Slashdot Mirror

On this subject: I finally got around to reading Anandtech's very long article about the current crop of SSD drives. I feel like it was pretty educational, which is good because it took a long time to digest.

In its discussion of performance degradation as drives are used, the article explains that individual pages of NAND memory can't be rewritten. Early in a drive's life, page are remapped when they are rewritten by the OS. As the drive is used, the drive runs out of pages to remap and is forced to copy a block (typically a 512KiB collection of 4KiB pages) to cache, erase the block and then rewrite the block with the new pages. That explains pretty well why write performance degrades, since writing to a block that has data must perform a read and erase operation in addition to the write. However, that explanation also leaves open the question of how the drive prevents data loss if it loses power. Worst case, the OS issues a write and the drive copies a 512KiB block to cache and erases the block, and then loses power. Due to remapping, literally anything could be in that half a MiB. The data loss could corrupt the file that was being modified, obviously, but also any other file on the drive, or parts of the filesystem itself.

I figure there's got to be protection against data loss built-in, but I'm not able to find details regarding any individual drive or manufacturer's approach to solving that problem. Does anyone know more about this subject?

I can say that the "sweet spot" in graphics is between a 6200 and a 7600 on the Nvidia side with the 6600 and 7600 being quite popular and widely used, with the X1650PRO being quite popular on the ATI side of the pond.

While I think you have the right idea, your recommendations are a little out of date.

I still have a 6600GT myself, and I love it, but my motherboard is still running AGP, so I can't really get a new video card without upgrading my whole machine. For buying a new card though, I can't recommend the 6- or 7-series today.

http://anandtech.com/guides/showdoc.aspx?i=3538

The first couple pages are the price range you're talking about; cards that work for the majority of gamers, without killing their bank accounts. For ~$75, you can get a 4670 or 9600GT. Spending any less than that is pretty much a waste.

There's still plenty reason to blame monopoly. Check out this thread on the AT forums, user VinDSL comes up with some interesting facts + sources about these supposed "Higher return rates" with Linux. Apparently it's all FUD marketing from MS:

Some netbook retailers are slamming Linux for boosting their product-return rates. Here's why you should take their protests with a grain of salt.

Lately, quite a few netbook makers and resellers are saying that buyers return Linux machines far more often than identical models running Windows XP. Last year, for example, the director of U.S. sales for MSI told Laptop Magazine that customers return Linux netbooks four times as often as Windows netbooks.

As Computerworld.com contributor Eric Lai pointed out, however, such claims can be misleading.

According to Lai, MSI's numbers weren't based on the company's actual netbook return rates.

In fact, at the time, MSI wasn't even shipping a Linux-powered netbook model.

Source.

A few weeks ago /. linked to a really wonderfully written article by Anand Lal Shimpi about SSD drives. In the article he includes some simple and clear explanations of how flash memory works, its lifespan, and how it handles writes and deletes to maximize the life of every block of storage.

http://www.anandtech.com/printarticle.aspx?i=3531

The only think missing from the article is a description of the behaviour of a failing drive.

One thing about this research paper is that they used only one model MemoRight GT MR25.2 in 8/16/32 GB capacities to do their testing before 2008-11-11 publication of the paper in the United Kingdom.

I'm concerned that the research test and results are largely skewed against SSDs because they used only that one model to do all their testing with based on only one price point for the SSDs.

There is a very large difference in performance between many various SSD drives based on the original flawed JMicron JMF602 chipset (stuttering/freezing on write), newer JMF602B (smaller stuttering), Samsung's chipset, Intel's chipset (fastest random writes by 4x), and the newest Indilnix Barefoot chipset (balanced sequential/random read/write). Additionally the huge drops in prices in the last 6-12 months ($1,500->$400) is a big change in the SSD arena. These price, capacity, and performance changes are going to continue fluctuating for the next few years yielding much better drives for the consumers.

I believe that the research in the paper will be shortly obsolete, if it isn't already, given the latest products on the market and price points and the Q3/Q4 new upcoming products from Intel and others.

I'm helping a friend of mine build an all-in-one HTPC / Desktop / Gaming system and I've been doing research into SSDs for the past few weeks based on reviews and benchmarks so I wanted to share my info.

Basically there are only two drives to consider and I list them below. A good alternative at this time is to purchase smaller SSDs and create RAID-0 (stripping) sets to effectively double their performance instead of buying a single large SSD. The RAID-0 article below shows great benchmark results to this effect.

Intel X25-M

The Intel X25-M series of drives is the top performance leader right now, and the 80GB drive is barely affordable for a desktop system build if you consider the increased performance of the drive.

Intel X25-M SSDSA2MH080G1 80GB SATA Internal Solid state disk (SSD) - Retail - $383.00 USD ($ 4.7875 / per GB)

OCZ Vertex

The new OCZ Vertex series of drives with the newer 1275 firmware is the price/performance leader and they are much more affordable than the Intel drives. When you combine two of these smaller 30/60 GB drives into RAID-0 (stripping) you get double the performance at still acceptable prices.

OCZ Vertex Series OCZSSD2-1VTX30G 2.5" 30GB SATA II MLC Internal Solid state disk (SSD) - Retail - $129.00 USD ($ 4.3 / per GB)

OCZ Vertex Series OCZSSD2-1VTX60G 2.5" 60GB SATA II MLC Internal Solid state disk (SSD) - Retail - $209.00 USD ( $ 3.483 / per GB)

Reviews

Required Reading:
AnandTech - The SSD Anthology: Understanding SSDs and New Drives from OCZ

AnandTech - Intel X25-M SSD: Intel Delivers One of the World's Fastest Drives

AnandTech - The SSD Update: Vertex Gets Faster, New Indilinx Drives and Intel/MacBook Problems Resolved

RAID-0 Performance:
ExtremeTech - Intel X25 80GB Solid-State Drive Review - PCMark Vantage Disk Tests

BenchmarkReviews - OCZ Vertex SSD RAID-0 Performance
(Be Warned about BenchmarkReviews! Synthetic benchmark results only, no real-life benchmarks such as PCMark Vantage.)

One thing about this research paper is that they used only one model MemoRight GT MR25.2 in 8/16/32 GB capacities to do their testing before 2008-11-11 publication of the paper in the United Kingdom.

I'm concerned that the research test and results are largely skewed against SSDs because they used only that one model to do all their testing with based on only one price point for the SSDs.

There is a very large difference in performance between many various SSD drives based on the original flawed JMicron JMF602 chipset (stuttering/freezing on write), newer JMF602B (smaller stuttering), Samsung's chipset, Intel's chipset (fastest random writes by 4x), and the newest Indilnix Barefoot chipset (balanced sequential/random read/write). Additionally the huge drops in prices in the last 6-12 months ($1,500->$400) is a big change in the SSD arena. These price, capacity, and performance changes are going to continue fluctuating for the next few years yielding much better drives for the consumers.

I believe that the research in the paper will be shortly obsolete, if it isn't already, given the latest products on the market and price points and the Q3/Q4 new upcoming products from Intel and others.

I'm helping a friend of mine build an all-in-one HTPC / Desktop / Gaming system and I've been doing research into SSDs for the past few weeks based on reviews and benchmarks so I wanted to share my info.

Basically there are only two drives to consider and I list them below. A good alternative at this time is to purchase smaller SSDs and create RAID-0 (stripping) sets to effectively double their performance instead of buying a single large SSD. The RAID-0 article below shows great benchmark results to this effect.

Intel X25-M

The Intel X25-M series of drives is the top performance leader right now, and the 80GB drive is barely affordable for a desktop system build if you consider the increased performance of the drive.

Intel X25-M SSDSA2MH080G1 80GB SATA Internal Solid state disk (SSD) - Retail - $383.00 USD ($ 4.7875 / per GB)

OCZ Vertex

The new OCZ Vertex series of drives with the newer 1275 firmware is the price/performance leader and they are much more affordable than the Intel drives. When you combine two of these smaller 30/60 GB drives into RAID-0 (stripping) you get double the performance at still acceptable prices.

OCZ Vertex Series OCZSSD2-1VTX30G 2.5" 30GB SATA II MLC Internal Solid state disk (SSD) - Retail - $129.00 USD ($ 4.3 / per GB)

OCZ Vertex Series OCZSSD2-1VTX60G 2.5" 60GB SATA II MLC Internal Solid state disk (SSD) - Retail - $209.00 USD ( $ 3.483 / per GB)

Reviews

Required Reading:
AnandTech - The SSD Anthology: Understanding SSDs and New Drives from OCZ

AnandTech - Intel X25-M SSD: Intel Delivers One of the World's Fastest Drives

AnandTech - The SSD Update: Vertex Gets Faster, New Indilinx Drives and Intel/MacBook Problems Resolved

RAID-0 Performance:
ExtremeTech - Intel X25 80GB Solid-State Drive Review - PCMark Vantage Disk Tests

BenchmarkReviews - OCZ Vertex SSD RAID-0 Performance
(Be Warned about BenchmarkReviews! Synthetic benchmark results only, no real-life benchmarks such as PCMark Vantage.)

One thing about this research paper is that they used only one model MemoRight GT MR25.2 in 8/16/32 GB capacities to do their testing before 2008-11-11 publication of the paper in the United Kingdom.

I'm concerned that the research test and results are largely skewed against SSDs because they used only that one model to do all their testing with based on only one price point for the SSDs.

There is a very large difference in performance between many various SSD drives based on the original flawed JMicron JMF602 chipset (stuttering/freezing on write), newer JMF602B (smaller stuttering), Samsung's chipset, Intel's chipset (fastest random writes by 4x), and the newest Indilnix Barefoot chipset (balanced sequential/random read/write). Additionally the huge drops in prices in the last 6-12 months ($1,500->$400) is a big change in the SSD arena. These price, capacity, and performance changes are going to continue fluctuating for the next few years yielding much better drives for the consumers.

I believe that the research in the paper will be shortly obsolete, if it isn't already, given the latest products on the market and price points and the Q3/Q4 new upcoming products from Intel and others.

I'm helping a friend of mine build an all-in-one HTPC / Desktop / Gaming system and I've been doing research into SSDs for the past few weeks based on reviews and benchmarks so I wanted to share my info.

Basically there are only two drives to consider and I list them below. A good alternative at this time is to purchase smaller SSDs and create RAID-0 (stripping) sets to effectively double their performance instead of buying a single large SSD. The RAID-0 article below shows great benchmark results to this effect.

Intel X25-M

The Intel X25-M series of drives is the top performance leader right now, and the 80GB drive is barely affordable for a desktop system build if you consider the increased performance of the drive.

Intel X25-M SSDSA2MH080G1 80GB SATA Internal Solid state disk (SSD) - Retail - $383.00 USD ($ 4.7875 / per GB)

OCZ Vertex

The new OCZ Vertex series of drives with the newer 1275 firmware is the price/performance leader and they are much more affordable than the Intel drives. When you combine two of these smaller 30/60 GB drives into RAID-0 (stripping) you get double the performance at still acceptable prices.

OCZ Vertex Series OCZSSD2-1VTX30G 2.5" 30GB SATA II MLC Internal Solid state disk (SSD) - Retail - $129.00 USD ($ 4.3 / per GB)

OCZ Vertex Series OCZSSD2-1VTX60G 2.5" 60GB SATA II MLC Internal Solid state disk (SSD) - Retail - $209.00 USD ( $ 3.483 / per GB)

Reviews

Required Reading:
AnandTech - The SSD Anthology: Understanding SSDs and New Drives from OCZ

AnandTech - Intel X25-M SSD: Intel Delivers One of the World's Fastest Drives

AnandTech - The SSD Update: Vertex Gets Faster, New Indilinx Drives and Intel/MacBook Problems Resolved

RAID-0 Performance:
ExtremeTech - Intel X25 80GB Solid-State Drive Review - PCMark Vantage Disk Tests

BenchmarkReviews - OCZ Vertex SSD RAID-0 Performance
(Be Warned about BenchmarkReviews! Synthetic benchmark results only, no real-life benchmarks such as PCMark Vantage.)

Any chances that you still have the link(s)?

Because my reading of Anand's research tells me that in active, non-stop use SSD would fail in about the same time as normal laptop 1.8"/2.5" harddrives - 1-1.5 years. Limit on number of rewrite cycles is high (~100k), yet is quite easy to reach.

The article you cite does not contain the 1-1.5 years figure anywhere. How did you get that number? For what it's worth, I've been using solid state drives in both my laptops for more than a year now, with no problems whatsoever.

Another very important point which often gets ignored is that a solid state drive failure is far more benign than a spinning platter drive failure. When a solid state drive fails, you lose the ability to write data, but you can still read data. On the other hand, failure of a spinning platter drive means that you can't read your data anymore, at least not without sending it to a very expensive data recovery firm.

Any chances that you still have the link(s)?

Because my reading of Anand's research tells me that in active, non-stop use SSD would fail in about the same time as normal laptop 1.8"/2.5" harddrives - 1-1.5 years. Limit on number of rewrite cycles is high (~100k), yet is quite easy to reach.

I didn't read the hothardware article. Did they specify at which resolutions which card wins? Did they test with the newest 185 Nvidia drivers? They're moderately slower than the 182's.

Anandtech, my personal favorite reviewer (none of that 1 paragraph/page + 100 page article nonsense *cough tomshardware cough*) tells a different story.

In case you don't feel like clicking-- 4890 takes the cake hands down on 24" and sub 24" displays (1920x1200 resolution and lower). At 2560x1200, it's a tossup.

Considering you can buy the 4890 right now and the GTX275 won't be available for 2 more weeks, I think it's pretty clear which card to get.

Looking at a wider range of reviews, I think we can call this round a draw. That means the real winners are consumers, because the selling point will become price.

Or, if you read the most interesting review of these cards, you'll see why maybe nvidia will skip the price game this time and instead try (and fail) to sell their cards based on physx:

http://www.anandtech.com/video/showdoc.aspx?i=3539

Many people run headless Mac OS X systems and they're OK.

And how many of those people are slashdotted?

Lots of people don't know how to tune apache+whatever database in order to handle slashdot-type loads (hint: the defaults don't work).

And if they did, is there a way to 'tune' OS X to compensate for the horrible kernel I/O handling? There have been a handful of tests demonstrating OS X's 'server' inadequacies in this department - notably, here and here. Seriously: 1/10th the performance of the same software running on Linux is really quite absurd. Maybe it's better in 10.5, and maybe it'll be better still in 10.6, but my experience with dbs in 10.5 are still pretty lackluster, even if they're better (I don't know).

Yes, there's the --skip-thread-priority build option for MySQL, which supposedly increases OS X mysql performance. Are you referring to this? If so, how much does it improve performance?

Many people run headless Mac OS X systems and they're OK.

And how many of those people are slashdotted?

Lots of people don't know how to tune apache+whatever database in order to handle slashdot-type loads (hint: the defaults don't work).

And if they did, is there a way to 'tune' OS X to compensate for the horrible kernel I/O handling? There have been a handful of tests demonstrating OS X's 'server' inadequacies in this department - notably, here and here. Seriously: 1/10th the performance of the same software running on Linux is really quite absurd. Maybe it's better in 10.5, and maybe it'll be better still in 10.6, but my experience with dbs in 10.5 are still pretty lackluster, even if they're better (I don't know).

Yes, there's the --skip-thread-priority build option for MySQL, which supposedly increases OS X mysql performance. Are you referring to this? If so, how much does it improve performance?

1. Hard Drives do this already.

2. Some of the SSDs in question do this already. [TFA]

3. Most filesystems do equivalent things already by delaying and aggregating writes.

From the OS's point of view the write completes as soon as the data is in cache RAM, which is almost instantly.

As stated, this is a Bad Thing. Most of the time, it is acceptable, but the filesystem (which is what you meant) does occasionally need to be able to force data to be written to disk immediately.

What do they say on Everything2? "Your revolutionary ideas on the future of storage have already occurred to others." Depending on your specific needs, you should:

1) use a ramdisk
2) use a RAID array
3) put the extra RAM to use as general system memory
4) use a file system with a long write delay
5) use a UPS
6) any combination of the above

Also, not using Windows is going to directly affect i/o performance as well as make it simpler to implement the above.

Mod parent up. This is REALLY the best article I've EVER read about SSD performance. If you are interested in buying an SSD do yourself a favor and read it, and understand it.

Anand consistently delivers top notch articles. They've had the same, unobtrusive, useful interface for the last, oh, 8 years now? They haven't pumped the site full of advertisements either like some other *cough tomshardware cough* websites.

For other good reads from Anandtech check out the DX11 writeup or their Intel SSD article.

Mod parent up. This is REALLY the best article I've EVER read about SSD performance. If you are interested in buying an SSD do yourself a favor and read it, and understand it.

Anand consistently delivers top notch articles. They've had the same, unobtrusive, useful interface for the last, oh, 8 years now? They haven't pumped the site full of advertisements either like some other *cough tomshardware cough* websites.

For other good reads from Anandtech check out the DX11 writeup or their Intel SSD article.

I've always liked Anand's articles primarily because he's not afraid to be frank and say something that's bad is bad. He doesn't sugar coat. Sometimes, when a product launches, he reviews it, and he says it doesn't live up to the hype, or X thing is missing/wrong/etc, I get bummed, but he does quite well at putting it all into perspective.

At the same time, I "feel" this giddy-nerd-joy when he writes about something that is ground-breaking or game-changing (RV770, Nehalem, etc). Take a look at this article, it's about how the RV770 came to be, and if you liked the SSD article and the back story, you'll surely be impressed by this one: http://www.anandtech.com/printarticle.aspx?i=3469

3) Speaking of noise - WD300 Raptors? Congrats, you just put the noisiest modern hard drives in a machine "built to be quiet" - if no expense was to be spared, why is this thing not outfitted with Solid State Disks???

    That's not really accurate, the newest Velociraptors are 2.5" hard drives encased in a large 3.5" heatsink that also is very effective at quieting the drive. Anandtech measured extremely reasonable sound levels in its review, so I'd be careful before casting aspersions on that front.

http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3403

They showed that the non-Intel MLC drives can have some serious performance issues, so I'de stick with an Intel MLC. If designed correctly, an SSD is actually more reliable than a standard HD, since you know exctly when it's going to fail. Here's some key points from the article on reliability:

page 4: OEMs wanted assurances that a user could write 20GB of data per day to these drives and still have them last, guaranteed, for five years. ... Intel went one step further and delivered 5x what the OEMs requested. Thus Intel will guarantee that you can write 100GB of data to one of its MLC SSDs every day, for the next five years, and your data will remain intact. Intel actually includes additional space on the drive, on the order of 7.5 - 8% more (6 - 6.4GB on an 80GB drive) specifically for reliability purposes. If you start running out of good blocks to write to (nearing the end of your drive's lifespan), the SSD will write to this additional space on the drive.

page 5: Intel's SSDs are designed so that when they fail, they attempt to fail on the next erase - so you don't lose data. If the drive can't fail on the next erase, it'll fail on the next program - again, so you don't lose existing data. You'll try and save a file and you'll get an error from the OS saying that the write couldn't be completed. The beauty here is that the SSD knows exactly when it can't erase/program a block, and if the drive knows, then you can use software to ask the drive what it knows. In the near future Intel will be releasing its own SSD tool that will let you query two SMART attributes on the drive: one telling you how close you are to the rated cycling limit, and one telling you when you've run out of reallocating blocks. The latter is the most important because Intel fully expects these drives to outlast their rated limits. As bad blocks develop, the SSD will mark them as such and write to new ones - by telling you when it has run out of bad blocks (or nearly run out of bad blocks), you'll know exactly when you need a new hard drive.

There are very few benchmarks where the Phenom II beats an i7. Anandtech's benchmarks show this. Google for pretty much any Phenom II review and you'll see that. They're decent chips, and very competitive given their cost. The absolute highest performing however...

Please cite a benchmark where the 940 (3Ghz Phenom II X4) performs worse than the 2.6 GHz i7. I'd really like to see it.

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=3512&p=1

I don't think there's a single benchmark in this review where the Phenom II X4 940 (3.0GHz) beats the Core i7 920 (2.66GHz). Some are close, but the i7 920 is actually much faster overall.

The Intel chips most comparable in performance to the Phenom II 940 are the Q9400 and Q9550, old Penryn quadcores at 2.66 and 2.83 GHz. That's my subjective eyeball evaluation of the benchmark data, anyways. One could try to do something SPEC-like and normalize the benchmarks to a baseline machine and then take the geometric mean for greater formality, but I'm not doing that for an anonymous slashdot post. :)

In an update defined by new graphics chipsets that were build specifically to accelerate high definition video playback these geniuses are testing the processor performance.

If you are interested in some useful numbers, anandtech did some good competitive tests on the current generation of integrated graphics chipsets. No, these are not inside a Mac Mini, but it provides much more relevant information than this ridiculous article.

Those Anandtech high-def video benchmarks might be worthless for the Mac mini because, AFAIK, the OS X drivers have not yet enabled full hardware decode for MPEG-2, VC-1 and H.264 on the GPU. This had been a Windows-only feature for a long time and was just recently enabled in a Linux beta driver (NVIDIA binary).

The Linux driver looks like a good sign (for eventual OS X support), but high-bitrate 1080p H.264 playback will continue to be a CPU-hog on the Mac mini (with a whirring CPU fan) until the OS X driver gets full hardware decode support.

In an update defined by new graphics chipsets that were build specifically to accelerate high definition video playback these geniuses are testing the processor performance.

If you are interested in some useful numbers, anandtech did some good competitive tests on the current generation of integrated graphics chipsets. No, these are not inside a Mac Mini, but it provides much more relevant information than this ridiculous article.

Those Anandtech high-def video benchmarks might be worthless for the Mac mini because, AFAIK, the OS X drivers have not yet enabled full hardware decode for MPEG-2, VC-1 and H.264 on the GPU. This had been a Windows-only feature for a long time and was just recently enabled in a Linux beta driver (NVIDIA binary).

The Linux driver looks like a good sign (for eventual OS X support), but high-bitrate 1080p H.264 playback will continue to be a CPU-hog on the Mac mini (with a whirring CPU fan) until the OS X driver gets full hardware decode support.

This might interest you. Anand Tech did some recent benchmarks on the current integrated chipsets, including the same one the new Mini uses.
http://www.anandtech.com/mb/showdoc.aspx?i=3432
It's not a mini, but it is a similar setup. The difference is huge and quite intriguing. I'm giving serious thought to using a new Mini as a front-end myself. When OSX gets bluray support I'll swap out the drive for a sony slot-loading bluray laptop drive. Should make a very nice all-in-one media box.

These benchmarks are meaningless and worthless. The site itself says that these are artificial tests based almost entirely on processor power. So, similar processors with the same RAM is going to give the same 'score' regardless of OS, video card, hard drive performance or any other factor. In an update defined by new graphics chipsets that were build specifically to accelerate high definition video playback these geniuses are testing the processor performance.

These are not Mac benchmarks. They are intel processor benchmarks. You could have gotten the same numbers months ago (and many sites have) by testing the new intel processors as they came out.

If you are interested in some useful numbers, anandtech did some good competitive tests on the current generation of integrated graphics chipsets. No, these are not inside a Mac Mini, but it provides much more relevant information than this ridiculous article.

Taiwan != China. Actually TSMC has been making chipsets for the Atom for some time, so I'm told. The Atom itself was made by Intel, on its latest process. TSMC lags behing Intel in process technology, but apparently that no longer matters for Atom. As anantech put it

http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3522&p=2
The other thing to keep in mind is that Moorestown, the first Atom SoC, will be built at 45nm while the first 32nm CPUs are shipping from Intel. Another way of putting it is that Atom processors don't appear to need the latest manufacturing process, just one that's mature and good enough. TSMC is transitioning to 40nm now, so Atom SoCs that are made there won't really be that far behind those made at Intel, if at all.

Actually if you read the rest of the article, there's a deeper reason for this. Historically chips for something like a cellphone take an ARM core and some custom peripherals, integrate them onto a chip and then fab them at somewhere like TSMC. Intel has never done this - they selll chips not IP. In fact one of the reasons the XBox360 moved to PPC was because Intel would not license their core as IP to be integrated into an ASIC. Intel Atoms on a TSMC process would be cheaper, but the real benefit would be (as Anandtech put it)

The Lincroft and Langwell blocks are done by Intel. The PMIC and Evans Peak blocks are partly Intel and partly 3rd party IP that are intermixed. Evans Peak in particular looks like it's going to be home to all sorts of IP depending on the application. A smart phone Atom SoC design might integrate a 3G modem here, while an iPod would opt for something else.

This makes sense if Atom is supposed to be competing with ARM. Maybe in the future they will sell Atoms as a hard macro like Arm do.

"And the jobs never did come back to America- it took 7 years for the number of jobs to be EQUAL to that before the crash, and during that time we imported just about as many workers as we gained jobs."

I don't think you can show causation here. Outsourcing had already begun; it wasn't caused by the "dot-com bubble".

I did look at the second link, which by the way was not a link to a story, but a link to a link to a story, which is very bad form. Regardless, quote: "Whose last big processor, the Nehalem, was designed and built in Bangalore".

Completely false. The Dunnington core, NOT Nehalem, was designed in Bangalore, and was the first and only chip so far to be designed there. Further, Dunnington is nothing more than a multi-core Penryn, which does not say very much about the Bangalore team: http://www.anandtech.com/printarticle.aspx?i=3264

Also, it was not BUILT there, but only designed there. And further yet, that has almost nothing to do with software development, which is the topic under discussion.

And lastly, if you had been involved with any of the international software freelancing sites, you would know that places like Pakistan, India, and any number of third-world countries are full of software development firms and individuals who are, according to their own horn-blowing, experts in virtually everything. You would also know that although they very often under-bid other developers, their reputation is terrible. More and more I have been seeing job offers on the international sites for software developers from "North America, South America, and Western Europe ONLY!"

Sure, outsourcing is bad for America. But the sitation is not as bad as it might appear... especially for software developers.

Face it, Vista got a bad name for three reasons:

1. The lowest-end computers certified to run it were not really capable (since fixed).

Microsoft ran the certification program that certified those low end computers as being capable of running vista. This was under Microsoft's control.

2. Nvidia's drivers sucked for the first 6 months.

While Nvidia's drivers sucking is not under Microsoft's direct control, the certification program that signs the drivers for use in Vista is. Were those drivers signed?
I will agree that the signing of the drivers doesn't necessarily mean that they don't suck, just that they wont harm your system; so in that way this one really shouldn't be Microsoft's responsibility as long as the drivers weren't actually destructive.

3. The I/O subsystem was poorly designed (fixed in SP1), and the virtualization of video memory was a poor idea for Vista-32 that makes game memory usage balloon (hence the higher memory requirements for games under Vista, and problems running out of memory that players don't see on XP). REALITY: Vista should have pushed 64-bit as the primary OS.

clearly Microsoft's fault.

Only one of the above was really under Microsoft's control.

Two of them. Why do you think the first one is not Microsoft's fault?

I also don't agree that these are the only reasons Vista got a bad name, but I'm leaving that part alone.

These were bottom-rung machines bought by people who didn't give a shit. All they wanted was a computar thingy to access thar intarwebs.

When people like this walk into these stores to buy the cheapest computer they can, nothing can move that mountain. You can tell them time and time again that the performance will suck, that it won't work with newer operating systems, and they still won't pony up another dollar.

Face it, Vista got a bad name for three reasons:

1. The lowest-end computers certified to run it were not really capable (since fixed).

2. Nvidia's drivers sucked for the first 6 months.

3. The I/O subsystem was poorly designed (fixed in SP1), and the virtualization of video memory was a poor idea for Vista-32 that makes game memory usage balloon (hence the higher memory requirements for games under Vista, and problems running out of memory that players don't see on XP). REALITY: Vista should have pushed 64-bit as the primary OS.

Only one of the above was really under Microsoft's control.

Not all SSDs are equal. Why you should pay US$400 for the Intel X25-M if you can get another for under US$100? Check this AnandTech review, that spent a lot of time bashing JMicron JMF602 based SSDs.

I forgot to specify in my original post: it was a GF4-MX440, which was just a GF2 with an upgraded memory controller.

All I wanted was a card that would have a draw distance of more than 10 feet in Half-Life 2, and the FX5200 did the job.

I have no idea how your Celeron compares to your Atom, I'm just saying that it's possible that the Atom was designed for low-power in-order execution. The Celeron is low-power too, but it's an out-of-order core. If we look at current processors, a 1.6GHz Celeron will beat a 1.6GHz Atom in many benchmarks because of the in-order/out-of-order differences, but the Atom would use less power. I don't know how the generational differences would affect that.

As for the graphics cards, I'm saying there is a software difference. If you run a Half-Life 2 on a GF4 (let me clarify and say I'm talking about the GF4 MX), then it will run in DX7 mode, with none of the DX8 features. If you run an FX series card, then it will run in DX8.1 mode by default, and the 6-series will run in DX9 mode. You didn't specify what model Radeon your laptop runs, but look at the difference in work your card has to compute between DX7 mode and DX8 mode:
http://www.anandtech.com/video/showdoc.aspx?i=2281&p=9

Moving from the GF4MX to the FX5200, my average framerate dropped a bit, but the image quality was so much better. You can change between the DirectX modes by hand, but unless you do that, you may get reduced performance, despite an increase in image quality.

I thought Anandtech's recent buyer's guides did an excellent job of showing the kind of performance hardware you can get for cheap these days:

http://www.anandtech.com/guides/showdoc.aspx?i=3486

The Apollo computers only had to cope with up to a few thousand kilobits per second of telemetry data and the like. Decoding a high definition YouTube stream means converting a few million bits per second of h.264 video into a 720p30 video stream (which is about 884 million bits per second).

Given that h.264 video is enormously more complicated to decode than telemetry data, and that the volume of it is at least several thousand times greater, I would be outright surprised if web browsing required ONLY 10000 times as much CPU power as the Apollo landers.

But, to be honest, the chipsets are just as likely to come with dedicated video decoding hardware than can handle HD H.264 without breaking a sweat. Take a look at the Atom's Poulsbo chipset for example.

Additional disclaimer: I'm not a CPU engineer, and this is still based on things I read on public websites.

I can't find the article, but Anandtech explained this well. Apparently the high-k+ process that's used in 45nm and smaller Intel chips make for incredibly low leakage currents.

I did, however, find a graph that shows total system power consumption moving from 65nm (Conroe) to 45nm (Penryn), at the same clock speed: http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3137&p=6

AMDs chipsets provide better integrated graphics than Intel do however. And I guess I would go for someone better though still crappy graphics when somewhat faster / more power efficient CPU (if Intel really is.)

Also Intel notebook with Nvidia chipset may compare better to AMD.

Just to make things more clear, desktop Intel-CPU motherboards with NVIDIA's latest GeForce 9300/9400 chipsets (with integrated graphics) are also available.

So if you want to use a Core 2 Duo/Quad and the best NVIDIA integrated graphics (with about half the 3D performance of a $35 discrete graphics card), then that option is also available.

You could browse around here, but honestly I didn't have a hard time... but it was like 2 years ago now. The important thing is to search for an IPS or (P)VA panel and stay away from the TN stuff... those are the ones that change the most when you change angle. This site lets you know what kind of panel a monitor has. I ended up settling on an Acer AL2051W with an P-MVA panel. It is significantly better than a TN screen for viewing angle, but isn't as good as an IPS in that regard. Also it has a glossy screen which drives some people nuts... me too, sometimes!

Here's an awesome rundown at anandtech.

Some links for you:
Dells get a mention
Some discussion about the $$$ Apple monitors

Whatever you do, don't give in and buy cheap :) My wife (who only does office stuff) has a cheap TN panel and honestly, it hurts me to look at it even for web browsing :)

Do they have some magically awesome chipset that has never graced the consumer market?

Yes, Larrabee. It's a massively-multicored x86 processor designed to act as a GPU (it has some fixed-function GPU stuff tacked on).

In effect, Intel intends to build a GPU powerful enough to get software rendering (and all the flexibility and power that brings) up to the same speed as hardware-accelerated rendering. Intel is also going to be providing OpenGL/Direct3D abstraction layers so that existing games can work.

Larrabee is expected to at least be competitive with nVidia/AMD's stuff, although it might not be until the second generation product before they're on equal footing.

The response times are always cherry picked from the absolute best circumstances the panel can manage, so you should take it with a grain of salt to begin with. It's all but meaningless.

Take the Syncmaster 2493HM, with a stated response time of 5ms. You might think it can update the screen 200 times completely each second with a figure like that, but no: Here's an image of its ghosting.

The monitor takes input at 60hz, so it has 16.66ms to update the panel completely each cycle. Obviously it can't do that since you can see two images clearly, which means it takes at least 33.33ms to update.

You're now thinking, "Can you even notice it though?" I have the monitor, and yes you can. It's plainly visible sometimes. The most noticable thing is when you have scrolling high-contrast elements on the screen, such as in a game.

According to Anandtech, currently Creative still has the best game compatibility, because the game devs write to their cards, but Asus' Xonar line has better sound quality, and nearly the same level of game compatibility. I know if I were to build a new machine I'd take their advice on that, what with Creative's driver troubles, especially on x86-64.

http://anandtech.com/guides/showdoc.aspx?i=3497&p=5

Based on Valve's stats, it looks like only about 3.5% of Steam users have an X-Fi card. I do know of a large portion of people who were weary of the X-Fi series though, and kept buying Audigys, and people like me who kept their "Creative Live!" cards, which are likely a good portion of that 33% with "other" sound devices.

GP is right though; most people are perfectly happy with onboard sound. This is especially true in the laptop market, which last I heard was now well over 50% of total computer sales.

But, DirectX 11 will be supported on Vista too.

http://www.anandtech.com/video/showdoc.aspx?i=3507

"To be fair, the OS upgrade requirement also threw a wrench in the gears. That won't be a problem this time, as Vista still sucks but will be getting DX11 support and Windows 7 looks like a better upgrade option for XP users than Vista. Developers who haven't already moved from DX9 may well skip DX10 altogether in favor of DX11 depending on the predicted ship dates of their titles, all signs point to DX11 as setting the time frame we start to see the revolution promised with the move to DX10 take place. Developers have had time to familiarize themselves with the extended advantages of programmability offered by DX10, coding for DX11 will be much easier though OOP constructs and multithreaded support, and if the features don't entice them, the ability to run on downlevel hardware with a better coding environment might just seal the deal."

With older NVIDIA cards, I have worse performance than on my EeePC (GMA 900).

Your old cards aren't as good as your new ones? shock! horror! Say it isn't so!

The GMA 900 is slower in games than a Geforce 6200, a several generations old budget GPU from nvidia. And anyway, nvidia have recently released a driver update that to a great extent fixed the problem with KDE, which suggests (as was already known) that the performance problem was indeed a driver issue.

But don't let random facts get in the way in your holy crusade against the godless open source zealots.

http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3403&p=4

[quote]Given the 100GB per day x 5 year lifespan of Intel's MLC SSDs, there's no cause for concern from a data reliability perspective for the desktop/notebook usage case. High load transactional database servers could easily outlast the lifespan of MLC flash and that's where SLC is really aimed at. These days the MLC vs. SLC debate is more about performance, but as you'll soon see - Intel has redefined what to expect from an MLC drive.[/quote]

So.. 100GB/day/drive = 100GBx4 = 400GB/day for (5*365)= 400GB*1825(days) = 730TB of data transferred. If you seriously go through THAT much data, you are either a pirate or you actually own your own movie studio.

Obviously these numbers can change, based on case temperature, wear leveling, etc. HOWEVER, they are what Intel states the drives can handle...

Read this:

http://www.anandtech.com/mb/showdoc.aspx?i=3471

You can now buy a PC perfectly capable of playing TF2 on high settings for around $400-500, not including the OS or monitor.

http://www.anandtech.com/guides/showdoc.aspx?i=3486&p=3
Take that machine, and spend $100 on a video card, and you have a gaming PC. Now, if you spend a bit more for the "budget" system they have listed, then you'll have a complete, balanced computer. But spending "thousands of dollars" is no longer necessary for gaming; it's frivolous spending.

5: unused RAM is wasted RAM. So long as it frees up the RAM when a high priority application needs it, using spare RAM for caching can have huge benefits. Don't trot out the power usage argument. The difference in power between half full ram and full ram is miniscule

The only problem with this is, the OS isn't freeing-up ram that other applications could use. This is because Vista reserves a memory space proportional to the size of your video ram. This is in-addition to the normal PCIe memory window, and wastes much more memory than XP.

Why exactly does Vista need to allocate a separate memory space for video when Aero is already disabled (full-screen gaming)? I hope to hell they fix this stupidity before Windows 7 ships. It's already causing a lot of confusion, as recent game releases have had MUCH higher memory requirements for Vista than XP.

Well, you're not looking very hard then.

Here's Anandtech's review of the ATI 4870x2, the closest competitor to the GTX295 (which they haven't reviewed yet):

http://www.anandtech.com/video/showdoc.aspx?i=3372&p=6

Note how they test from 1280x1024 and up. While widescreen displays might be 1280x800, I think that's a pretty decent resolution for showing the "consumer capabilities" of a card.

From what I've read, the current Phenom II's are limited to DDR2 and AM2+; the AM2+/AM3 DDR2/DDR3 supporting ones are due next month.

Not even games require a great CPU, anymore. Some of the most innovative ones(Like World of Goo or Left4Dead) will run on old systems, like a 2ghz Athlon XP from 2002/2003. :P Faster CPUs are for... bragging rights? Encoding stuff?

More memory is for multitasking - you can never have enough memory! I'm amazed that I can hit 1GB memory usage when my OS and background software only consumes ~150MB. A year ago that wasn't the case, but now I just have more stuff running...

I think the next must have computer hardware will be high performance SSDs. That'll boost responsiveness more than a faster CPU can. People will become accustomed to their games loading instantly, and not having lag spikes. It's already been proven that a high end SSD can take the min FPS in a game like Crysis from the low 10's to the high 20's - that's more than the jump from Dual-core to Quad-core, or 8800GTX to GTX260.

http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3403&p=14

We already have an abundance of CPU power. The next jumps will definitely be in IO/storage performance, power consumption, and size.

It makes me wonder if ARM can wedge its way into the netbook market - power consumption is not a strength of x86.

the moment I saw the whopping 130w listed next to the i7 920 I immediately decided I didn't want one.

And don't forget -- this is an Intel power rating, which means "typical use". I have a computer with an AMD Phenom 9850, and that's rated at 120W, but that's worst case, not typical. In actual use the computer has been quiet and cool. (I ought to try recompiling the Linux kernel on all four cores or something.)

http://www.silentpcreview.com/article169-page3.html

AMD could corner the efficiency market.

Intel is selling far more CPUs than AMD can even make, and they have the top absolute performance, but AMD is selling plenty of Opterons into data centers because AMD does well on performance/power ratio. AMD is also selling some CPU chips with a maximum heat dissipation of 45W. AMD is already working on cornering the efficiency market.

http://enterprise.amd.com/us-en/AMD-Business/Technology-Home/Power-Management.aspx
http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=3003

steveha