboard provided heat spreaders, though they don't really count as a heatsink IMO as they do a poor job due to being completely flat.
  16. I don't know those specifically but IMO they're all extremely risky as many key sites are marketplaces allowing anyone to sell keys, which may be stolen. The only one I'm fairly sure of are CDKeys as I believe they use regional pricing to obtain legitimate keys themselves, rather than the others that lets anyone sell keys.
  17. Its still hitting 112C on hot spot though which is mighty concerning to me, as its only likely to get worse over time as the thermal pads dry out. The 4070 Ti is not a power hungry card, the hot spot should not be this high IMO although I can't check mine as its on Linux.
  18. Would have been nice of them to tell me on the device itself when it said "nope", its too late now.
  19. It would be very unlikely for there to be no WiFi 7 Revision 2 cards, though maybe it will be a silent revision rather than a BE210. Have you seen how many different WiFi 4 and WiFi 5 model cards Intel has made? Though granted a lot of those were different MIMO configurations They already have two 2.5Gbit ethernet adapter models released in the last few years, the first one which had at least three revisions due to bugs. We don't even know for 100% certain that the BE200 will fully meet the final WiFi 7 specification. Historically WiFi vendors push v1 to market early and then release better performing versions later. My own experience was the AX210 performed better on WiFi 5 than the AX200 did for example, though both were comparable on WiFi 6.
  20. Intake is best for cooling the CPU, Exhaust is best for cooling the GPU. It all depends which part you're trying to optimise the cooling for and if you have any other intakes that can direct air not already heated up by the AIO towards the GPU. For example I managed to squeeze a 120mm fan under my 240mm AIO front intake, so it added some cooler air towards the GPU.
  21. That's an awkward size, tower coolers tend to be bigger than that so we end up with things like the Thermalright AXP90-X53 Full Low Profile which are a lot smaller.
  22. Its not recommended but there's a reasonable chance it will work as long as the branches are not fully blocking line of sight. It just might not be 100% stable in high winds or rain/snow. I have a P2P WiFi using a couple of LiteBeam 5AC 23 over a short range with a tree trunk in the way, some branches and the other side mounted indoors behind a thick wall (I seem to pick it up via reflections through the windows). Its a fairly stable 80-100Mbit. Unfortunately the only way to know for sure is to try it.
  23. You can simply change the SSID on the second router to be the same as the first, then any WiFi devices will just connect to the strongest signal. It wont be as seamless as a mesh system, but that may require replacing both routers or at least not using the WiFi on either and having two mesh devices instead for the WiFi. But its probably not worth replacing the WiFi at all yet as most devices are going to top out around Gigabit anyway. For now I'd focus on getting a 10G switch with at least one SFP+ port and put it BEFORE the second router. That way you can have the media conversion done in the switch itself with a SFP+ adapter. Is the fibre plugged directly into a SFP+ adapter in the main router or is it copper going into a media converter at that end?
  24. That does seem rather high and my guess would be its the VRM which will likely shut-off at around 125C, which could explain the black screens. As to why, the only thing I could think is bad thermal pad installation during manufacturing? Does it happen with the PC side panel off? If so, I'd RMA the card as defective.
  25. That would suggest DNS servers are not being set by the routers DHCP on whatever device you are pinging from. The router may also have its own DNS forwarder and be setting the gateway (itself) as the DNS for the clients, which wont work if the router doesn't have any DNS servers defined. Nice of them to finally tell you. Why they don't define this on the endpoint device instead, I guess as technically you could have multiple services down the same fibre and the VLAN would be how you define which one to use?
×