Hardware innovations you'd like to see become mainstream

[Fushigi]

Just to clarify, what I mean is a jumper on the M/B that remaps at least the first 1M to the video card RAM. Or perhaps you can use all of the video card RAM except 1 or 2 MB so you have graphics capability and the use of almost all of the 32 or 64 or 128 MB on the video card. In the normal(operating) position you use the M/B memory banks as always, and the video card works normally. I don't see any downside to this. If the jumper stops working, it means you probably have a fried M/B anyhow. If the video card RAM goes bad, you need to replace it anyway. There is no RAM whatsoever on the M/B to go bad. Remapping isn't that big of a deal to do in hardware. I'm really surprised nobody's thought of this. It would be very convenient and almost trivial to do with today's chipsets.

I'd prefer it be located on/around the chipset and be high-speed RAM. Switchable as follows: Normal operation a.k.a. if DIMMs are found inthe memory slots: L3 cache. "Special" operation: No DIMMs found; 1MB (or whatever) system RAM. While the engineering isn't exactly easy, it really is not breaking much new ground and provides a real-world performance boost when the 'special' mode is not necessary.

BTW, memmaker and QEMM can both use the B000-B7FF area, which is used by a monochrome video adaptor(and hence not used on modern machines). QEMM's Vidram works a little differently. It disables all graphics(hence it's only useful for test programs) and makes use of the A000 to AFFF, or A000 to B7FFF areas to extend conventional memory to either 704 KB or 736 KB. And yes, it does use the video card RAM for this purpose because it is not possible for the processor to remap anything into the A000 to AFFF area. In fact, the program works on an 8086 with only 640K of physical RAM, and gives either 704 KB or 736 KB(I forgot which) of useable conventional memory. It does definitely use the video adaptor RAM since the 8086 processor is incapable of remapping memory.

My original PC was a Zenith Z-151 8088 (later NEC V20) with 320K conventional, CGA graphics, and dual 320K floppies. Pretty sweet for a 1984 box. One upgrade was to replace the RAM PAL chip and up the DOS conventional RAM to 704K fully DOS-addressable. This was done via hardware; no drivers or anything to load. Very handy when conventional RAM was all I had and I wanted conveniences like SuperKey and SideKick.

- Fushigi

 

[blakerwry]

jtr1962, it sounds like too much of a pain.. I'd rather not worry about whether this video card will work with this mobo..etc.. I'd just rather keep a spare stick of RAM around to test with. It's not like there is a shortage of PC133 or 266 around my house.

 

[Mercutio]

Yes, but it'd be really handy when you get a machine in for service that using those godsforsaken RIMMs. I dunno about you but I'm kind of short in the RDRAM department (OK, I admit it, I *do* have a couple... but I wouldn't if I didn't have to).
Not to mention POS boards that won't work without registered or ECC memory.

 

[blakerwry]

i've never had to do work on RIMM boards.... and most baords that i see that require ECC/Registed are usually server type boards... not typically what i would call POS.

 

[Mercutio]

Two words: DFI P2XBL.
What a pain in the ass those things are.

 

[Jan Kivar]

I'd really like to see cases that have built-in 120mm slots for fans & filters. Probably two in front and one in back (I'd like two, but there really is no space...). And with the fans, some software & temperature controlled fanbus. Basically stuff that will make my box less loud.

Cheers,

Jan

 

[Handruin]

Welcome Jan,

I would like to see a better air channeling rather than larger fans. If the air could be controlled more efficiently, I think that would be more effective than a brute-force method of air-cooling.

 

[blakerwry]

anybody seen shuttles ICE heatpipe?

http://www.shuttle.com/new/images/productimages/ph4.jpg

one part connects to the CPU like a normal HSF, however through conduction heat is transfered to a radiator type part where heat can be transfered out of the case through standard air cooling.

it works surprisingly well... eliminating the need for seperate CPU and case fans in most systems while keeping the temps low.

 

[Jan Kivar]

Welcome Jan,

I would like to see a better air channeling rather than larger fans. If the air could be controlled more efficiently, I think that would be more effective than a brute-force method of air-cooling.

Best alternative is to make a hole through the side panel and attach a fan & fan duct to the CPU heat sink. But the hole will let the sound come out even more. Some Intel specsheets planned a system that Dell for instance is using; one fan sucks the heat out and replaces the CPU fan.

Blakerwry, heat pipes are neat, but they have to be designed by the motherboard & case. I have seen though pics of a computer that has nearly half meter long heat pipe. The hot end is attached to a heat sink that has been installed through the top of the case...

I think water cooling is the way to go with the new processors. Over 70W is too much for a normal heat sink to handle. I'm currently using Swiftech's MCX-462. It cost nearly as much as my processor when I bought it...

Handruin, thanks for the welcome... I appreciate it.

Cheers,

Jan

 

[Mercutio]

Bigger fans spin more slowly with the same impact as smaller fans spinning quickly. So why not control the airflow from a bigger fan?

Since I spent a little while dealing with a FlexATX machine today... how about a small formfactor PC design that doesn't suck? Any PC that I need 45 minutes to install RAM in has something wrong with it. Continuing my string of positive comments about Apple, the flat-panel Imac, which is certainly a silly small design, has its SODIMM slots right under the bottom base-cover.

Why isn't there an international marker for "This is a lever, pull it" or "This is the button that opens the cute little door". Heck, I think a hardware improvement that a lot of techs would like to see is "instructions for opening the funny-looking vomit-box", printed on the underside of the machine.

 

[mubs]

Heck, I think a hardware improvement that a lot of techs would like to see is "instructions for opening the funny-looking vomit-box", printed on the underside of the machine.

I'll second that !!!

Reg. the Dell designs, if anyone's been near a newer Dell (< 6months old) they're awfully quiet, to the point where you wonder if they're on at all. The've done a good job keeping those suckers quiet. I haven't examined them in detail as to why this is so.

 

[Mercutio]

My gues is all that thick, soft plastic.

 

[Buck]

Dell has insulated things quite nicely with a two-layer case. The first layer being metal (beer can thickness) and then molding the outside in plastic. But they also keep things quiet on the cooling end. The CPU uses a heatsink only, and then an airduct creates a tunnel to an opening at the back of the case where there is a 80-mm low-rpm exhaust fan. Top this off with a decently quiet PSU fan and you have a new Dell. To top this off, the plastic rails that hold the drives in place use rubber grommets and plastic thumb screws to mount themselves to the drive.

 

[timwhit]

Buck, you forgot to mention how hard it is to figure out how these cases are to disassemble or get a drive out of. (At least for me...and I can figure out most Packard Bell cases without a problem)

 

[Buck]

Buck, you forgot to mention how hard it is to figure out how these cases are to disassemble or get a drive out of. (At least for me...and I can figure out most Packard Bell cases without a problem)

I suppose insanity comes with experience Tim, I've worked on quite a few Dells recently and have learned a few tricks about replacing products. The fact that you Packard Bell systems easier to disassemble highlights my idea that insanity comes with experience.

 

[P5-133XL]

Heck, I think a hardware improvement that a lot of techs would like to see is "instructions for opening the funny-looking vomit-box", printed on the underside of the machine.

Everyone spun their wheels for a while and then came up with a corollary to my one basic law - One single simple way to open and close the case.

 

[Splash]

If the computer industry could just stop the clock speed madness and concentrate on holding clockspeed at about 2.5 GHz, maybe by the time microprocessors are produced using 90nm manufacturing process a microprocessor would only require a passive heatsink for normal operations with a small fan that only switches on during unusually hot ambient room temperature conditions.

Use of the computer case with an augmented with a heat spreader would be the ideal approach. The heat spreader would eliminate any hot spots on the case.

 

[Mercutio]

Having just broken one of the pegs off the ZIF socket of a gav7rxp...

A heatsink retention technology that doesn't suck.

 

[blakerwry]

Having just broken one of the pegs off the ZIF socket of a gav7rxp...

A heatsink retention technology that doesn't suck.

heatsinks used to attach to the CPU's themselves... personally I would much rather use the p4 or the socket7/370/A type of heatsinks than the old socket 1/2/3.

oh, and splash.. I think what you're looking for is called a VIA C3... maybe not a .9 micron process yet, but can be passively cooled and runs at about 1gHz... give it another year or so and you might see 2ghz models @ .9micron.

 

[Platform]

...oh, and splash.. I think what you're looking for is called a VIA C3... maybe not a .9 micron process yet, but can be passively cooled and runs at about 1gHz... give it another year or so and you might see 2ghz models @ .9micron.

Normally, I tend to run away in horror when I hear the name Via.

However, I wouldn't mind sampling a decent piece of computing hardware using a C3 microprocessor, hopefully one with a power supply that has also has no cooling fan (ONE emergency cooling fan would be acceptable, though).

 

[Cliptin]

However, I wouldn't mind sampling a decent piece of computing hardware using a C3 microprocessor, hopefully one with a power supply that has also has no cooling fan (ONE emergency cooling fan would be acceptable, though).

I have one I will loan you or sell cheap. C3 ~500 and set-top box with external power supply.

 

[Mercutio]

I would *love* to see a PC-based component that could handle most of the functions of a HT receiver. A digital in and out, plus some banana plugs for real speakers... it'd be sweet.

 

[Pradeep]

I don't know whether we will see banana plugs on PC's anytime soon. The Europeon governments are worried that their slaves...oops citizens will plug their appliances into them and fry.

 

[mangyDOG]

Dual AGP slot support. Supposedly it came in with AGP 8x, but none of the toy vendors (Abit, MSI, Asus) have done it yet.

Mmm, imagine dual Matrox Parhalia's - 6 monitors - thats full surround 360 degree gaming!!!!

Even if the frame rate is a bit lower that the latest 9700pro who wants to play with one monitor if you can have 6....

 

[Tea]

Did zomebody zay banana?

But let'z face it, when did audio manufacturerers agree about plug standards any better than computer industry morons .. er .. I mean "designers"? Well, OK, no-one is as bad as computer people, but the audio industry iz ztill pretty bad. Can you imagine it? 17 different pairs of plugs on the back of your PC so that you can use whichever one fits, depending on which beand of speakers you have?

(Stop exagerating, Tea.)

(Zorry.)

 

[Jan Kivar]

I would *love* to see a PC-based component that could handle most of the functions of a HT receiver. A digital in and out, plus some banana plugs for real speakers... it'd be sweet.

You can do as much as a pre-amp. There is little sense putting more heat producing components (=amplifiers) in the box. Plus the computer is really electronically noisy inside. You could build one if You'd really like though.

However, computer is better just as source.

Cheers,

Jan

 

[blakerwry]

yes.. get a real reciever if you want a HTPC

 

[Mercutio]

Real speaker connectors would be cool, and the "card" in question could be a low-EMI firewire device...

 

[Jan Kivar]

So what You'd basically need is a receiver with computer connectivity. I think that JVC and H/K had (have?) some products that had an USB port.

Jan

 

[Mercutio]

Sort of... USB is WAY too much of a load on a system CPU, and the only products I know of (Yamaha's, HK, Creative) are all USB, and don't have the functionality I'd like to see - eg video switching (especially switching to a VGA connector). If we're going to go outside the PC, those would be essential to me.

 

[blakerwry]

reciever built for computers? why dont you just use the computer like any other source(DVD, VCR, etc).... use the digital audio cables or get a mini-stereo to RCA adapter and use analog audio...

as far as video goes if your reciever supports video try to get one that has multiple S-video inputs... the only reciever i've seen that had video inputs only had 1 s-vid input/output, but had multiple RCA video inputs...

 

[Jan Kivar]

I have Yamaha RX-V596RDS, and it has... let me check... four S-VHS in (one with dedicated out) back panel, one S-VHS out (monitor out), and one S-VHS in the front panel as well. And same inputs have composite video too.

Newer HT amps have even component in/out ports, I don't know whether You could connect computer via these ports. Most amps just pass through the picture (from similiar port to another), so You'd have to have a TV/HDTV/projector that has the appropriate input.

Cheers,

Jan

 

[Jan Kivar]

Hmm... I meant S-Video. They use same connector though...

Cheers,

Jan

 

[blakerwry]

So what You'd basically need is a receiver with computer connectivity. I think that JVC and H/K had (have?) some products that had an USB port.

Jan

so why did you write this? i don't understand why a reciever would need "computer connectivity" or anything else special to use your computer... like a USB port....

 

[Mercutio]

Blake, the goal of such a device would be to eliminate some of the many, many cables and expensive devices that are connected to an HTPC.
Right now I can't even consider removing my HTPC from its receiver, even though all that receiver is, to my computer, is a way to hook up "real" speakers and a decent remote control.

 

[double bit CRC error]

hmm... I would estimate from computer to reciever you would have

2 or 3 x mini stereo -> RCA cables (4 to 6 channels of sound)
1 x digital Coax or TosLink cable (Digital sound is Uber|337)
1 x S-Video or RCA cable (got to show off that desktop)


From sources into the HTPC you would have:

1 x RCA or S-video cable (for digitally recording TV or whatnot)
1 x RCA to mini-stero cable (so you can have your cake and eat it too)



Then you have the standard wireless keyboard/mouse (dis)conencted to the computer... and maybe a remote...

What am I forgetting?


Certainly a single cable breakout box such as the hercules Game theatre Xp uses would make (dis)connecting audio alot easier... same thing with a DIN cable like the Fortissimo II or some of the Philips cards use.

 

[Mercutio]

The following components are involved in mine:

VCR
SVHS VCR
TV
DVD Changer
PC
Receiver
Projector
5 speakers, 1 sub
2.4GHz wireless video distributor (doesn't pass macrovision, which is happy for dubbing)
2.1 PC speaker set for occasional sounds from PC

I've got the VCRs plugged into the receiver for audio and video. The TV (video-only) plugged into the receiver's composite output, the computer plugged into the receiver's svideo output, the 6 speakers plugged into the receiver, the computer plugged into the receiver through SPDIF, the computer plugged into the projector's VGA input, the DVD changer plugged into the projector's component input, composite AV out from PC to wireless distributor, svideo out from PC to SVHS deck...

And I think that's everything. I think. In a perfect world, being able to do all that switching on a PC directly would make things a lot easier, but for the moment, I can watch any source I have, through the PC to the projector or through receiver to TV, I can record to PC, SVHS or plain old VHS, and I can watch any of the above on one of the two other computers in my apartment with a VCR & tuner card.

Of course, given all of the above, mayhap I shouldn't be complaining about cables...

 

[blakerwry]

2.4GHz wireless video distributor (doesn't pass macrovision, which is happy for dubbing)

That's a neat idea....

 

[Mercutio]

Look on Ebay. Those things cost like $25 used. That's where I got mine, anyway (then paid the new price for an extra receiver, sigh).

 

[Jan Kivar]

So what You'd basically need is a receiver with computer connectivity. I think that JVC and H/K had (have?) some products that had an USB port.

Jan

so why did you write this? i don't understand why a reciever would need "computer connectivity" or anything else special to use your computer... like a USB port....

Blake, the goal of such a device would be to eliminate some of the many, many cables and expensive devices that are connected to an HTPC.
Right now I can't even consider removing my HTPC from its receiver, even though all that receiver is, to my computer, is a way to hook up "real" speakers and a decent remote control.

Mercutio, You are right on the spot. I don't want to quote your cabling here, but You could do it all with just one FireWire.

blakerwry, this is what I meant:

Current amps have inputs for digital audio, so You could record it directly (is it even recording if You just write the PCM stream??).

You could record the 5.1 channel SACD audio in six discrete channels (if You just have the format, maybe pack it to DD or DTS stream). You could record any video source. These two will have some sort of degradation, as the signal goes DA->AD along the way. Here in Europe digital TV is becoming a reality, so You could record the stream directly if You had DTV tuner card in your HTPC.

All this with one cable... Actually I don't know if it's possible to transfer video in FireWire directly to the monitor, especially if there is "something" else in the cable than just the video.

Cheers,

Jan