Jump to content
Keatah

A consultant's advice is better than a gamer's advice..

Recommended Posts

When it comes to building systems, a consultant's advice is better than a gamer's advice because their commentary and suggestions, tips & tricks, result in a better, more practical, stable, and capable build. And their philosophy tends to apply to both state-of-the-art and vintage hardware equally. Time agnostic.

 

The systems I built that have a "consultant techniques" incorporated tended to last longer with less downtime caused by tweaking and configuring and just generalized messing with settings.

 

The gamer's philosophy produced machines that sometimes became unstable, were certainly not cost effective (think $1,000+ graphics cards), and required tweaking and settings changes all too frequently - oftentimes on a per-game basis.

 

Without complaining further, the gamer's builds turned out to be less usefull all around compared to a consultant's build. The latter could go back and forth between productivity, hobby, and gaming very smoothly. What the gamer build was generally stuck on gaming and didn't adapt well to general usage.

 

---

 

I see that professional advice comes without all the blaring ?music? and noise and circus antics. But when watching a gamer's vid, it's full of ridiculous sound effects, jerky camera action, high-on-energy-drinks-scripting. And mostly full of this "back-n-forth" indecisive tedium. Think some kid screaming for subscribers and likes and exhaustively detailing FPS benchmarks on a lineup of 10 graphics boards that vary in price by $3.00 each. Only one will be a winner and the other 9 are complete utter garbage not worth considering.

  • Like 1

Share this post


Link to post
Share on other sites
Posted (edited)

Not necessarily paying one to build your machine. But just following their philosophy.

Edited by Keatah

Share this post


Link to post
Share on other sites

Well in that case I'm a consultant myself. My fees are reasonable but my clients have to put up with my sarcasm and generally pissy attitude.

 

  • Like 1

Share this post


Link to post
Share on other sites

I think there is a certain level of truth to this. One of the biggest problems I see in much of the vintage gaming and computing community, where it concerns people designing and selling projects of their own, is overestimating the market. They will have a good idea, and try to bring it to fruition, only to discover no one buys it. Why? Well, it could be due to price point...it simply isn't worth it's cost. Or it could be due to practicality. Possibly it is a redundant product, where another better or cheaper solution exists. Or maybe it simply fills a niche that the vast majority of gamers simply do not fall into.

 

Whatever the case, as a gamer, it's easy for us to put the blinders on when we have something in mind that sounds like a good idea. Its much more difficult to take the Warren Buffett approach, and look at the product for it's potential.

  • Like 1

Share this post


Link to post
Share on other sites

When it comes to building systems, a consultant's advice is better than a gamer's advice because their commentary and suggestions, tips & tricks, result in a better, more practical, stable, and capable build. And their philosophy tends to apply to both state-of-the-art and vintage hardware equally. Time agnostic.

 

The systems I built that have a "consultant techniques" incorporated tended to last longer with less downtime caused by tweaking and configuring and just generalized messing with settings.

 

The gamer's philosophy produced machines that sometimes became unstable, were certainly not cost effective (think $1,000+ graphics cards), and required tweaking and settings changes all too frequently - oftentimes on a per-game basis.

 

Without complaining further, the gamer's builds turned out to be less usefull all around compared to a consultant's build. The latter could go back and forth between productivity, hobby, and gaming very smoothly. What the gamer build was generally stuck on gaming and didn't adapt well to general usage.

 

---

 

I see that professional advice comes without all the blaring ?music? and noise and circus antics. But when watching a gamer's vid, it's full of ridiculous sound effects, jerky camera action, high-on-energy-drinks-scripting. And mostly full of this "back-n-forth" indecisive tedium. Think some kid screaming for subscribers and likes and exhaustively detailing FPS benchmarks on a lineup of 10 graphics boards that vary in price by $3.00 each. Only one will be a winner and the other 9 are complete utter garbage not worth considering.

 

Putting on the consultant voice for a moment, I'll add one to the pile of perplexing hardware decisions that appear in a surprising number of gamer-built PCs: RAID cards and ridiculous amounts of storage.

 

If you fundamentally do not understand the purpose of different RAID levels, how and when to use them, and why, just use a decently-quick SSD or two in a JBOD arrangement and call it good.

 

If the goal is to have more than, say, 4TB of online storage in a single machine, fine - but your desktop or gaming PC is not the machine that should also be acting as your home NAS.

 

Oh, and 10Gbit network interfaces have virtually zero benefit to online gaming at this time unless your Internet connection is higher than 1Gbit. Save your money and get a decent Intel 1Gbit adapter.

 

Seriously, though: wasting money in the pursuit of trying to build pseudo-enterprise-grade technologies into what's essentially a desktop PC is pointless. Spend the money where it counts, not where it's the equivalent of a +5hp sticker.

Share this post


Link to post
Share on other sites
Posted (edited)

The 10Gb ethernet is for lan parties, still overkill but you can copy files super fast so you can start gaming sooner. Raid setups are for zero downtime because a gamer can't take that risk. Raid still doesn't save you from backing up your data. Back in the 1990s a $600 gaming card was no match for a $6000 professional 3D graphics card used in 3D cad; not to mention the thousands spent on ECC system ram. And cad applications had their own graphics settings and sometimes their own adapter drivers. You weren't necessarily using a consultant but you might have been buying the hardware from the same place you get the software. Things are different today.

Edited by mr_me

Share this post


Link to post
Share on other sites

The 10Gb ethernet is for lan parties, still overkill but you can copy files super fast so you can start gaming sooner.

Sure... But the same could be said of any network where files are being moved across it for any purpose. And, ultimately, in that scenario, LAN speed is going to be dictated by a bunch of other factors:

  • Speed of the TCP/IP stacks in use on both the local and remote machine
  • Speed of the network protocol (e.g., CIFS/SMB, and more on this further down) in use to transfer files
  • Performance of the network infrastructure being traversed
  • Buffers on the network adapters at each end
  • Buffers on the network infrastructure
  • Speed of the storage at either end, but particularly the write speed on the remote side
  • Etc.

Given that SATA disks (which are what are most commonly used in this class of machine) top out at 6Gbps of transfer under ideal conditions, that's the absolute best that can be expected. Of course, those are numbers obtained under lab conditions; real-world numbers will be much lower. Realistically, TCP/IP stacks and network protocols (in this case, most likely CIFS/SMB) have significant overheads of their own - and if you're getting better than 600Mbit with CIFS/SMB, you're doing very well. But that's still only a fraction of what a 10Gbit drop can conceivably offer.

 

Even using something like rsync won't get 10Gbit. Hell, even a commercial product like Aspera, which is UDP-based, will struggle to get to that point, and it's built for speed.

 

Not to mention that if the remote machine or switching infrastructure on the LAN still uses or traverses 1Gbit interfaces, 10Gbit is a moot point.

 

I totally agree with the speed equation, but there are so many bottlenecks with 10Gbit at a desktop level at this point in time that it's really not cost-effective because it can't be used at even close to its full potential. That time will come, sure, but it's still 3 to 5 years out IMHO.

 

Raid setups are for zero downtime because a gamer can't take that risk.

Except that RAID doesn't guarantee zero downtime; it just reduces the risk of downtime and data loss if and only if the correct RAID level is in use.

 

The number of gamer recommendations that I've seen or heard to use only RAID 0 or 1 because 'performance is better' is surprisingly high - and neither of those RAID levels provide a solid degree of redundancy, though RAID 1 is better than RAID 0 from the standpoint of at least having a duplicate of the data. Unless you're running at least RAID 5 (or preferably 6 or 10), however, the benefits are minimal at best with the same (or enhanced) risks present at worst.

 

Raid still doesn't save you from backing up your data.

Totally agreed. But in the gamer world, RAID seems to be seen as a magic bullet for any and all potential data storage risks and considerations. It's not. Even in enterprise environments - where the budgets to do this right are more typically available - drive failure and data loss still happens.

 

I've seen first-hand what happens when an Isilon NAS loses an entire (144TB!) shelf of storage. It's not pretty, and that specific device was one that was designed to do nothing but handle heavy-demand storage needs. Now imagine keeping all of your personal data on a RAID inside the same PC that gets used for gaming, Internet activity, and general-purpose computing. It's scary to think that someone would believe that a) this is a good idea in the first place, or that b) RAID would keep them safe from data loss when they're doing everything you should never do on a storage system.

Share this post


Link to post
Share on other sites
Posted (edited)

You're correct, in the old days Raid 0 or 5 on a desktop was for performance. Raid 1 does have redundancy protection but no performance benefit. Raid 0 has no redundancy but more capacity than raid 5. The reason raid 1 or 5 doesn't save you from regular data backup is because it doesn't protect you from fire or theft. Raid 0 is particularly risky because if one drive fails you lose everything.

Edited by mr_me

Share this post


Link to post
Share on other sites

Repeat til understood:

RAID IS NOT BACKUP

 

It's useful until your RAID controller screws up and your data is corrupted in two places simultaneously. :lol:

  • Like 2

Share this post


Link to post
Share on other sites
Posted (edited)

Usually if the controller fails, your data is intact, just tricky to access. So yes data backup also protects you from this rare event. In the old days raid in the desktop was for performance; not so much today. It still has the benefit of quicker recovery from a failed disk drive but having a mix of solid state and disk drives complicates things. Of course if you don't bother to replace a failed raid drive, you're screwed when the second one fails.

Edited by mr_me
  • Like 1

Share this post


Link to post
Share on other sites
Posted (edited)

Putting on the consultant voice for a moment, I'll add one to the pile of perplexing hardware decisions that appear in a surprising number of gamer-built PCs: RAID cards and ridiculous amounts of storage.

Oh, and 10Gbit network interfaces have virtually zero benefit to online gaming at this time unless your Internet connection is higher than 1Gbit. Save your money and get a decent Intel 1Gbit adapter.

 

Seriously, though: wasting money in the pursuit of trying to build pseudo-enterprise-grade technologies into what's essentially a desktop PC is pointless. Spend the money where it counts, not where it's the equivalent of a +5hp sticker.

 

I did RAID once in a homebuilt rig and found it way more than I needed. And way too expensive for the ever so slight performance improvements to my workflow - doing emulators and music and pictures. So JBOD works better for me and allows me to upgrade piecemeal (though there's always DROBO). And today's NVME speeds are more than sufficient for most all my tasks anyways.

 

JBOD also lets me roughly separate my stuff a little. All my simulation material on one disk. All my media-photos-music on another. Emulation yet another.

 

I also find that most Enterprise technologies are increasing in their internal parallelism to support data streams from many directions. From the chip level through outside interface. All that is wasted on games. Games are not simply parallel optimized yet and don't know how to use anything that is.

 

Besides, today's $150 NVME drive sticks are "Enterprise Class" compared to 300 baud modems and teletypes and 140KByte floppy drives.

 

---

 

Gamer PCs can never have enough storage. Some of my simulations are as much as several hundred gigs. And all my astronomy stuff even more with all the star charts and databases.

Edited by Keatah
  • Like 1

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...