Looking forward to your review of the Intel SSD. Kind of. I don't know if it there's much point to this review actually. All the things you usually measure are known beforehand in this case:


No noise and vibration, whatsoever.

Idle power consumption is pretty much zero. Load power consumption is just a couple watts.

It's going to get crowned on all fronts. Probably pretty much the holy grail of silent computing.

Only thing we don't know yet is if we can afford to get one or not.

SSDs are a mistery to me. Do they run hot? Would one need case airflow to cool? What is the max temperature of a SSD?

Yes they run warm/hot. Kind of like a laptop hard drive. They are low wattage devices but have no fan.

OCZ lists Operating Temp: -10C ~ +70C so if your PC is 35C ambient the drive would have to have 35C rise before that would be an issue. I seriously doubt it's any big deal.

I'd say plug it in and forget about it.

Of course that's assuming you have the money for one. If not you can always make do with the ever cheapening 7200 RPM drives like the WD6400AAKS ($80+S&H gets it to my door at under $90).

I'm really looking forward to SSDs but I'm also willing to wait it out another year or two and go with rotating media one more time.

at least now that intel is pushing the SSD technology, prices should drop even quicker than they have been. I like that they went with 2.5" and 1.8" form factors only, dropping the 3.5" standard. Its about time desktop HDD's moved away from the larger 3.5" form factor.

Now we just need to get some mainstream case manufacturers to start pumping out cases that allow native 2.5" drive mounting.

Yeah interesting news for sure - thanks as always for keeping us up to date from a silent perspective. I have been waiting for years for SSD's to get seriously mainstream, with Intel doing this it's going to force the issue quickly. Everyone has known the bottleneck has been HDD's for ages, even with RAID arrays and perpindicular technology it's still mechanical. I hope quantum computing becomes mainstream in my lifetime

Can anyone say whether SSD's are intrinsically more or less expensive to make than rotating magnetic platter HDD's? In other words, is it just scale/volume and supply/demand factors that will cause it's price to eventually beat magnetic HDD's or is it more complex and expensive to manufacture?

Low power CPU's that can do 1080p HDTV, low power chipsets, now SSD's - gonna have very small passively cooled instant on devices at cheap prices soon for every computing and household need . . . unreal.

There's probably no definitive way to answer this, but my own opinion is that it's almost always scale/volume, supply/demand factors that dictate technology pricing. Memory chips and ICs are highly complex, as are fast spinning magnetic disks that must be balanced and "read", but both are mass manufactured, not hand-crafted. With high enough demand (and thus competition among producers to meet that demand), the machines and processes to build the products improve, and pricing invariably drops.

New socket, "low" 550g weight limit...

They really haven't been paying much attention to the heatsink market recently, have they?

What's the equivalent for LGA775?

I don't know... I don't see myself interested in SSDs until a terabyte-sized drive becomes affordable, especially with hardware in Europe costing double that in the USA.

disappointing to see that push pins are here to stay. They are fine for stock coolers, but they tend to either not provide enough pressure or are a pain to get at with alot of the larger heatsinks. Butt hen if nehelems are pretty much all cooler than previous gens, maybe we wont see many more 1kg heatsinks..

Intel are seemingly pretty much keeping the same TDP range for Nehalem and the extra features suggest from early previews that the CPU itself consumes more power than Penryn at the same speed. It does have a useful boost in performance with multithreaded applications but I wouldn’t expect it overall to consume less power.

480g, so the maximum mass appears to have increased roughly in line with the socket surface area. But it still seems woefully inadequate. I can see their logic though - keep cooling adequate and costs down for the mass market, let enthusiasts carry on using backplates if they feel the need.

Im anxious to see the power consumption related benchmarks for Nehalem. Its great to put all the marketing stuff on slides - but the real meat is when it gets in the hands of reviewers.

And like everyone else im waiting to see what intell will charge for their SSD drives.

I have to agree with this and the other comment about Intel. Intel is the 800 pound gorilla of the computing market. The perfect example of Intel's influence was on the page following the SSD page in Mike's article. VIA has been making a fantastic tiny CPU product for years now. I am very satisfied with my comparitively ancient Via 533mhz Eden chip running my mini-ITX system. Even with as good as VIA's products are, it wasn't until Intel's Atom that the sector suddenly was on fire and every company has an Intel Atom product.

I mentioned this in the other thread, but already Intel's first crack at SSD pretty much goes for broke - 250mb/sec sequential read is almost the limit of SATA 3.0gbit. By the time that SATA 6.0gbit is finalized, they may have SSD's to max that out too, which IMHO is GREAT! The fact that Microsoft wasted their time on Readyboost further illustrates how much the industry insiders understand that the hard drive is the biggest limiting factor in human-perceived speed.

To this day, I still encounter people who are not technically savvy that rank the speed of their computer based on how fast it boots up, or how quickly they can start MS Word and load a document. That's exactly the kind of stuff that tends to be I/O bound by the hard drive. I know SSD is still extremely young technology, but I'm very excited to see Intel, essentially the largest player in affordable computing, seeingly pouncing with intent to beat all comers.

'fantastic' is not the adjective i would choose. "overpriced" and "slow" spring to mind energetically. what exactly are you using your 533Mhz Eden for? file serving?

Yeah, file and web server mainly, very light duty workload, well suited to the low power Eden. Given that nobody else made anything in the miniITX or nano / pico ITX before Intel's Atom, it was simply the price to pay for miniaturization. I agree you pay a lot of money for not a lot of raw performance, but still, what else was there in that form factor before Atom joined the scene?

I agree. PC Perspective's comparison of the Intel Atom mITX vs a new 1.8ghz Nano-cpu mITX from VIA stated that the former could be had for $70-75, and that the latter would be no more than $100. I totally disagree with the speculation about the VIA price. It should be no more than $100, but if it is, it will only be because of pressure from Intel. Look at any mini-itx etailer catalog and you'll see than any VIA boards with CPUs clocked at 1ghz or higher are priced at min $150. For a 1.8ghz CPU part, it'll be closer to $300. I've always thought they were a bit of a ripoff but for quite a while, they were the only game in town, and VIA seemed to establish the pricing for all the others in the sector -- m-itx boards were generally quite pricey for what you got. That's until the first D201GLY Intel mITX appeared last year. $70 was unheard of for a m-itx till then.

Interesting comparison Mike. However there were a couple of things that I thought I should point out.

First - like you said, the price of VIA's solution is still unknown. Hopefully they'll put it at a price that's competitive of the Atom, but their past lineup of products doesn't give me much confidence.

Secondly - The dual-core Atom is due out soon. I'm looking to build a system of my own based around this board (provided that the price hike is reasonable, and the power consumption doesn't increase dramatically). A lot of software released these days are starting to make use of multi-core setups, and I'm sure that the dual-core Atom will make it's own dent in the market.

Let's just say that I'm (im)patiently waiting for the dual-core Atom to appear in stores.

What's wrong with putting an extra twist in the suspension on a Solo/P150?

Why bother making the drive hard mount when you can easily suspend any size drive?

I don't care if next month they decide to start selling 2.9782" drives or 3.159123" drives. They'll still fit in my suspension methods.

Yeah I keep looking for someone to bring these out. I know there are ways to mount 2.5 drives in 3.5 bays but would love a case to come out with it already built that way.

Wait, you weren't suggesting you would still suspend an SSD, were you?

Doesn't the new technology negate the revolutionary adaption to dealing with the problems of the mechanical drive age?

- Dave

I'd hard mount an SSD harder to aid heat dissipation...

That's what I was thinking. There have been some comments on these boards that SSDs do produce heat that isn't negligible, so I'd think we'd now easily go back to hard mounting them to metal frames for more efficient heat transfer since vibration and noise are not an issue.

Regarding the D945GCLF2: Does it have 1024kB of cache, or does it have 1024MB of memory soldered?

SSDs should not make noise but some probably will make some kind of electronic noise some of the time. (like mobos, PSUs and video cards.) If it bothers you, though, you probably need to find a way to reduce your aural sensitivity.

But I don't think these things will produce much heat. It is directly related to power consumption, and at idle in a PC, they barely pull 0.1W. Intel gave a spec of 150mW for typical use in a desktop PC. Even 1~2.5W notebook drives generally do fine in minimal airflow suspension setups; I just can't see how self-heat of a 150mW SSD will be a problem in any PC.

My bad. kb. corrected.

With that thermal comment in mind, Mike, do you think we are close to having the ability to build a small desktop PC for everyday use that is finally without fans or moving disks?

I was encouraged to feel this is close after reading your article, but for the mention that while the CPU on the one board is passively cooled, they couldn't keep the design on the chipset down to avoid a small fan. I wonder if a third-party chipset heatsink would allow for no fans at all in a system based on that board.

Yes.

-- all the technology developed over the last few years to move heat using heatpipes

-- very low power envelope of all key components, specifically the CPU, IGP and the chipset

-- passive cooling using the entire case

It would still require some customization, but because there is so little heat involved, it will be much cheaper and simpler to do than before. Obviously, DIYers will have to be creative, but for manufacturers it should be easy.

Actually, I have a passively cooled mini-ITX case design in the pipeline.

Excellent! I didn't want to pressure an article or review, but since I feel you've been successful getting the message across that the quiet effort starts with selecting quiet components instead of struggling to silence a noisemaker, it seems that low power CPUs and the SSD make for the near-term reality. I'd easily give up the terabyte drive to take what SSD is offered to have such a computer.

You were right a couple months back when I asked about the ASUS EAH3650, it works great right out of the box, and meets all my everyday video needs while never getting above 52C. (I don't game on this machine.) My friends are all impressed as a result, so there's certainly business to come for your sponsors that bring these last few pieces to market.