kksnowbear

FWIW, even though I'm not in the UK and have not actually done business with them, I did - briefly - look over the PCSpecialist site as mentioned above. It does appear that they allow you to pick, individually and specifically, from named components (and the brands are well-known). So it does seem like a decent alternative, for sure. And endorsements from fellow simmers here, that's a plus. That said, it should also be mentioned that, of course, the final cost will matter...for example, it doesn't matter how good PCSpecialist is, if they charge twice as much for the desired components, probably still not the best approach. To be accurate, I did *not* look at pricing at all. Secondly, even though they do list components by make/model, they do still have better and worse brands and models among the choices. For example, while they do have GPUs from Asus, the only other manufacturer they specify is PNY; the rest are back to 'generic' specs without a manufacturer or model. In power supplies, it appears they only have Corsair units...nothing wrong with that, necessarily, Corsair makes some great PSUs. But there are none of the other good manufacturers (Antec, eVGA, Seasonic, etc...) which you might get better pricing on...so you are somewhat limited. Don't get me wrong - this does appear to be worth considering, just same as above, you have to pay attention to what you're getting - and at what price. The Scan site seemed a good deal more expensive, but I haven't gone back for another look yet. A business - almost any business - makes (or loses) money on inventory. So if PCSpecialist limits the inventory choices in order to control costs, nothing wrong with it at all. It is *definitely* a plus that they specify which brands/models and let you pick...just gotta watch price One small thing I saw that I didn't like...PCSpecialist has Arctic MX4 thermal compound as an option, for an additional 9 GBP...that's a lot. I use Artic MX4 as 'standard' on every build I do, at no extra cost. So that does make me wonder about what the final build cost might tally up to (without actually looking at it, mind you).

Looking over a few examples at the link you posted, I found the problem with the link is essentially the same as with the list above: They are builders, and they're not disclosing many of the key specifics. For example: No motherboard make/model listed. No RAM make/model; RAM speeds are listed, but no CAS levels. No power supply make/model - and no indication of whether the units are 80+ rated. No SSD mfr/model. No GPU mfr/model... All this should start to sound very familiar lol As I mentioned previously, this is what they do. They cite specs that are (very) general in nature, leaving all sorts of places where they can cut costs by putting in lesser brands, lesser models etc. And again, the reason all these details matter is because, between them, they could alter the cost of a machine several hundreds of (insert currency here). (Some of the high-end systems on that Scan site are 7000 GBP, and I'm not even sure how taxes would apply...they start to seem *way* overpriced...more on that in a separate post...) Note there's absolutely nothing wrong with having someone else do a build for you, especially if you're not comfortable doing it yourself. I do it for people all the time, and if it weren't for those who don't want to do it themselves, I'd be out of a job lol But paying someone else to build it does *not* automatically mean you shouldn't be given details of the parts they're using. This is the key. In fact, what I sometimes do is take parts a customer has chosen for themselves (typically with advice, mind you) and then charge a fee to assemble/commission/overclock etc. This removes any doubt about 'mark-up' on the parts themselves, and the client gets to pick/pay what they want (again, with advice to make sure it's compatible). TBH it usually results in a more expensive system overall, but this approach offers complete transparency about cost of parts and labor. Some people just prefer that; whereas some prefer to save a little by letting me spec parts - but even then, I always tell them exactly what they're getting. Let me try this, so you can see what I mean. When I give a quote to a customer, here's an example of what it might look like: Motherboard: MSI X570-A Pro CPU: AMD Ryzen 7 5800X3D GPU: MSI GamingZ Trio Radeon RX 6900XT RAM: 4x8G XPG Spectrix D50 DDR4-3600 CAS18 RGB PSU: eVGA 850W 80+ Gold CPU cooler: ID Cooling 240mm AIO liquid RGB Boot drive: Samsung 980 Pro 250G NVMe SSD OS: Win10 Pro x64 Storage: Samsung 980 Pro 2TB NVMe SSD Case: Antec DF600 Flux, Mid-Tower, Tempered Glass Side Panel (Mind you this is only an example, intended strictly to illustrate my point about the details in a component listing/quote. It is *not* intended as a list of anything you should be considering for your current build, nor to solicit opinions about the stuff I listed - I'm not quoting anything to anyone here, just showing how the list itself should look). Notice that, for each item, there's a manufacturer and model specified, as well as 'secondary details' (like CAS level for RAM, or 80+ rating for power supplies). It doesn't matter (yet) if you don't recognize the make/model or other details in the list, but they need to at least be present so you know specifically what you're considering. Unfortunately, since (as with most all the pre-built vendors) they don't give details, it's really not possible to say whether any of the Scan systems is a good value for the price. You could always throw money at the most expensive setup they offer - many do this, I regret having to say - but I seriously recommend not doing that. You might actually be better off dealing with a local shop/custom build - but it's just going to depend on who's willing to give you the details as to which is a better deal. One final thought here: I'm not bad-mouthing the pre-builts (lol well not entirely...) If someone doesn't mind the difference in brands/models and paying someone else to build it, those can be a reasonable option. Some people just prefer to find the absolute lowest cost possible, and have no desire to build a system. What I'm saying is that, in any case, I think it's important to know what you're paying for, regardless of cost or who builds it. Hopefully all this makes sense.

Also on the connection question: What you could try is using the wireless at first, if it proves to be problematic, go another way. You should only require one of the three from the house to the outbuilding: Wired, wireless, or power line adapter. Try them in order of least costly (wireless) to most - including or skipping the power line adapters as you prefer. going instead directly to wired if that makes sense. NOTE: This will mean your PC has to support wireless, so it might be worth considering to get a motherboard that has wireless on it. Alternatively, small adapter cards are cheap and readily available. I would suggest avoiding USB interfaces/devices whenever possible.

On the whole connection thing, that's going to be a separate project I think...but I will say I think you're better off by far with a wired solution (*not* over power lines) if at all possible. Those power line adapters can be finicky, and wireless often leads to lag due to latency in the wireless radio signal.

I would seriously consider taking Qcumber up on his very kind offer. The experience gained could be priceless to your venture...and he makes two very good points about the GPUs. One of my "selling points" to potential buyers is that I will offer to let them sit down and "test drive" a proposed system, even for a couple hours if they want, so they can see how it performs as a system. I feel this helps distinguish me from a lot of other sellers, online or brick and mortar. If you have a chance to see a setup like his/close to what you're considering, I'd jump on it. If he's the sort to have a drink time to time, I think I'd offer to bring him a little something as a gift, too

Well... You can turn settings down - but there would be a limit to how much detail you're willing to sacrifice. You can turn settings up - but then (of course) frame rate will vary in inverse proportion...and there will also be a limit (in FPS) you're willing to tolerate. Trade-off, same as it ever was "Good" frame rate means different things in different circumstances...some people are willing to accept lower frame rates simply because they aren't interested in paying for the (stupid expensive) high-end GPUs. (As a 4090 owner, I gotta say, they're right!!! hahaha) Some people already *have* stupid expensive GPUs and still would welcome higher frame rates lol...(this is probably the arrangement that leads to a 4000GBP price tag) So, what is 'good' for you? 45...75...90...? I think the comment(s) above suggesting you get the best GPU you can afford are very good advice. VR is exceptionally demanding of graphics performance, and it seems like only the more recent GPUs are getting people up to the kind of performance that most seem to want. Sucky part is that these high-end GPUs are usually sinfully expensive The AMD AM5 platform is a good choice, because you can hopefully get reasonable performance now, and then later on (assuming a decent motherboard) have room to upgrade - likely that any AM5 board will accommodate processors all the up to the 7950X3D, but other features on the board will vary. AMD has committed to future socket compatibility as well, so you should be set for quite some time - your budget will dictate whether to invest now in a board that supports PCIe Gen5 (though I rather suspect this is one area where that quote you have includes a lower-end spec unit to save cost)...you can consider that further once you see the list's details. Incidentally - for my own edification, if you please - can you comment on the taxes I mentioned? For example, of that 1400 quote, how much is actual equipment cost and how much is because of taxes? I'm just wondering how much difference it's going to make if we were to try comparing prices I might see here to what you'll be looking at there. I don't think a currency converter takes that into account. As an example, a $1700 retail system "all in" (taxes included) would mean I paid essentially 1603 for the actual hardware and 97 in tax. I'm not knowledgeable as to how your taxes would work out exactly, but if you pay considerably more of the 1400 for taxes, it leaves proportionally less for the actual hardware - which obviously affects the component options. SO perhaps you'd be good enough to elaborate on that. As far as the list goes, Tuesday it is What I might honestly expect is something like this: Lower-tier AM5 motherboard, ASRock brand, probably B650 chipset (lowest of the four types). ASRock is very common in 'pre-built' systems, and it's among the areas they save cost. Off-brand PSU (not sure of brands available there in the UK but online sources can usually tell what's what). Another area where prebuilts usually try to save cost. Lower-tier GPU manufacturer (i.e. not Asus, MSI, Gigabyte - though again I'm not informed as to brands where you are). Some examples I put at the lower end (just my own opinion) might be ASRock, PNY, Zotac, or Palit. Off-brand M.2/NVMe drive (i.e., not Samsung, WD, etc) Possibly lower-end RAM, probably lower speed and higher CAS rate. What the prebuilts typically do is cite popular, well-known component details like "32G RAM" (without saying who made it, the speed or CAS level); a "4070" GPU (no manufacturer name or model); an "850W Gold" PSU (no mfr name/model number). Even among high-end manufacturers, they might make two or three different models of 4090 GPU, or several models of 850W PSUs - and the features of these different levels of components can vary a lot. It's not a lost point that the list in the quote not only doesn't specify a brand or model of the motherboard, but also doesn't identify which chipset - and that can vary the cost (and features) by hundreds of dollars. Seriously. Here in the states, Alienware is seriously among the worst of these pre-built brands. It's a well-known name. But if I look at those systems from a background of 40 years' experience and formal training, they cut corners as I've outlined, yet are looked upon by uninformed buyers as top-end systems. IMO, they aren't, really. I'm not saying it's immoral or illegal mind you, but what these places do is assume that anyone who's paying them to build it really doesn't know that much about specs etc, and that leaves a lot of room for cutting cost. I can honestly say that, when I've compared offers like those by looking at the details, it's common to find the areas I described above are where they're making money. Anyone who's preference is to have someone build a system for them (and there's nothing wrong with that) needs to pay attention to these details, and insist on getting specifics about the components being used. Anyhow, we shall see I suppose Happy "Summer Bank Holiday"

Yeah, see that should illustrate what a huge difference can be made depending on exact hardware etc... But I would agree 1400 is fairly conservative (without actually studying the matter, and no real details, mind). That said, however, 1400 GBP is around 1700 USD if I'm looking at it correctly (though you guys pay a lot in taxes as I understand it...my sales tax here is 6%) . 1700usd is not outrageous for a good system but it does depend a lot on the specifics. If you're really interested in top notch performance you may wind up a bit lot closer to 4000 lol But for sure, let's see what the list says.

Oh, ok...in that case lol

I'd love to, honestly.... However, considering that I'm on the "other side of the pond", I'm afraid it wouldn't be very practical

Not so sure I would agree with that TBH. Even some models of 4090 from the same big three recommend 850W. I have a Asus TUF 4090, and the most I've measured it drawing is ~650W and that's with a fairly heavy load. Asus recommends 850W for this model. I have a 1300W PSU, but there are other reasons for that beyond 'requirements' or even 'recommendations'. However, there are reliable sources that indicate 4090s can 'spike' up to 700W, and very few meters of any type are going to "catch" this, especially if they're reading the line side of the PSU - which is practically *everyone*, unless they're doing professional work and have the (very expensive) equipment to measure with. Certainly not 'wall-wart' Kill-A-Watt type meters, nor the software (or hardware display) on any battery backup/UPS/SPS or surge suppressor. I don't know factually that this (the spiking) doesn't apply to other 40-series GPUs besides the 4090, but the card specs don't seem to indicate it would apply. For example, even a 4080 has a TDP of 320w which is fully 130W *less* than a 4090, so it's hard to imagine that unit, even with 'spiking', drawing more than 500W (meaning an 850W PSU would almost certainly be adequate, provided it's a high-quality unit. (Admittedly, I haven't studied this particularly and do not have the hardware, so I'm basing my perspective on general research and available data). A 4070 has a TDP of 200W (and a 4060Ti just 160W) so a 1000W PSU would *definitely* be gross overkill, and thus unnecessarily expensive (unless the concern is more about efficiency, but this whole discussion changes substantially in that case).

I might be able to come up with a figure...but it wouldn't be comparable to what you' d expect to pay for two reasons; one I build everything myself, and two I do this professionally so I don't usually pay the retail "going rate" for hardware. FWIW The price you listed above didn't sound horrible, considering...but you have to be careful when ordering 'pre-built' machines as there are always ways in the BoM that they're shaving off costs. For example, in your post above, you list an 850W Gold PSU, but what model and brand? Same goes for the motherboard, RAM, GPU and M.2/NVMe drive. Those things combined could make a *huge* difference in the final cost of a system, if they're using off-brand or bottom tier hardware. Also, FWIW, I would agree with the comment above that you should probably skip the (more costly) over-clocking model motherboards, stick with reasonable brands but not top-end models. Most of the features you really need are available in mid-range boards which can shave $100 off the price of a board alone. I personally build more Asus systems than anything, but there are plenty of MSI and Gigabyte fans. I think I'd go with Asus or MSI - my own personal preference, but based on a decades of first-hand experience. I'm not sure about the 64G RAM vs 32 question TBH ...it does seem a lot of people say DCS eats up 32G rather quickly (depending, of course)...so if you can afford the extra cost, maybe consider 64G.

For reference, on the PSU post I made above: Gigabyte https://www.gigabyte.com/Graphics-Card/GV-N4090GAMING-OC-24GD-rev-10-11/sp#sp Asus https://rog.asus.com/graphics-cards/graphics-cards/rog-strix/rog-strix-rtx4090-o24g-gaming-model/spec/ MSI https://www.msi.com/Graphics-Card/GeForce-RTX-4090-SUPRIM-24G/Specification (As the fourth in the 'big four' GPU manufacturers - eVGA - dropped out of the business after their 30-series, that pretty much means the biggest three in the world).

If you change to a 4090, you'll require a 1000w power supply - at least per 3 of the biggest AIB manufacturers' recommendations for systems running their top-tier 4090 GPUs: Some have attempted to insist an 850 is enough, but again that's not what the mfrs say, and it also doesn't stand up to educated academic scrutiny either. Foolish to spend that kind of $ on a high end system and then run the risk of starving it for power by trying to save $50.

I'd suggest a gigantic grain of salt if using that website... First of all, the data it uses for DCS appears to be very dated, citing (for example) the game requirements as shown here: A GTX560...really? Also, if you look, you can see that the data it is relying upon to make these calculations isn't exactly scientific. Anyone can basically use this section of the site to input values, and those values don't even have to be actual, first-hand experience, nor vetted or verified in any way. The inputs shown below all indicate using "anticipated" values is adequate, meaning it's not at all constrained even to reality, only what somebody thinks they'd get with said hardware. This is hardly what could be called reliable information by any stretch of the imagination. Don't get me wrong, it happens I am familiar with the site, long before now, and I've used it as a reference many times over the years... However, those references are again always take with a huge grain of salt, and always with very careful consideration of other, more proven reliable information.

Ummm... Both Intel and AMD have made CPUs with (far) more than 18 cores for quite some time now (as in years). And that's "full on" cores (if you must). Problem isn't core count. If it were, there have been many CPUs that could easily overcome the issue, having core counts of 56, 64 (and beyond). Problem is cores aren't utilized effectively at all. I do fully understand the difficulties involved - but that doesn't change the fact that lacking cores isn't the problem, therefore more cores isn't the answer. Software isn't using the hardware effectively as it is. More hardware (cores) won't help.

Price lowered - now asking $250 plus actual shipping (US48 only)

House have coax for television? If so, MoCA might be an option. It's wired, so lag isn't usually an issue compared to just about any kind of wireless (including mesh networks). HTH

Because of my ongoing technical work, I have a monitor I'm interested in selling. It's a 27" Acer Nitro XV3 4k unit (model XV273K). It has an IPS panel, 1ms response time, Gsync compatible and supports 144Hz refresh rates using two DP cables (single cable = 120Hz; HDMI is 2.0/60Hz refresh). There are no scratches, blemishes or dead pixels whatsoever on the screen. I've never noticed any cosmetic issues on the chassis or base. It is in excellent shape overall. I'm not a display spec fanatic, but I'd say the picture is fantastic. Others who have visited my shop and seen it first hand have commented about how good the display looks. This is an excellent opportunity for someone who wants to run 4k and has a GPU that can perform at the level. For me, Nvidia 2080Ti/3080Ti3090/3090Ti/4090 all have done great. Even a 2070 Super and 1080Ti can manage reasonable frame rates with appropriate settings. It works as expected but the AMD Freesync implementation seems to be buggy, which is why I'm replacing it (my work requires being reasonably able to test all functionality like GSync/Freesync and higher refresh rates). It will sleep as configured, but doesn't wake as it should, and this is only when FreeSync is used with an AMD GPU. The Freesync functionality itself, as well as FSR and RSR, work as expected. Just the stupid sleep thing. I am certain it's not 'bad hardware'; rather, a design/implementation issue. Nvidia GPU Gsync works flawlessly - so I am suggesting that it is a good choice for someone with an Nvidia GPU (or someone who doesn't intend to use Freesync for whatever reason). Also, since it's HDMI version 2.0, it only supports 60Hz refresh over HDMI and I really need to be able to test at least 120 (which requires HDMI 2.1) for higher-end cards I sometimes work with. This wasn't an issue honestly until more recently, now that a fair number of cards can actually perform at that level. So even if it weren't for the Freesync sleep thing, I decided to replace it so I can test a broader range of hardware on my "4k bench". As of this writing, this monitor is selling New on Amazon for 551.65 (see pic, although to be fair it's a third party and not the best price). From what I can see on the Bay they're going for over $350 before shipping and taxes (and the shipping is being listed at well over $50, plus they charge taxes on the shipping, so you'd be looking at over $425 Used). I paid over $500 for it, in April last year. It's only been used for bench testing in my shop (~3-5 hours/week on average), not "full-time" usage by a considerable margin. I would like to get $300 $250 plus reasonable shipping costs (should be able to come in <$50 on that). I am not sure if I have the original box, but I will definitely arrange for a suitable shipping package one way or another. Please PM me if interested. I'd be happy to answer any questions, and I will try to get some pics up soon. Thank you.

Beautiful I've seen many people mount radiators in the front of a case, and it's not a good choice (typically for a few reasons). Mostly it's because they don't have a case to begin with that is adequate to properly mounting a 360 radiator so it will exhaust heat. So they wind up mounting it in front, as an intake - which as I'm sure you know, according to the Law of Conservation of Energy, just makes everything inside warmer. Really bad idea. Much better to just get/use a proper case that is designed to accommodate big radiator(s) in optimal mounting configuration - as you obviously had the foresight to do Never ceases to amaze me, people who obviously have money to buy very high-end hardware but then insist on cheap or under-rated power supplies, cases that aren't properly sized or ventilated, and/or components that are going to force poor build decisions (like dumping heat from a radiator inside the case). Personally I believe if someone has money for really good high end components, then they should match the balance of a system build with other parts that are appropriate to the core components (GPU CPU, motherboard, RAM). Why spend $2500+ on these and then use a small case or PSU just to save $50...??? On the PSU: I did a little quick looking around and, as I thought may be the case, it's a brand that's just not common in the states...but as you pointed out, it's made by the same Chinese outfit that makes other brands which are common here. This is actually very common, with almost any brand of PSU these days. Good show! I am sure you'll be very happy with the new build

Looks like it's a great setup which should perform well for quite some time to come And it's also very good to see that you wisely decided to go with the manufacturer's recommended 1000W PSU. (Admittedly I don't recognize the brand, but since you're not in the USA that doesn't necessarily mean anything). I do sincerely hope you mounted the 360 radiator in the top, oriented to exhaust heated air from the case - yes?

We agree completely Sorry if it seemed I was arguing - I guess in a way it *is* an "argument" - but not at all negative I believe you are perfectly correct in that RAM particularly will have less of an overall impact than other major components like CPU/GPU. The thing is, precisely because of your point, the impact of RAM on performance is very difficult to test/quantify in a system's overall game performance; it seems to get 'buried' within the whole. So the benchmarks, especially those specific to memory testing, can help 'bring out' the difference in performance that is due to RAM itself, instead of it being lost in all the other factors.

Of course, you are correct about the difference in benchmarks and 'real-world' performance, especially in any specific application/game. I agree completely. However, there are many instances where benchmarking is useful - at least two of which are: 1. A case where someone doesn't already have the hardware in question, and is looking to determine how best their budget is applied. In this case, you might want to know how a given upgrade/hardware purchase will affect performance generally, as a measure of how it might affect performance specific to one app/game. An example might be "RAM kit X seems to increase performance by (measurement)...but for 10% less cost, RAM kit Y seems to provide the same performance increase". 2. Where comparing system A to system B, either when they are same/similar hardware or even different, but looking at effect of exact same hardware upgrade on each/both. As you say, it won't necessarily tell you what any given specific app will do on either machine being compared...but it will help establish the overall impact of the same change, comparatively, to both systems. One example might be "Changing a 3090 to a 4090 on a VR system is worth the cost, but may be less of a 'value' if only running 1440 on a conventional monitor". Naturally this doesn't apply to everyone, because many (most?) people don't have more than one setup to test anyway...but it does apply when considering/comparing one's setup to others, or if a person has access to more than one set of hardware. I run a shop and have 20-30 setups at any one point that I can try various things with, and this is very helpful to me in making changes and/or recommendations to my own system, and others' (which I do a lot of). So benchmarks in general, while not necessarily translating directly to 'real world' performance, are far from being meaningless. If understood and used properly, they can be a very effective tool in an overall approach to system design/building/maintenance/upgrading. And they're not just academic, depending on the circumstances. Because I do a lot of this type work, I keep very extensive records (>1,800 test cases), all involving synthetic benchmark scores. Without this data, I'd have much less in the way of tools to compare/consider options, and effectively advise others as to their options.

Out of curiosity, what model is the Kill-A-Watt unit you used originally? (And brand, if it's not actually a Kill-A-Watt brand, there are tons of knock-offs.) Also, could you tell me the length of time that the CyberPunk benchmark takes to run (is it always the same amount of time)?

I said earlier (still there, go look) that it was recommended 1000w, required 850. In spite of your obvious internet prowess lol, you're still off base...because using the minimum required PSU is always a good idea ("not") - especially in a system where someone drops a load of cash to put in a 4090 to begin with. Two other AIB vendors with models of the 4090 that recommend 1000w PSU: - Asus Strix OC edition - Gigabyte 4090 Gaming OC 24G Since eVGA is no longer in the GPU game, that pretty much covers the 'big three', here in the States anyhow. You have no legitimate, confirmed information from an official source that proves which system those manufacturers were considering when they made their recommendations... ... do you? So, there's no way for anyone here (including you) to say what they took into account. And they don't specify "Well, if you have a 7800x, then it's really only 850" And again, the recommendation is 1000w for (now three) GPU manufacturers' 4090 cards. You're not arguing with me, bud - looks like you're arguing with them.

The only point I made about 230 vs 120 is that it's more efficient, thus being among the reasons you two will see different power consumption. Note well: I said "among the reasons". I said it is becuase...it is. Fact: 220 line is more efficient than 120. And it most assuredly *is* on point once you turned the discussion to "why is your load different from mine if they're similar systems" (even though you did this while posting readings from a different device, measuring by a different means on a system that is...different.) I cited a number of reasons that apply to your question. Simple: They're different. And I went on to say it's important to identify the reasons for the differences... ...while you keep arguing with a manufacturer's specs for their own hardware, simply because it doesn't agree with your assertion that 850w is enough. Not according to MSI it's not. But what does MSI know? They only design, manufacture, and sell *thousands* (millions??) of these things.