Hey guys, it's been a month and a half and I still don't see the Mac Pro CPU upgrade article. I want to do one of these mods myself, and I'm waiting for the full article to come out before I plunk down the big bucks.
If you're interested in the possibility of upgrading an i7 or Xeon Mac Pro check out this discussion thread and benchmark result to see what Anand's blog has inspired:
Please do change over! For those of us that waited around on XP, and made the jump to w7, it is much more helpful to have reviews in w7 than it is in Vista. Plus, testing the latest hardware on the latest software helps us see where the real topend results are :-)
Hey, it occurred to me that the massive 8-way 8-core nehalem-EX platform is actually connected in a cube! Each vertex is a processor, each edge is a QPI link. It's the most symmetrical way to make an 8-way system using processors with 3 links (the 4th connects to the I/O hubs).
Thinking of a cube is easier than understanding the 2D-projection, I think... And it's cool! Too bad that system will probably be insanely expensive. I'm a physicist and I have a bunch of colleagues that would love having such a system. Many of them actually have massively parallel software!
[Quote]How much would you guys hate me if I switched to Windows 7 sooner rather than later?[/quote]
Exactly the opposite, the only thing I'd want to see is Windows 7 used, and perhaps XP so show any differences with SSD performance. There seems to be a bit of an XP or W7 only mentality, but Vista's performance would be easy to ballpark from just using Window 7's results. I was not even aware Vista supports the Trim command, but was of the opinion XP does not?
As regarding the CPU selection for the next influx of old CPU to be included for the the Anandtech bench, I have heard you mention about Pentium 4's, Athlon XP but why not Athlon 64's of socket 939? It is one of the greatest processors of its time and running on single core and DDR1, we would like to see how it fare against the current line up of new processors.
Personally, for those of us who remember all the old CPU's dating back 10-15 years, it just simply helps to know how the Atom compares to older cpu XYZ. Is it like a PIII for example in performance? Because once we know how it performs in relation to an older CPU, everybody "in the know" can extrapolate how it'll perform compared to something modern.
Great Stuff there Anand, looking forward to your next few articles. I like to see how my old BP6 with dual Celerons would fair against the newer beasts of today. Other than that keep the updates coming on the SSD front, eventually will make the switch but for now Raptor is best.
I read about the athlon xp and p4 etc. Don't forget the dual socket systems out there. We have a lot of them, many users went to them when core 2 duo and quads meant severe upgrades of everything. Please add the dual socket workstation class cpu's and mothierboards you have on hand to the bench. If you need help with some of them I am sure some readers out there would be glad to lend a hand
Switch to win7 as soon as possible. Lots of folks are running winXP on new hardware but are so distrustful of upgrading that they're not getting everything out of their hardware that they could. Nobody trusts anyone who says that vista works fine, so nobody uses vista if they don't have to.
But if you switch to win7 and it works out good for you, then I bet lots of people will finally upgrade from winXP. I'm running win7 beta on my laptop and if MS doesn't price win7 out of my reach, I have 4 computers that will need to be upgraded when it's released, assuming I decide I can trust MS again.
Anand, common! how could we ever hate you? we love all your articles! Go for 'pioneering' and do your benches on win7, we will all be using it within the 'reasonable' future anyway.
Also, please please please! feed us more info on SSD's. I keep checking your 'Storage' page DAILY for that update you promised in your last article from march(some secretive stuff from Intel).
I was thinking about getting an intel disk but keep stalling due to uncertainties about when the next gen. SSD's will be around the corner and how much an upgrade/price reduction they will bring to the table.
...What I'd like to see is some older GPUs used as well, particulary the 8800GTX. It was ages now since I saw a single review at anandtech that featured 8800GTXes, SLI'd or otherwise, and I'd love to see how they stack up against today's offerings.
Now I get to see how my i7 stacks up against my ancient barton xp 2500
Just make sure that old motherboard - it better be nforce2 based, has good caps.
WINDOWS 7 RC is already my main OS... it's a whole lot easier to live with than Vista. The difference is very palpable, especially with networked Win 7 machines, but especially with the 'smarter' file Explorer and better PC resources management.
I'm very much looking forward to your newest SSD test results, especially the OCZ Vertex EX SLC drive. I think it should be a worthy contender to the Intel solution.
And honestly, I'd love it if you switched over to Windows 7 as soon as possible. Windows 7 works FABulously with solid state drives, intelligently disabling performance tweaks meant for traditional, rotating drives, and aligning the drive properly on initial format. From what I've read, the RC currently sends the trim command, but I don't know if any drives have firmware to support it yet. I'm excited about the new OCZ Vertex firmware, which will support trim. Until then we're all stuck with Wiper.exe. But really, Windows 7 and SSDs just go hand in hand. I installed 7 the night I got my Vertex 60GB, and the experience has been flawless.
SSD's rock. Used them since their early appearance on ebay and was ammazed by the speed of my freshly installed system.
The older ones are too slow to use anymore but still enjoy the speed of the OCZ core and solid.
The Vertex is a must and i will get one when i get home in 2 months.
Hope you have your SSD update done by then.
Thanks for the effort you put in your SDD stories Anand, have read them all.
Big thanks, gr Morbid
They technically are mobile chips, but it'd be great if you could bench some original Yonah Core Duo chips. They were sold as mobile on desktop applications and had early use as media boxes so they have a relevant comparison point for desktops besides being the origins of the Merom and Nehalem micro-architectures.
And while you are testing the Nehalem Mac Pro, it'd be great if you could add those numbers to your charts as well. If you have older Mac Pros, it'd be great to see how dual Clovertowns, dual Harpertowns, and dual Gainestown compare to Penryn and Bloomfield.
I'll put in a plug again for the 3.73GHz Pentium Extreme Edition 965 dual core Presler as well.
First time I've been even mildly intersted in an Apple product since the Lisa. It's awesome that you will be reviewing it with your processor upgrade article!
I'm hoping this leads to a next-gen dual CPU mainboard market; come on Asus, EVGA, if you're going to charge us $400 for a mainboard let's get some dual-CPU action too!
Win7: Yes, I installed the RC1 just to test it, but almost never utilized back the XP.
And no. Is not the same performance that Vista. Vista did NOT worked. Win7 works.
SSD: what matters on the end of the day, is how much performance I can buy on a budget.
I can buy a fast 120 Gb OCZ Vertex. but would consider also investing on those options at the same cost:
1-120 GB OCZ Vertex.
2-Same cost on RAID0 of smallest (and cheaper) SSD disks.
3-JBOD RAID of 30/60 Gb SSD and 1Tb HDD, with a good defragmenter software.
4-Same cost on 4 (or maybe 8) 7200rpm 1Tb Hdd, using only the first 120 Gb. since it would use only the fastest tracks, it would have an amazing read /write speed, and the heads would no move, so the latency (and noise) would be reduced at the minimun, and the caché will add (4x1Tbx32Mb would have 120Mb of caché).
I understand where people are coming from when they say use Win 7 for SSDs but I don't think you should until it's gone gold. Regardless of how good the RC is or how many people have no problems it's still not a retail OS and shouldn't be used exclusively for testing right now.
I read the Windows 7 blog on SSD and they made a bunch of changes to optimize the OS, so saying that Vista is no different than 7 is not true.
I say you should move to 7 for testing SSD. I am also curious to know which SSDs currently on the market are properly communicating with Windows 7 at first install and telling the OS that the drive is a SSD and not a HDD so 7 properly optimizes.
Also, what is the status on TRIM for Intel SSDs? Is it true that Intel has been working on their own SSD defrag utility??? (possibly just a rumor considering most believe defrag is not needed).
I've been using 7 for just about 6 months and it's been a great experience. Really brings some love to Microsoft.
Since there's no difference between Vista and 7 I say go ahead an switch. The SSD review will be much more real that way.
Keep in mind that 7 could change until the RTM (not very likely but possible).
Vista was fun for the past 2.5 years but now we can let it die just like XP should have and let 7 take the reigns :)
Thanks for all your hard work, love the articles. Your SSD reviews are like no others.
Instead of getting rid of the "shorter bar = better" benchmarks (they are useful after all), why not make a simple interface change and keep them? I have two thought on how this could be done.
1) Move all the "shorter = better" benchmarks to the bottom of the page, and keep the "larger = better" ones at the top. That way you're not switching back and forth as you read through the benchmarks.
2) On "shorter = better" benchmarks, have the bars originate on the right side of the graphic instead of the left side. This way the shorter bars would still be better, but now there would be a consistent "farther to the right = better" that might be helpful in avoiding confusion.
How about reparamaterizing the shorter is better bars to something that works as longer is better. One way to do this would be to consider the absolute value of the difference between every piece of hardware and the worst performer. The worst performer would be equal to zero.
There is probably an infinite number of other ways to make something like that work, and although the information's value would be diminished because it is only based on the results (and doesn't truly depict the actual results), but it would be valid in comparing relative performance. If something like this is done, I think it'd be important to keep the actual results viewable to ensure that there is no bias (I love you guys, but there are always jerks that are like "wow way to give AMNvidintel an advantage to make them look better")
except that you're trying to get a relative idea of the performance of the CPUs
if 1 cpu is twice as fast as another, that should be readily apparent without having to judge what the length of a bar means in different sections of the graph
so it might only be 1 second on this task of this size, but it could be 100 seconds on a different or larger task
just for example, imagine a comparison of cpus from 286 to Nehalem that measured how long it took pkzip to compress 20MB
an athlon XP would be indistinguishable from an i7 920
in reality the 920 trounces the athlon XP, but the legend is so compressed it becomes hard to tell
- win 7 : GO for it. i've just purchased a GSKILL falcon SSD (brother or OZC vertex), and done lots of benchmarks for photoshop. i compared XP64 and win7 64, win7 is much more efficient. so SSD + win7 are the winners! on my old setup, vista is much slower and problematic than win7.
- RAID SSD : as many readers, i would like to see the performance difference between OCZ vertex SSD :
2 RAIDed 30GB, 2 RAIDed 60GB, single 60GB, single 128GB.
because 2x60GB cost about the same as 1x128GB.
Go ahead to Vista 2.0, errr, I mean Win7. I've been dual-booting it and it feels ready for launch!
One request: Can you test a G.Skill Falcon 128Gb SSD. Price/Gb looks pretty good, and they're quoting some impressive performance #'s. I'd like to see you do your REAL world testing on it! PLEASE!
I think the article should include xp & win 7. Vista yuk...
With explanation with REALLY how 'worn down' does a mlc drive become with heavy usage. I have a vertex & its fast w reads & writes. But I keep hearing these stories on them 'not lasting' ? How long do they last? compared against SLC? do the storage cells wear out???
Windows 7? Go for it, Anand. I think many of your readers are already there, and many more are considering it. Best OS they've ever released, even as an RC.
if other benchmarks have shown no tangible difference between 7 and Vista, by all means, go with 7. we're here for the relative performance anyhow, not the exact cloning of your rig for exact times.
on older pc's. to me, the base of the world is a p3-500. the old gateway/compaq cast iron horses still run, still browse the net, and with a new harddrive are zippy enough. i see tons of cel-1200 tualatins that effectively match the p4-1.3-1.5's to cover the 'painfully almost' catagory. and consider a p4-2.4 just because there really are that many dell dim2400's still out there.
on the ssd's: with intel slc 32gb ~$400usd, 64gb ~$780usd, i only lost $20 going with raid0 32gb's on an ich10r. hdtach3 296MB/sec avg, 496MB/sec burst. since so many of the ssd's are ~2x the cost as you go up in size, PLEASE consider putting raid0 smaller ssd's on benchmarks. 60GB is enough to survive on a boot drive. two 32gb supertalent's for ~$270 total should plausibly destroy an intel slc x25-e 64gb at $500 cheaper, and 3.5" brackets mounting two 2.5" drives make it all too convenient.
on pivotal boot and app load tests, consider distinguishing "with same drive swap file", "with other drive swap file", and "with zero swap file".
1 more vote for switch now. I'm using it exclusively, and if there's a difference with respect to SSD performance related to trim support or anything else, I want to know.
I would like to see the P4 2.8E (worst overclocker) in the bench comparison, as well as the Core2 E6600. I jumped on early with the Core2, able to pick one up before they hit the market. I think it's about time to get an upgrade.
The other thing that might be nice to see is the performance w/ water cooling and regular hair cooling (fan or just a spreader).
As for SSDs, I'll be looking forward to your SLC comparison :) as well as the results from the 8core NeX (Nehalem-Ex)
"How much would you guys hate me if I switched to Windows 7 sooner rather than later?"
No hate at all. I'm in the process of switching everything over to Windows 7 RC (my main desktop has been on Windows 7 since the Beta), so Windows 7 results are most important to me.
I'm sure there will be those that disagree, though.
"Although our recent article showed that it’s not really any faster than Vista"
Dammit. I was really hoping that Windows 7 would offer some type of performance increase over Vista, not just better SSD performance. Oh well, I guess I'll take what I can get. I really dislike the overlordishness of mainstream operating systems these days, but if it helps improve overall ease of use then I suppose it's a sacrifice I'm willing to make.
I must admit I am as excited as the next guy when it comes to SSDs, i can't wait for them to become an affordable viable option for a boot drive, nevermind them becoming big enough to use as my main drive (minimum of 1TB currently).
However I am thinking of replacing my aging 500gb HDD with a nice new spangly 1TB drive and have been looking for reviews of said beasties and have been unable to find a review done in the last 12+ months anywhere....
So how about seeing if you can grab the latest and greatest HDDs on the market and giving them a spin to tell us what to get and what to avoid.
Definitely would like to see you move to Windows 7. I've been running the x64 edition on my daily use laptop since Beta 1 and have had no unexpected significant issues with it. Go for it!
* thanks for the status report
* that WePC project looks a great iniciative, let's see what comes out of it
* thanks for letting me know I don't want an SSD drive yet
* that 64 cores nehalem system looks like a 3D artist's wet dream, but I guess the price is still going to be quite a bit higher than that of a comparable render farm based on more standard equipment
Hello Anand.
What about checking the SSD Optimizations made in the Chipset level?
Are the newer chipsets of Intel / AMD / nVidia any better using those drives? Does each of them has plans on making optimized chipset for SSD?
I've been using it as my primary OS since the RC release and it's seems to generally perform better than Vista. I've only had one slight problem and that was getting my Colormunki monitor calibrator working. Win7 wasn't able to find a driver but I found it myself and got it working.
So I say "GO GO GO" in regards to switching to Win7.
OK, I'm already using the RC, but it really is great. Plus, it's not different enough from Vista to really make difference if you go ahead and switch, and if the adoption of the Beta and RC are any indication then Win7 adoption will be out the roof!
People already hate vista so switch to vista 2 by all means.
I would also suggest that the poll system you implemented a while back would be perfect to use if you want to ask a straightforward question like this. I wouldn't have to ramble on in my broken English and I wouldn't have to read every answer to see the results. Just a thought...
After reading this update, really amazed with such a wide scope coverage that you are currently working hard on. Much anticipating with what will coming next from AnandTech.
On the Bench, the main reason is that people want to know how much performance we are getting for our money. I think the recent economy has changes some of our thinking in spending.
The past few years we had experience upgrading parts with not much performance difference. It is interesting to see how previous CPU fair with current generation. And finally realize the days we had improvements like 386 to 486 and Pentium are over.
SSD - The mystery of multiple channel SSD has yet to solved. Is it possible to have more then 10 channel SSD drive in an 2.5" SSD. If not what is the maximum possible channel, i.e what performance should we expect from future SSD.
EX seems to me an destructive product. Previous Benchmarks shows how much performance / watt the Nehalem was able to deliver. May be WoW servers can finally fit more players inside a Single Realm?
The biggest jump, of course, was the 286, arguable the best processor Intel ever made for its time. It was extremely fast, so much so that I couldn't believe it the first time I used it. Going from a PC to a PC/AT was a jump that was mind-boggling. On top of that, it added virtual memory, and the ability to multitask. It was a huge improvement.
The 386 sucked, although it did move to 32-bit and added the now worthless Virtual 86 mode. The performance was poor though, as it ran like a 286 at the same clock speed, clock normalized. But, it moved the platform forward.
The 486 was a nice processor, for sure. It didn't add anything new, but ran like a raped ape. I remember how we fought over the first 486 PC made, an upgraded version of the PS/2 Model 70-A21, which was called the B21. It cut compiles down from 30 minutes to about 18 compared to a 386/25, and it was only running at 25 MHz, and did not have any L2 cache, unlike the 386 which had a 64K external cache.
The Pentium wasn't so great. It ran at poor clock speeds at first, and when I used it, I found the performance quite poor. I used to say Intel's odd numbered processors were all bad, and their even ones good and this was further proof. I went from a 486 at 100 MHz, which were available, to a Pentium at 66 MHz, and was entirely unimpressed. It ran hotter than Hell too. But, in time, they got it right, with shrinks and the Pentium MMX (although MMX wasn't so important, there were other enhancements). The FDIV bug was kind of funny too, and how idiots overreacted to it. Still, I have one of the processors and I hope it will go up in value.
The Pentium Pro was another initial disappointment. It ran 16-bit code very poorly, although the fault lies more with Microsoft than Intel since they gave Intel guidance that 32-bit code would dominate by 1995 when the Pentium Pro was released. It was really expensive too, because of the on-processor L2 cache (which was really on the same packaging, but not part of the processor), which could not be tested before joined with the processor, so if one were bad, they were both thrown out. The Pentium II addressed the segment register issue of the Pentium Pro, so improved 16-bit code somewhat, and obviously what we use today is still based on the Pentium Pro design. It was, of course, and even number (P6) design, so was good.
The Pentium 4, an odd number design, of course, sucked far worse than the Pentium and 386. It was much hotter, like the Pentium compared to the 486, but the performance sucked too, which made it a double threat.
Now we are back to the Pentium Pro design, so it's only reasonable we don't see dramatic changes since it's really not a new processor at all, but just improvements to an existing line, like the Pentium MMX, or Athlon XP (I'd use Prescott as an example too, but was it an improvement?????).
It makes sense though. Really, virtually nothing in PCs is new, and never has been. Virtually every concept has been taken from mainframes, or supercomputers, and microcomputers were just low cost implementations of those concepts. When you had a scarcity of transistors, there were many very good things you could not implement, but you wanted to. That low lying fruit is gone, and adding transistors gives greatly diminished returns. It's like the improvement with any technology. If you look at cars, jets, etc... once the technology matures, the improvements slow down and you get gradual improvements.
I'd like to see x86 die, but I guess it will not happen. Processors cost more, perform worse, and use more power because of this miserable instruction set. It's not a lot per processor, but when you multiply it by the millions upon millions of processors that are afflicted with the x86 instruction set, the costs are staggering. That's not even counting the migraines it has caused by people who have to code in assembly for it. It's gruesome.
Great post, but you forgot the first Celeron, the 300a that overclocked to 450MHz without breaking a sweat. In my book, the best bang-for-the-buck Intel CPU right up through the Core2Duo.
This was in the early days of overclocking, just over a decade ago, and it was a dream to be able to run a chip at 150% its rated speed with zero changes to things like voltages (heck most motherboards didn't even support voltage changes back then, iirc).
I agree with what you said about the economy changing things. It isn't just that though. Computer technology isn't gaining nearly as quickly, and the need for new systems - at least in the consumer space is less pressing. The dollar is far more important now than it was. Purchases even at the business level are determined even more by cost than in the 90s, when changes to performance was more heavily weighed factor in IT decisions.
This is why businesses that want to survive have to be broadly invested. For example, if a company that manufacturers hard disk drives exclusively fails to get going on SSD early enough, it will be facing extinction.
If you're GM, it pays to invest in something other than just big gas guzzling automobiles. :-D Let's hope the Chevy Volt has some real value. I have my doubts that it will be as popular as they'd hoped, but I do think the car merits an award for ingenuity. They took conventional thinking and flipped it on its head, I just wish they'd done it 20 years ago. With their marketing dept they might've been able to put Ford out of business.
The trade off is a more expensive motherboard, though, that has buffer memory. It wasn't clear to me where this memory would be, separate ICs soldered to the mb or integrated into the chipset. My guess is the former, at least to begin with, in which case RAM manufacturers won't hate the change nearly as much.
Instead of subsidizing FB-DIMM's AMB chip on each DIMM sold because of extra power consumed, now Intel can charge for the SMB chips individually(8 socket system would require a whooping 32 SMB chips).
Depending on how much Intel wants to price those SMB chips, it could make or break Nehalem-EX's platform. (A simple $10 per SMB would easily mount to 320 dollars cost to the motherboard manufacturers, which could then turn into a 600 dollars+ premium on the motherboard)
This approach is indeed the correct one from a technical perspective. Basically Intel wants all the memory premium associated with Nehalem-EX platform by forcing the memory makers into producing commodity DDR3 DIMMs, not FB-DIMM2s.
If they are of a similar size/cost then obviously it's fair to compare them as people looking for a drive of that size/cost will be able to pick either.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
69 Comments
Back to Article
AZREOspecialist - Saturday, July 11, 2009 - link
Hey guys, it's been a month and a half and I still don't see the Mac Pro CPU upgrade article. I want to do one of these mods myself, and I'm waiting for the full article to come out before I plunk down the big bucks.Any idea when this will be posted?
Tutor - Friday, June 26, 2009 - link
If you're interested in the possibility of upgrading an i7 or Xeon Mac Pro check out this discussion thread and benchmark result to see what Anand's blog has inspired:http://forums.macrumors.com/showthread.php?t=73073...">http://forums.macrumors.com/showthread.php?t=73073...
http://browse.geekbench.ca/geekbench2/view/143750">http://browse.geekbench.ca/geekbench2/view/143750
Tutor - Saturday, June 27, 2009 - link
Update: Thread is at:http://forums.macrumors.com/showthread.php?t=71393...">http://forums.macrumors.com/showthread.php?t=71393...
jeramhyde - Wednesday, June 3, 2009 - link
re: Windows 7Please do change over! For those of us that waited around on XP, and made the jump to w7, it is much more helpful to have reviews in w7 than it is in Vista. Plus, testing the latest hardware on the latest software helps us see where the real topend results are :-)
So a big +1 from me!
ddobrigk - Saturday, May 30, 2009 - link
Hey, it occurred to me that the massive 8-way 8-core nehalem-EX platform is actually connected in a cube! Each vertex is a processor, each edge is a QPI link. It's the most symmetrical way to make an 8-way system using processors with 3 links (the 4th connects to the I/O hubs).Thinking of a cube is easier than understanding the 2D-projection, I think... And it's cool! Too bad that system will probably be insanely expensive. I'm a physicist and I have a bunch of colleagues that would love having such a system. Many of them actually have massively parallel software!
Kougar - Friday, May 29, 2009 - link
[Quote]How much would you guys hate me if I switched to Windows 7 sooner rather than later?[/quote]Exactly the opposite, the only thing I'd want to see is Windows 7 used, and perhaps XP so show any differences with SSD performance. There seems to be a bit of an XP or W7 only mentality, but Vista's performance would be easy to ballpark from just using Window 7's results. I was not even aware Vista supports the Trim command, but was of the opinion XP does not?
Seramics - Friday, May 29, 2009 - link
As regarding the CPU selection for the next influx of old CPU to be included for the the Anandtech bench, I have heard you mention about Pentium 4's, Athlon XP but why not Athlon 64's of socket 939? It is one of the greatest processors of its time and running on single core and DDR1, we would like to see how it fare against the current line up of new processors.valnar - Thursday, May 28, 2009 - link
Personally, for those of us who remember all the old CPU's dating back 10-15 years, it just simply helps to know how the Atom compares to older cpu XYZ. Is it like a PIII for example in performance? Because once we know how it performs in relation to an older CPU, everybody "in the know" can extrapolate how it'll perform compared to something modern.plummerb - Thursday, May 28, 2009 - link
Great Stuff there Anand, looking forward to your next few articles. I like to see how my old BP6 with dual Celerons would fair against the newer beasts of today. Other than that keep the updates coming on the SSD front, eventually will make the switch but for now Raptor is best.abnderby - Thursday, May 28, 2009 - link
I read about the athlon xp and p4 etc. Don't forget the dual socket systems out there. We have a lot of them, many users went to them when core 2 duo and quads meant severe upgrades of everything. Please add the dual socket workstation class cpu's and mothierboards you have on hand to the bench. If you need help with some of them I am sure some readers out there would be glad to lend a handflensr - Thursday, May 28, 2009 - link
Anand,Switch to win7 as soon as possible. Lots of folks are running winXP on new hardware but are so distrustful of upgrading that they're not getting everything out of their hardware that they could. Nobody trusts anyone who says that vista works fine, so nobody uses vista if they don't have to.
But if you switch to win7 and it works out good for you, then I bet lots of people will finally upgrade from winXP. I'm running win7 beta on my laptop and if MS doesn't price win7 out of my reach, I have 4 computers that will need to be upgraded when it's released, assuming I decide I can trust MS again.
danielk - Thursday, May 28, 2009 - link
Anand, common! how could we ever hate you? we love all your articles! Go for 'pioneering' and do your benches on win7, we will all be using it within the 'reasonable' future anyway.Also, please please please! feed us more info on SSD's. I keep checking your 'Storage' page DAILY for that update you promised in your last article from march(some secretive stuff from Intel).
I was thinking about getting an intel disk but keep stalling due to uncertainties about when the next gen. SSD's will be around the corner and how much an upgrade/price reduction they will bring to the table.
FaaR - Thursday, May 28, 2009 - link
...What I'd like to see is some older GPUs used as well, particulary the 8800GTX. It was ages now since I saw a single review at anandtech that featured 8800GTXes, SLI'd or otherwise, and I'd love to see how they stack up against today's offerings.Randomblame - Thursday, May 28, 2009 - link
Now I get to see how my i7 stacks up against my ancient barton xp 2500Just make sure that old motherboard - it better be nforce2 based, has good caps.
dEad0r - Thursday, May 28, 2009 - link
Windows 7 ftw!evilsopure - Thursday, May 28, 2009 - link
WINDOWS 7 RC is already my main OS... it's a whole lot easier to live with than Vista. The difference is very palpable, especially with networked Win 7 machines, but especially with the 'smarter' file Explorer and better PC resources management.Ikonomi - Thursday, May 28, 2009 - link
I'm very much looking forward to your newest SSD test results, especially the OCZ Vertex EX SLC drive. I think it should be a worthy contender to the Intel solution.And honestly, I'd love it if you switched over to Windows 7 as soon as possible. Windows 7 works FABulously with solid state drives, intelligently disabling performance tweaks meant for traditional, rotating drives, and aligning the drive properly on initial format. From what I've read, the RC currently sends the trim command, but I don't know if any drives have firmware to support it yet. I'm excited about the new OCZ Vertex firmware, which will support trim. Until then we're all stuck with Wiper.exe. But really, Windows 7 and SSDs just go hand in hand. I installed 7 the night I got my Vertex 60GB, and the experience has been flawless.
Morbidrod - Wednesday, May 27, 2009 - link
SSD's rock. Used them since their early appearance on ebay and was ammazed by the speed of my freshly installed system.The older ones are too slow to use anymore but still enjoy the speed of the OCZ core and solid.
The Vertex is a must and i will get one when i get home in 2 months.
Hope you have your SSD update done by then.
Thanks for the effort you put in your SDD stories Anand, have read them all.
Big thanks, gr Morbid
ltcommanderdata - Wednesday, May 27, 2009 - link
They technically are mobile chips, but it'd be great if you could bench some original Yonah Core Duo chips. They were sold as mobile on desktop applications and had early use as media boxes so they have a relevant comparison point for desktops besides being the origins of the Merom and Nehalem micro-architectures.And while you are testing the Nehalem Mac Pro, it'd be great if you could add those numbers to your charts as well. If you have older Mac Pros, it'd be great to see how dual Clovertowns, dual Harpertowns, and dual Gainestown compare to Penryn and Bloomfield.
I'll put in a plug again for the 3.73GHz Pentium Extreme Edition 965 dual core Presler as well.
Sabresiberian - Wednesday, May 27, 2009 - link
First time I've been even mildly intersted in an Apple product since the Lisa. It's awesome that you will be reviewing it with your processor upgrade article!I'm hoping this leads to a next-gen dual CPU mainboard market; come on Asus, EVGA, if you're going to charge us $400 for a mainboard let's get some dual-CPU action too!
:D
marraco - Wednesday, May 27, 2009 - link
Win7: Yes, I installed the RC1 just to test it, but almost never utilized back the XP.And no. Is not the same performance that Vista. Vista did NOT worked. Win7 works.
SSD: what matters on the end of the day, is how much performance I can buy on a budget.
I can buy a fast 120 Gb OCZ Vertex. but would consider also investing on those options at the same cost:
1-120 GB OCZ Vertex.
2-Same cost on RAID0 of smallest (and cheaper) SSD disks.
3-JBOD RAID of 30/60 Gb SSD and 1Tb HDD, with a good defragmenter software.
4-Same cost on 4 (or maybe 8) 7200rpm 1Tb Hdd, using only the first 120 Gb. since it would use only the fastest tracks, it would have an amazing read /write speed, and the heads would no move, so the latency (and noise) would be reduced at the minimun, and the caché will add (4x1Tbx32Mb would have 120Mb of caché).
MadMan007 - Wednesday, May 27, 2009 - link
I understand where people are coming from when they say use Win 7 for SSDs but I don't think you should until it's gone gold. Regardless of how good the RC is or how many people have no problems it's still not a retail OS and shouldn't be used exclusively for testing right now.Zorlac - Wednesday, May 27, 2009 - link
I read the Windows 7 blog on SSD and they made a bunch of changes to optimize the OS, so saying that Vista is no different than 7 is not true.I say you should move to 7 for testing SSD. I am also curious to know which SSDs currently on the market are properly communicating with Windows 7 at first install and telling the OS that the drive is a SSD and not a HDD so 7 properly optimizes.
Also, what is the status on TRIM for Intel SSDs? Is it true that Intel has been working on their own SSD defrag utility??? (possibly just a rumor considering most believe defrag is not needed).
Nihility - Wednesday, May 27, 2009 - link
I've been using 7 for just about 6 months and it's been a great experience. Really brings some love to Microsoft.Since there's no difference between Vista and 7 I say go ahead an switch. The SSD review will be much more real that way.
Keep in mind that 7 could change until the RTM (not very likely but possible).
Vista was fun for the past 2.5 years but now we can let it die just like XP should have and let 7 take the reigns :)
Thanks for all your hard work, love the articles. Your SSD reviews are like no others.
slashbinslashbash - Wednesday, May 27, 2009 - link
Instead of getting rid of the "shorter bar = better" benchmarks (they are useful after all), why not make a simple interface change and keep them? I have two thought on how this could be done.1) Move all the "shorter = better" benchmarks to the bottom of the page, and keep the "larger = better" ones at the top. That way you're not switching back and forth as you read through the benchmarks.
2) On "shorter = better" benchmarks, have the bars originate on the right side of the graphic instead of the left side. This way the shorter bars would still be better, but now there would be a consistent "farther to the right = better" that might be helpful in avoiding confusion.
uuwuu - Thursday, May 28, 2009 - link
How about reparamaterizing the shorter is better bars to something that works as longer is better. One way to do this would be to consider the absolute value of the difference between every piece of hardware and the worst performer. The worst performer would be equal to zero.There is probably an infinite number of other ways to make something like that work, and although the information's value would be diminished because it is only based on the results (and doesn't truly depict the actual results), but it would be valid in comparing relative performance. If something like this is done, I think it'd be important to keep the actual results viewable to ensure that there is no bias (I love you guys, but there are always jerks that are like "wow way to give AMNvidintel an advantage to make them look better")
tynopik - Thursday, May 28, 2009 - link
the problem isn't so much confusion (although that is an issue) as it is non-linearityespecially when charting widely different capabilities (from Atom to Nehalem), the top-end get compressed to nothingness
for instance going from 10 seconds to 9 seconds (10% improvement) is more impressive than going from 200 seconds to 190 seconds (5% improvement)
yet the 5% difference is 10 times longer than the 10% difference!
> (they are useful after all),
fortunately it's easy to convert them into linear scales
for instance, instead of saying how long it took to compress 1GB in winrar, say how many MB/sec it could compress
strikeback03 - Thursday, May 28, 2009 - link
Except that saving 10 seconds at a time is more useful than saving one second per operation if it is a common operation.tynopik - Thursday, May 28, 2009 - link
except that you're trying to get a relative idea of the performance of the CPUsif 1 cpu is twice as fast as another, that should be readily apparent without having to judge what the length of a bar means in different sections of the graph
so it might only be 1 second on this task of this size, but it could be 100 seconds on a different or larger task
just for example, imagine a comparison of cpus from 286 to Nehalem that measured how long it took pkzip to compress 20MB
an athlon XP would be indistinguishable from an i7 920
in reality the 920 trounces the athlon XP, but the legend is so compressed it becomes hard to tell
fredsky - Wednesday, May 27, 2009 - link
hi anand !- win 7 : GO for it. i've just purchased a GSKILL falcon SSD (brother or OZC vertex), and done lots of benchmarks for photoshop. i compared XP64 and win7 64, win7 is much more efficient. so SSD + win7 are the winners! on my old setup, vista is much slower and problematic than win7.
- RAID SSD : as many readers, i would like to see the performance difference between OCZ vertex SSD :
2 RAIDed 30GB, 2 RAIDed 60GB, single 60GB, single 128GB.
because 2x60GB cost about the same as 1x128GB.
chrisf6969 - Wednesday, May 27, 2009 - link
Go ahead to Vista 2.0, errr, I mean Win7. I've been dual-booting it and it feels ready for launch!One request: Can you test a G.Skill Falcon 128Gb SSD. Price/Gb looks pretty good, and they're quoting some impressive performance #'s. I'd like to see you do your REAL world testing on it! PLEASE!
chrisf6969 - Wednesday, May 27, 2009 - link
P.S. Are you going to RAID two Vertex 30Gb's as they're so cheap, seems like you could get some really impressive # for $200+.DrSmith - Wednesday, May 27, 2009 - link
I think the article should include xp & win 7. Vista yuk...With explanation with REALLY how 'worn down' does a mlc drive become with heavy usage. I have a vertex & its fast w reads & writes. But I keep hearing these stories on them 'not lasting' ? How long do they last? compared against SLC? do the storage cells wear out???
ciparis - Wednesday, May 27, 2009 - link
Windows 7? Go for it, Anand. I think many of your readers are already there, and many more are considering it. Best OS they've ever released, even as an RC.coda6 - Wednesday, May 27, 2009 - link
But can we get an in-depth HTPC article for the Anand site? :Dgwolfman - Wednesday, May 27, 2009 - link
Thanks Anand, greatly appreciated!onisaur - Wednesday, May 27, 2009 - link
if other benchmarks have shown no tangible difference between 7 and Vista, by all means, go with 7. we're here for the relative performance anyhow, not the exact cloning of your rig for exact times.on older pc's. to me, the base of the world is a p3-500. the old gateway/compaq cast iron horses still run, still browse the net, and with a new harddrive are zippy enough. i see tons of cel-1200 tualatins that effectively match the p4-1.3-1.5's to cover the 'painfully almost' catagory. and consider a p4-2.4 just because there really are that many dell dim2400's still out there.
on the ssd's: with intel slc 32gb ~$400usd, 64gb ~$780usd, i only lost $20 going with raid0 32gb's on an ich10r. hdtach3 296MB/sec avg, 496MB/sec burst. since so many of the ssd's are ~2x the cost as you go up in size, PLEASE consider putting raid0 smaller ssd's on benchmarks. 60GB is enough to survive on a boot drive. two 32gb supertalent's for ~$270 total should plausibly destroy an intel slc x25-e 64gb at $500 cheaper, and 3.5" brackets mounting two 2.5" drives make it all too convenient.
on pivotal boot and app load tests, consider distinguishing "with same drive swap file", "with other drive swap file", and "with zero swap file".
thebandit - Wednesday, May 27, 2009 - link
Aren't we still missing an article on ddr2 vs ddr3 ? what happened to that ?h0kiez - Wednesday, May 27, 2009 - link
1 more vote for switch now. I'm using it exclusively, and if there's a difference with respect to SSD performance related to trim support or anything else, I want to know.vol7ron - Wednesday, May 27, 2009 - link
I would like to see the P4 2.8E (worst overclocker) in the bench comparison, as well as the Core2 E6600. I jumped on early with the Core2, able to pick one up before they hit the market. I think it's about time to get an upgrade.The other thing that might be nice to see is the performance w/ water cooling and regular hair cooling (fan or just a spreader).
As for SSDs, I'll be looking forward to your SLC comparison :) as well as the results from the 8core NeX (Nehalem-Ex)
vol7ron - Friday, May 29, 2009 - link
by hair, I meant *air*VaultDweller - Wednesday, May 27, 2009 - link
"How much would you guys hate me if I switched to Windows 7 sooner rather than later?"No hate at all. I'm in the process of switching everything over to Windows 7 RC (my main desktop has been on Windows 7 since the Beta), so Windows 7 results are most important to me.
I'm sure there will be those that disagree, though.
JonnyDough - Wednesday, May 27, 2009 - link
"Although our recent article showed that it’s not really any faster than Vista"Dammit. I was really hoping that Windows 7 would offer some type of performance increase over Vista, not just better SSD performance. Oh well, I guess I'll take what I can get. I really dislike the overlordishness of mainstream operating systems these days, but if it helps improve overall ease of use then I suppose it's a sacrifice I'm willing to make.
cordas - Wednesday, May 27, 2009 - link
I must admit I am as excited as the next guy when it comes to SSDs, i can't wait for them to become an affordable viable option for a boot drive, nevermind them becoming big enough to use as my main drive (minimum of 1TB currently).However I am thinking of replacing my aging 500gb HDD with a nice new spangly 1TB drive and have been looking for reviews of said beasties and have been unable to find a review done in the last 12+ months anywhere....
So how about seeing if you can grab the latest and greatest HDDs on the market and giving them a spin to tell us what to get and what to avoid.
Automan - Wednesday, May 27, 2009 - link
Definitely would like to see you move to Windows 7. I've been running the x64 edition on my daily use laptop since Beta 1 and have had no unexpected significant issues with it. Go for it!samspqr - Wednesday, May 27, 2009 - link
* thanks for the status report* that WePC project looks a great iniciative, let's see what comes out of it
* thanks for letting me know I don't want an SSD drive yet
* that 64 cores nehalem system looks like a 3D artist's wet dream, but I guess the price is still going to be quite a bit higher than that of a comparable render farm based on more standard equipment
Drazick - Wednesday, May 27, 2009 - link
Hello Anand.What about checking the SSD Optimizations made in the Chipset level?
Are the newer chipsets of Intel / AMD / nVidia any better using those drives? Does each of them has plans on making optimized chipset for SSD?
Thanks...
Pointwood - Wednesday, May 27, 2009 - link
I've been using it as my primary OS since the RC release and it's seems to generally perform better than Vista. I've only had one slight problem and that was getting my Colormunki monitor calibrator working. Win7 wasn't able to find a driver but I found it myself and got it working.So I say "GO GO GO" in regards to switching to Win7.
feraltoad - Wednesday, May 27, 2009 - link
Switch to Win7!OK, I'm already using the RC, but it really is great. Plus, it's not different enough from Vista to really make difference if you go ahead and switch, and if the adoption of the Beta and RC are any indication then Win7 adoption will be out the roof!
Live - Thursday, May 28, 2009 - link
People already hate vista so switch to vista 2 by all means.I would also suggest that the poll system you implemented a while back would be perfect to use if you want to ask a straightforward question like this. I wouldn't have to ramble on in my broken English and I wouldn't have to read every answer to see the results. Just a thought...
Live - Thursday, May 28, 2009 - link
I forgot to add that I really appreciate articles like this one and that a edit function is needed.thebeastie - Wednesday, May 27, 2009 - link
Yeah looking forward to it, just like the latest Terminator movie :)JonnyDough - Wednesday, May 27, 2009 - link
Or any Christian Bale movie...Souka - Wednesday, May 27, 2009 - link
Check out "The Machinist" if ya havn't seen it...FlameDeer - Wednesday, May 27, 2009 - link
After reading this update, really amazed with such a wide scope coverage that you are currently working hard on. Much anticipating with what will coming next from AnandTech.Thank you Anand & keep up the good work. :)
iwodo - Wednesday, May 27, 2009 - link
On the Bench, the main reason is that people want to know how much performance we are getting for our money. I think the recent economy has changes some of our thinking in spending.The past few years we had experience upgrading parts with not much performance difference. It is interesting to see how previous CPU fair with current generation. And finally realize the days we had improvements like 386 to 486 and Pentium are over.
SSD - The mystery of multiple channel SSD has yet to solved. Is it possible to have more then 10 channel SSD drive in an 2.5" SSD. If not what is the maximum possible channel, i.e what performance should we expect from future SSD.
EX seems to me an destructive product. Previous Benchmarks shows how much performance / watt the Nehalem was able to deliver. May be WoW servers can finally fit more players inside a Single Realm?
TA152H - Thursday, May 28, 2009 - link
If you think about it, it does make sense.The biggest jump, of course, was the 286, arguable the best processor Intel ever made for its time. It was extremely fast, so much so that I couldn't believe it the first time I used it. Going from a PC to a PC/AT was a jump that was mind-boggling. On top of that, it added virtual memory, and the ability to multitask. It was a huge improvement.
The 386 sucked, although it did move to 32-bit and added the now worthless Virtual 86 mode. The performance was poor though, as it ran like a 286 at the same clock speed, clock normalized. But, it moved the platform forward.
The 486 was a nice processor, for sure. It didn't add anything new, but ran like a raped ape. I remember how we fought over the first 486 PC made, an upgraded version of the PS/2 Model 70-A21, which was called the B21. It cut compiles down from 30 minutes to about 18 compared to a 386/25, and it was only running at 25 MHz, and did not have any L2 cache, unlike the 386 which had a 64K external cache.
The Pentium wasn't so great. It ran at poor clock speeds at first, and when I used it, I found the performance quite poor. I used to say Intel's odd numbered processors were all bad, and their even ones good and this was further proof. I went from a 486 at 100 MHz, which were available, to a Pentium at 66 MHz, and was entirely unimpressed. It ran hotter than Hell too. But, in time, they got it right, with shrinks and the Pentium MMX (although MMX wasn't so important, there were other enhancements). The FDIV bug was kind of funny too, and how idiots overreacted to it. Still, I have one of the processors and I hope it will go up in value.
The Pentium Pro was another initial disappointment. It ran 16-bit code very poorly, although the fault lies more with Microsoft than Intel since they gave Intel guidance that 32-bit code would dominate by 1995 when the Pentium Pro was released. It was really expensive too, because of the on-processor L2 cache (which was really on the same packaging, but not part of the processor), which could not be tested before joined with the processor, so if one were bad, they were both thrown out. The Pentium II addressed the segment register issue of the Pentium Pro, so improved 16-bit code somewhat, and obviously what we use today is still based on the Pentium Pro design. It was, of course, and even number (P6) design, so was good.
The Pentium 4, an odd number design, of course, sucked far worse than the Pentium and 386. It was much hotter, like the Pentium compared to the 486, but the performance sucked too, which made it a double threat.
Now we are back to the Pentium Pro design, so it's only reasonable we don't see dramatic changes since it's really not a new processor at all, but just improvements to an existing line, like the Pentium MMX, or Athlon XP (I'd use Prescott as an example too, but was it an improvement?????).
It makes sense though. Really, virtually nothing in PCs is new, and never has been. Virtually every concept has been taken from mainframes, or supercomputers, and microcomputers were just low cost implementations of those concepts. When you had a scarcity of transistors, there were many very good things you could not implement, but you wanted to. That low lying fruit is gone, and adding transistors gives greatly diminished returns. It's like the improvement with any technology. If you look at cars, jets, etc... once the technology matures, the improvements slow down and you get gradual improvements.
I'd like to see x86 die, but I guess it will not happen. Processors cost more, perform worse, and use more power because of this miserable instruction set. It's not a lot per processor, but when you multiply it by the millions upon millions of processors that are afflicted with the x86 instruction set, the costs are staggering. That's not even counting the migraines it has caused by people who have to code in assembly for it. It's gruesome.
yacoub - Thursday, June 4, 2009 - link
Great post, but you forgot the first Celeron, the 300a that overclocked to 450MHz without breaking a sweat. In my book, the best bang-for-the-buck Intel CPU right up through the Core2Duo.This was in the early days of overclocking, just over a decade ago, and it was a dream to be able to run a chip at 150% its rated speed with zero changes to things like voltages (heck most motherboards didn't even support voltage changes back then, iirc).
Ah... those were the days. :)
JonnyDough - Wednesday, May 27, 2009 - link
I agree with what you said about the economy changing things. It isn't just that though. Computer technology isn't gaining nearly as quickly, and the need for new systems - at least in the consumer space is less pressing. The dollar is far more important now than it was. Purchases even at the business level are determined even more by cost than in the 90s, when changes to performance was more heavily weighed factor in IT decisions.This is why businesses that want to survive have to be broadly invested. For example, if a company that manufacturers hard disk drives exclusively fails to get going on SSD early enough, it will be facing extinction.
If you're GM, it pays to invest in something other than just big gas guzzling automobiles. :-D Let's hope the Chevy Volt has some real value. I have my doubts that it will be as popular as they'd hoped, but I do think the car merits an award for ingenuity. They took conventional thinking and flipped it on its head, I just wish they'd done it 20 years ago. With their marketing dept they might've been able to put Ford out of business.
dragunover - Wednesday, May 27, 2009 - link
Removing the necessity for expensive server RAM? Someone's gonna hate that, and someone's gonna love that.Personally I'm the latter of the two.
Jaybus - Wednesday, May 27, 2009 - link
The trade off is a more expensive motherboard, though, that has buffer memory. It wasn't clear to me where this memory would be, separate ICs soldered to the mb or integrated into the chipset. My guess is the former, at least to begin with, in which case RAM manufacturers won't hate the change nearly as much.tshen83 - Wednesday, May 27, 2009 - link
Instead of subsidizing FB-DIMM's AMB chip on each DIMM sold because of extra power consumed, now Intel can charge for the SMB chips individually(8 socket system would require a whooping 32 SMB chips).Depending on how much Intel wants to price those SMB chips, it could make or break Nehalem-EX's platform. (A simple $10 per SMB would easily mount to 320 dollars cost to the motherboard manufacturers, which could then turn into a 600 dollars+ premium on the motherboard)
This approach is indeed the correct one from a technical perspective. Basically Intel wants all the memory premium associated with Nehalem-EX platform by forcing the memory makers into producing commodity DDR3 DIMMs, not FB-DIMM2s.
128 threads + 128 DIMMs...this is porn.
KentState - Tuesday, June 2, 2009 - link
A $600 premium isn't going to add much to the overall cost. I would bet that an 8-way server with 128GB of ram is going to run $40k+.sprockkets - Wednesday, May 27, 2009 - link
Looking forward to it.vol7ron - Thursday, May 28, 2009 - link
Did he say he's gonna compare the new SLCs to the X25-M. Isn't the X25-M an MLC? I know Intel puts out a decent SSD, but is that a fare comparison?vol7ron - Friday, May 29, 2009 - link
by fare I meant fairsbuckler - Monday, June 1, 2009 - link
If they are of a similar size/cost then obviously it's fair to compare them as people looking for a drive of that size/cost will be able to pick either.vol7ron - Monday, June 1, 2009 - link
I thought cost was almost negligent with regards to the enthusiast market :)I thought you compare on comparable performance and then evaluate the costs; since, after all, prices change due to rebates, sales, and demand.
dipakmr40 - Thursday, December 3, 2020 - link
very nice post all the article of your website have been read by me ..I recently write a new article please have a look.
https://www.wishingboxs.com/best-gaming-mouse-unde...