Slashdot Mirror

The Celeron D isn't dual core. It's just a Prescott core Celeron with 256KB of L2 cache.

Time to update the firmware on our CPU name/model decoder rings.

In more recent history, they patented (John) Carmack's Reverse rendering technique, and then used it against id to force them to include EAX functionality in the Doom3 engine.

This sure is the strangest 10-30% I've ever seen.

Yeah. Some tests were only won by a few percentage points. Others absolutely obliterated the 4200rpm drive by over 300%. Overall effect? Who knows... depends on how you use it. Either way, I'd say it's worth the price difference.

While that may be true, let's not forget that the CoreDuo is still intended as a mobile processor. The fact that it is even competitive with the X2s is impressive. The real test will come when the Conroe comes out. Early reviews (granted, on intel supplied hardware) show Conroes kicking the crap out of even the fastest AMD processors (overclocked even) in everything from gaming to media encoding and synthetic benchmarks. They are cooler and more efficient to boot. Unless AMD pulls some magic out of its hat, my next rig will be Intel based. See here: http://www.pcper.com/article.php?aid=219&type=expe rt&pid=1 here: http://techreport.com/etc/2006q1/conroe/index.x?pg =1 here: http://www.anandtech.com/tradeshows/showdoc.aspx?i =2713 and here: http://www.hexus.net/content/item.php?item=4843

Actually, I suspect the grandparent is remembering is most likely when Nvidia had special detection for 3dmark03.exe. Doing optimizations by filename isn't _necessarily_ evil, provided that it's disclosed, done to correct things in the game that could have been done better on your hardware, and doesn't degrade the quality of the output.

For example, if Game X were drawing water a specific way, and nVidia cards could render the exact same water more quickly using a different way while retaining the same quality, it wouldn't be unreasonable for nVidia to include such an optimization. This should, however, be disclosed, so that reviewers can test with (more representative of what a user can expect), or without (testing the raw hardware) this optimization.

Detecting a synthetic benchmark is just cheating, and they were actually lowering the quality of the output as well. It wasn't even an "this could be done better this way" type of tweak.

Also reviewed at The Tech Report, with more extensive testing against a wider range of processors.

Most large scale systems that use SSD's to increase DB performance do so using DRAM (mainly) or SRAM based units with battery backup...The units are ridiculously expensive, but far faster than anything you'd manage to get with flash or harddisks

You can get versions in the $150 range that use standard SIMMs, then fill them up with whatever size you want. Definitely not the same as the high-end stuff, but pretty cool. Gigabyte's i-ram comes to mind--it's battery-backed, kind of (like 12 hours completely off power or something, draws from the PCI bus to maintain state/recharge).

http://techreport.com/reviews/2006q1/gigabyte-iram /index.x?pg=1

1) I'm using a Gigabyte nforce4 ultra motherboard (the passively cooled one) with 8 onboard sata controllers and another cheap over the counter 4 port sata pci card for a total of 12 drives.
2) I made a mistake here by using all of the same drives bought from 1 place - I only read the horror stories after :( - It seems very possible for multiple drives to fail within 20 minutes of eachother after 8 months (!!!) if they have consecutive serial numbers and part of a bad batch. Dont make this mistake! - Since I couldnt return the drives, I made another raid1 array with different drives for critical storage and OS, the raid6 is mostly media backup and recording shows.
3) I use software raid. SCSI hardware raid is far too expensive and the performance of software raid is generally better than those onboard raid controllers. I also like having the ability to resize the array (although not tried it yet) and I just trust it more. I've been using software raid for half a decade now and it withstood all of my abuse so far while I saw expensive scsi raids fail from simple things like rebooting while the array is rebuilding.
4) Another stupid mistake here, I'm using reiser4 and its caused problems for me already. No data loss, but just try using it with a berkleydb database or anything like that and watch the kernel panic :) - I also had reiser4 corrupt partitions at work on development boxes (no data loss, but unexpected behaviour and reboots required), there is also other strange dmesg errors for reiser4 and if for example you umount a reiser4 partition and mkfs.ext3 over it, it just refuses to work, you have to reboot. I wouldnt recommend reiser4 for now!
5) I have an ANTEC PSU/Phantom 500 GB 500W ATX12V v2.01 PSU. To be honest I havent hooked up the failsafe PSU yet but its an option on the coolermaster stacker I'm hoping to use soon. I havent had any problems yet, but I guess it does strain the PSU when the drives wake up and the PSU fan does turn on for a while when drives spin up. Looking at the data sheet, it seems the drives do spin up at 30W even though operating peak is only 12.3W (7W idle) and my system is otherwise like the one tested here: http://techreport.com/onearticle.x/7417.

The would mean:

Hard drives: 84W idle 147W peak 360W spin-up
System: 110W idle 150W peak 150W
-------
Total: 194W idle 297W peak 510W max (ouch!)

The PSU is 85% efficient so I guess 510W is quite a bit over the realistic maximum of 420W, but the system is not only idle most of the time but the CPU auto throttles to 1GHZ almost all of the time, so while I havent measured this, I bet the idle consumption of the system is closer to 90W which means 330W available for the drives. Now that I think of it, I probably need more power, lol.
6) Only issue is slow wake up time. It takes around 15-30 seconds before I get a file listing.
7) Cooling tips, well, fanless motherboard, fanless (for the most part) PSU means less worry about fans :) - I have 3x120mm 20db fans drawing air in and cooling hard drives on the way, 1x120mm 25db fan sucking air out the back, 1x25db fun sucking air out on the top and the cpu cooling is the Zalman CNPS9500 (the huge vertical cooler) that also helps with the airflow. I also have dust filters on all intake (included with coolermaster stacker).

As for silence, at night it is no way near silent.. but in a good way :) - What I mean by that is there is no whine or noticeable noise, instead there is a very low tone hum. If you walk past the PC at night you may not notice it is switched on until you switch it off - its basically like the sound of the ocean from a far, you dont want to turn it off :) - Then again we are in the countryside so chances are it will be plain inaudiable for most in the city.

while you have an interesting idea, i don't think anything like that will be made in the near future for amyn reasons. first off, it would go against an industry form factor (PCI type expansion slots) that has been in place for a lomg time, and companies like nvidia and ATI would not want to risk making a venture outside such a long standing form.

another problem with changing the form for graphics units is comaptibility, as i for one do not have a free 5.25" bay in which to put something (i already have 3 5.25" hard drive bays and a DVD drive, as i ran out of 3.5" slots). this is also link to the fact the the current pci type clot is a long running standard. making a cable from an external device would be tricky, but would probally work if heavily shielded.

a video card interfacing through the memory bus? highly unlikely, as there are significant differences between the memory bus and a video bus. you would have to create a hybrid bus that could handle both data types, which would complicate things and just make everything slower. the GPU fetches a lot of data from the main memory, but placing the entire GPU on that memory bus is a bad idea.

currently, video cards are not at the point where such measures need to be taken, and i hope that time does not come. as i'm sure other people have said on here, i think there should be more work on making cards run cooler and use less power. everything has been a chase for performance without much concern for heat and power use. when it reaches the point where the cards are just getting too hot, i think they will making more of an effort for make the more efficent. the new geforce 7900 cards did this by performing better with LESS transistors than the previous generation. i think once heat and power use become too much of a factor for the cards to sell, the companies will start to make them cooler and more efficent.

RAM- you need enough RAM so you don't hit swap. For today's games, thats 1 GB. After that, the number 1 thing you can do to improve system performance is to get low latency RAM. Your CPU will be waiting for RAM, minimize the time that it is.

Well, ok, it's not trash-junk, but it's not up to the hype either. There was a review on TechRepublic a while ago that I'm pretty sure made it to slashdot (if not, then digg). Basically it showd that Low Latency RAM in itself made little to no difference and more RAM was always the way to go.

That said, Low Latency ram is not entirely a waste of money. Low Latency ram has a better shot at overclocking (like turning DDR 400 into DDR 450 by relaxing the timings and pumping the clock rate.) It's also more likely to be higher quality and thus less likely to go bad on you. As an asside, if you're interested in RAM in general, I've found this site very informative/userful.

But I would minimize the value of Low Latency in and of itself.

http://www.amdzone.com/modules.php?op=modload&name =Sections&file=index&req=listarticles&secid=13

http://www.anandtech.com/video/showdoc.aspx?i=2717

http://www.guru3d.com/article/Videocards/326/

http://pcper.com/article.php?aid=213

http://techreport.com/onearticle.x/9529

http://www.tomshardware.com/2006/03/09/ati_and_nvi dias_same_day_mega_launch_mayhem/

Their PC is way more powerful than it needs to be. Should have saved some money on cpu/graphics/memory. Those savings then could be used to make that thing so much quieter. Number one change in my opinion would be to use a sp;od state storage, such as http://www.techreport.com/reviews/2006q1/gigabyte- iram/index.x?pg=1(possibly the only use I would have for one of these currently). Also I'd use a near-silent psu and water cooling...

Yeah... I found the benchmarks pretty impressive given that the 2.0GHz Intel part was compared to a 2.4GHz AMD part (as per the test systems specs. And it wasn't even the latest version of the Intel parts (it wasn't a Yonah based part). Yonah has a few architectural improvements over the Dothan even, both in power management and in performance. I'm not a brand-loyal customer either, I buy what I think is best at the time I buy (which is why I have two Athlon64, one Athlon64 X2, and three Athlon XP machines). From the looks of things, my next machine may very well be an Intel based one if I decide to upgrade one at the end of 2006.

Sorry guys, I left out the AMD CPU in question (to compare to the Solo): mTurion MT-34 (1.8GHz 25W TDP)

Since TFA showed a 2.0GHz Pentium M outperforming a 2.4GHz Turion in most of the important benchmarks, I think the 1.66GHz Core Solo (with its architectural improvements over the Pentium M) might outperform a 1.8GHz Turion by even more.

I think TFA is pretty weak, though. They emphasize the importance of performance per watt, but they don't include the widely available Turion MT. They also used freakin' desktop chipsets with these mobile processors. Doesn't this make the "system power consumption" numbers useless for most readers? I thought the mobile chipsets were supposed to have important power-saving technologies in their FSB, memory contollers, wirless modules, etc.

Sorry guys, I left out the AMD CPU in question (to compare to the Solo): mTurion MT-34 (1.8GHz 25W TDP)

Since TFA showed a 2.0GHz Pentium M outperforming a 2.4GHz Turion in most of the important benchmarks, I think the 1.66GHz Core Solo (with its architectural improvements over the Pentium M) might outperform a 1.8GHz Turion by even more.

I think TFA is pretty weak, though. They emphasize the importance of performance per watt, but they don't include the widely available Turion MT. They also used freakin' desktop chipsets with these mobile processors. Doesn't this make the "system power consumption" numbers useless for most readers? I thought the mobile chipsets were supposed to have important power-saving technologies in their FSB, memory contollers, wirless modules, etc.

Anandtech

Techreport

Neoseeker

And still working at that.

http://www.pcper.com/article.php?aid=202&type=expe rt
http://www.hardocp.com/article.html?art=OTUz
http://www.techreport.com/reviews/2006q1/radeon-x1 900/index.x?pg=1
http://www.hothardware.com/viewarticle.aspx?articl eid=777&cid=2
http://www.beyond3d.com/reviews/ati/r580/

The Tech Report's in-depth analysis is flowing freely with narry a hiccup. Check it out here.

Tech Report (FX60 beats out 955 in most of the benchmarks, if not by a large margin then at least consistently).

Also check out AMDZone, AnandTech, Björn3D, FiringSquad, HEXUS, HotHardware, LostCircuits,
PC Perspective, t-break, and TrustedReviews who all have reviews as well.

Here's where the 'sweet spots' in terms of performance/price are, in my opinion; choose depending on your budget. (Of course, if your goal is to waste money, there's plenty of components available at or near $1000 prices, as well, but they don't provide much more than a 20% or so performance increase over the $200-300 options.)


Processor:
Athlon 64 3200+ ($160)
A 2GHz Athlon 64 with 512K cache. As is widely known, these beat the pants off of Pentium 4s.

Athlon 64 X2 3800+ ($320)
Two 2GHz Athlon 64s with 512K cache (dual core).



Motherboard:
Abit KN8 SLI ($110)
SLI doesn't carry much of a price premium any more these days, so it can't hurt to have the extra upgrade capability. Other brands like DFI, Asus, MSI, EPoX, are fine as well.



Memory:
2x 512MB Crucial PC-3200 ($95)
2x 1GB Crucial PC-3200 ($170)
Two is so you can run them in dual channel mode. Other good brands include Corsair, Kingston, Mushkin, OCZ.


Video card:
GeForce 6600GT 128MB ($125)
8 pixel pipelines at 500MHz = 4 Gigasomethings

GeForce 6800GS 256MB ($190)
12 pipelines at 425MHz = 5.1 Gigasomethings. This also has double the memory and memory bandwidth of a 6600GT, so it'll handle higher resolutions and antialiasing levels much better.

GeForce 7800GT 256MB ($270)
20 pipelines at 400MHz = 8 Gigasomethings. This is almost exactly double a 6600GT in many respects (double the pixel pushing power, memory, and memory bandwidth).


If you want to find things out for yourself, I recommend browsing around at The Tech Report and AnandTech; I've found these two to consistently have the highest quality reviews and comparisons out there. Their system guides don't completely suck, either. (Neither do Ars Technica's, but they don't do hardware reviews).

Saw this funny thing today at TechReport, a PCIe card where you can mount two mobile graphics modules/cards (MXM) for SLI in desktop systems. Weird prototype thingy. Not sure what market this is targeting...

I just saw this http://www.techreport.com/onearticle.x/9182

It's ok if you can copy music, but ALL software has to have DRM of some kind INCLUDING open-source. That one is repsonsible for what one's application does, and if it can be used illegally and is/has done so, then you face fines/jail time. That sounds damn right hilterriffic to me...

Great, it's the Carmack's Reverse situation all over again...

Look, your view of an improved console is not the same as MS' or Sony's. I totally respect that. But that doesn't mean that MS made a bad or faulty design. And it doesn't mean other people agree with you. I highly recommend you not buy a 360, it doesn't seem appropriate for you. But your arguments on this are good for those in your position only. Since HDTV resolution is 3-4X times that of regular TV, I personally am looking for more than "several times" faster equipment. It would take that just to tread water, graphics-wise at HDTV resolutions, and I (and MS) think improved graphics are a better idea than that.

As to you not being clear about cooling, you again, for the 3rd time just don't know what you are talking about in this area. If you put the unit in a box it will heat up far above normal temperatures. You can't ensure you have headroom if you can't control the cooling solution, and if the user puts the unit in a box, you can't control the cooling solution. Okay. Do you get it now? I also explained why you cannot always warn the user. I explained a few practical reasons why not and a legal/liability one. I finally took the time to explain that you have NO IDEA what the incidence of the problem is. If it happens 5 times out of a million, then MS shouldn't spend a single dime fixing it. If it happens 5,000, it's quite important. Until you know which it is, you should refrain from explaining how MS did a crappy job and you'd do better.

At room temperatures, heat doesn't produce significant changes in electrical resistance in copper, tin, etc. We're not talking about superconductors here. Also note that there is no evidence that more heat makes chips wear out quicker, at least not until you get to the catastrophic failure mode. Additionally, as a 360 owner, I can say that your comments about "robust operation" are pretty presumptive. You just have no idea if they are unreliable. Mine is reliable at at least 116F-140F. Frankly, that seems pretty good to me. So step off about your comments as to Xbox 360 would be more reliable if it ran cooler.

Finally, all those comments are AGAIN tangential. MS felt they needed a certain level of performance. That level of performance generates a certain amount of heat. Putting that much heat in a stereo cabinet doesn't work. So no matter how much you complain, you're not going to change the situation. It seems likely to me MS did about as well as reasonably could be done given the cost and design constraints.

And making it "a few percent" slower would only reduce the power consumption by a few percent. For example, look at an AMD A64. http://www.amdcompare.com/us-en/desktop/ Select 90nm from CMOS and A64 (regular) from processor. Also look at http://techreport.com/cpu/ for the voltage ratings.

A 90nm 4000+ runs at 2.4GHz and 1.4V and has a design power of 89W. A 3700+ runs at 2.2GHz and runs at 1.4V and has a design power of 89W. A 3200+ runs at 2.0GHz and 1.4V and has a design power of 67W (make sure to use socket 754 version, other suffers from reduced performance due to bad FSB).

Power is proportional to frequency and with the square of voltage. So, let's calculate. Performance is proportional to frequency.

So, a 3700+ is 91% as fast as a 4000+, and uses 91% as much power.
A 3200+ is 83% as fast and uses 83% as much power.

If a 3200 could run at 1.3V (not impossible, it would than take (1.3/1.4)^2*(2.0/2.4) as much power. That'd be 71% as much power.

In that case, I'd save as much as 20W (30%), but lose about 20% of my power. If you go by TDP ratings on AMD's site, it comes out to about the same (62W TDP versus 89W, for about 70% as much power, although these are rough numbers, AMD seems to class TDP ratings by what heatsink they use, not actual TDP). Anyway, how does this jive with your statement that it would use MUCH LESS than 80% of the power?

It just happens to be that personally, you don't find the tradeoffs

The Tech Report got SLI working with this chipset. http://techreport.com/reviews/2005q4/via-k8t900/in dex.x?pg=5

"We were able to obtain a set of older NVIDIA drivers, revision 71.24, that don't include a chipset-based lockout for SLI. These drivers aren't new enough to support monsters like the GeForce 7800 GTX 512, but they work just fine with a couple of GeForce 6800 Ultras. Here's the K8T900 going head to head against the nForce4 SLI in SLI mode."

I hear the 6800 GS is the best value for money right now. (Just as as fast as the plain 6800, but cheaper.)

Personally I just bought a 6600GT because 150 euros is all I can afford, and my Radeon 9100 isn't cutting it any more.

With 500GB drives, it would take 7340.032 drives to attain 3.5PB... with NO redundancy.

Height (mm) 25.4
Width (mm) 101.6
Depth (mm) 146

1. First.
2. Second.
3. Third.
4. Fourth.
5. Fifth.

http://www.pcper.com/article.php?aid=182&type=expe rt
http://www.hardocp.com/article.html?art=ODg1
http://www.techreport.com/reviews/2005q4/geforce-7 800gtx-512/index.x?pg=1
http://www.anandtech.com/video/showdoc.aspx?i=2607
http://www.hothardware.com/viewarticle.aspx?articl eid=751&cid=2
http://www.guru3d.com/article/Videocards/285/

http://www.techreport.com/etc/2005q3/hires-gaming/ index.x?pg=1

Several games benchmarked at 2048x1536

I want to upgrade my graphics card, but my options are limited without having to upgrade my 300 watt power supply

You could look for fanless video cards; they're probably low-ish power since that's less heat to lose efficiently.

Or read The Techreport's power and noise page in their 6800GS review. None of them topped 300 watts.

Listed alphabetically.

http://www.anandtech.com/video/showdoc.aspx?i=2593
http://www.avault.com/hardware/getreview.asp?revie w=evga6800gs
http://www.computerbase.de/artikel/hardware/grafik karten/2005/test_nvidia_geforce_6800_gs/
http://www.driverheaven.net/reviews/X16_GS/index.h tm
http://www.guru3d.com/article/Videocards/278/
http://www.hardocp.com/article.html?art=ODgy
http://www.hothardware.com/viewarticle.aspx?articl eid=747&cid=2
http://www.noticias3d.com/articulo.asp?idarticulo= 541
http://www.nvnews.net/previews/geforce_6800_gs/ind ex.shtml
http://techreport.com/reviews/2005q4/geforce-6800g s/index.x?pg=1

There are lots of other reviews out there, too. Looks like the 6800GS kicks the X1600 where it hurts. Over and over and over again.

- [H]ard|OCP
- Avault
- Computer Base
- Driver Heaven
- Guru3D
- Hartware
- HotHardware
- Noticia3D
- nV News
- The Tech Report

I shamelessly stole this list from Hardocp.com

Pretty decent review here I read earlier:

nVidia 6800GS

And 3800+ vs 820:

http://techreport.com/reviews/2005q3/athlon64-x2-3 800/index.x?pg=13

"The system based on the X2 3800+ draws less power at idle and under load than anything here but the single-core A64 3800+. Under load, the Pentium D 840-based rig draws 292W at the wall socket, while the X2 3800+ system draws 166W. And the X2 3800+ outperforms the Pentium D 840 more often than not. The performance-per-watt picture on the X2 3800+ is impressive indeed."

Load: 166W vs 251W (85W difference)
Idle: 117W vs 160W (43W difference)

The issue is that the second link is to the TDP of the X2 3800+, not the actual power consumption which will be lower than 89W, more likely 70W.

What matters is system power consumption however. CPU + Chipset + Everything Else. Of course you can test with Everything Else being the same, so it comes down to the CPU + Chipset. AMD have an on-die memory controller, so that is a couple of Watts saved over the Intel chipset, however Intel's chipsets are traditionally quite efficient (although whether or not the chipset for dual-core processors is I don't know). Best bet is to measure at the socket.

http://techreport.com/reviews/2005q2/athlon64-x2/i ndex.x?pg=15

Under load:
Pentium D 840 uses 292W at the socket.
Athlon 64 X2 4200+ uses 178W at the socket.

Difference is 114W. Plug that into your calculator!

After the other review posted today about fast memory doing almost nothing for transcoding:
"moving to tighter memory timings or a more aggressive command rate generally didn't improve performance by more than a few percentage points, if at all, in our tests."
"Mozilla does show a difference between the settings, both on its own and when paired with Windows Media Encoder. Still, the differences in performance between 2-2-2-5 and 2.5-4-4-8 timings, and between the 1T and 2T command rates, are only a couple of percentage points."

Have to agree with AC on the cpu issue, taken from the http://techreport.com/reviews/2005q2/athlon64-x2/i ndex.x?pg=16

Conclusions
Let's start by talking about the Athlon 64 X2 4200+. This CPU generally offers better performance than its direct competitor from Intel, the Pentium D 840. Most notably, the X2 4200+ doesn't share the Pentium D's relatively weak performance in single-threaded tasks like our 3D gaming benchmarks. The Athlon 64 X2 4200+ also consumes less power, at the system level, than the Pentium D 840--just a little bit at idle (even without Cool'n'Quiet) but over 100W under load. That's a very potent combo, all told.

In fact, the X2 4200+ frequently outperforms the Pentium Extreme Edition 840, which costs nearly twice as much. Thanks to its dual-core config, the X2 4200+ also embarrasses some expensive single-core processors, like the Athlon 64 FX-55 and the Pentium 4 Extreme Edition 3.73GHz. Personally, I don't think there's any reason to pay any more for a CPU than the $531 that AMD will be asking for the Athlon 64 X2 4200+.

If you must pay more for some reason, the Athlon 64 X2 4800+ will give you the best all-around performance we've ever seen from a "single" CPU. The X2 4800+ beats out the Pentium Extreme Edition 840 virtually across the board, even in tests that use four threads to take best advantage of the Extreme Edition 840's Hyper-Threading capabilities. The difference becomes even more pronounced in single-threaded applications, including games, where the Pentium XE 840 is near the bottom of the pack and the X2 4800+ is constantly near the top. The X2 4800+ also consumes considerably less power, both at idle and under load.

The X2 4800+ gives up 200MHz to its fastest single-core competitor, the Athlon 64 FX-55, but gains most of the performance back in single-threaded apps thanks to AMD's latest round of core enhancements, included in the X2 chips. The X2 4800+ also matches the Opteron 152 in many cases thanks to Socket 939's faster memory subsystem. Remarkably, our test system consumes the same amount of power under load with an X2 4800+ in its socket as it does with an Athlon 64 FX-55, even though the X2 is running two rendering threads and doing nearly twice the work. Amazing.

There's not much to complain about here, but that won't stop me from trying. I would like to see AMD extend the X2 line down two more notches by offering a couple of Athlon 64 X2 variants at 2GHz clock speeds and lower prices. I realize that by asking for this, I may sound like a bit of a freeloader or something, but hey--Intel's doing it. No, the performance picture for Intel's dual-core chips isn't quite so rosy, but the lower-end Pentium D models will make the sometimes-substantial benefits of dual-core CPU technology more widely accessible. If AMD doesn't follow suit, lots of folks will be forced to choose between one fast AMD core or two relatively slower Intel cores. I'm not so sure I won't end up recommending the latter more often than the former.

Beyond that, the giant question looming over the Athlon 64 X2 is about availability, as in, "When can I get one?" Let's hope the answer is sooner rather than later, because these things are sweet.

Really, why would someone choose to use a chip that is less powerful, intrinsically costs more to operate, and costs more to cool?

Regrettably, because it has the Intel logo on it. I'm lucky working in a company where if I say I want AMD, I get AMD. I'm sure there's plenty of hardware geeks on /. who've asked for a shiny new Opteron server and been smacked back by either a company "Intel-only" policy, or their reseller's "Intel only" policy.

FWIW, AMD recently launched the new single-core Opteron 254 and it utterly trounces the Intel competition. Even in benchmarks that have been traditional strongholds for the Netburst architecture.

Well John Carmack has said

http://techreport.com/etc/2005q3/carmack-quakecon/ index.x?pg=1

"Graphics accelerators are a great example of parallelism working well, he noted, but game code is not similarly parallelizable. Carmack cited his Quake III Arena engine, whose renderer was multithreaded and achieved up to 40% performance increases on multiprocessor systems, as a good example of where games would have to go. (Q3A's SMP mode was notoriously crash-prone and fragile, working only with certain graphics driver revisions and the like.) Initial returns on multithreading, he projected, will be disappointing."

Mind you if you could really get a 40% speed up it would mean that a X2 4800 at 2.4Ghz would outperform a FX-57 at 2.8Ghz (the fastest available dual core when the X2 4800 was released). In fact it should outperform anything up to 3.3Ghz.

Actually half that, +20%, should be enough to level the playing field between the fastest available dual core and the fastest single core, assuming the 20% clock diff bewteen the X2 4800 and the FX-57 is typical.

Reads:
The Tech Report has posted an in-depth review

Should read:
The Tech Report has posted an in-depth review

(Thanks to synthparadox for the link)

Missing link?
http://techreport.com/reviews/2005q4/soundblaster- x-fi/index.x?pg=1

obviously you missed the fact that when gaming at 1600x1200 and are using 4x antialiasing and 8x anisotropic filtering that the x1800xt beats the shit out of the gf7800.

if you're buying a 500 dollar card, are you seriously worried about benchmarks that are run without aa+af? this card even does HDR (hi dynamic range) plus AA, something that the gf7800 can't.

this card is way more sophisticated and highly refined that the brute force 7800. the 7800 isn't bad but that this card can do with 16 pipelines what the 7800 can't do with 24, says a lot.

and that's just raw performance with todays games. never mind the fact that the 1800xt comes with 512megs of super fast ram... ready for well into the next generation of games, whereas 256meg 7800's are already obsolete for the high end of the next generation. sure 256 will be enough if you pare down the resolution and lower the texture detail. one example is the game F.E.A.R... on the 1800xt it absolutely trounces the 7800 in performance.

my advice... read ALL the reviews you can get your hands on. there are too many discrepencies if you only read one or two. if you want to get a more full picture, get to reading.

http://www.anandtech.com/video/showdoc.aspx?i=2552

http://www.beyond3d.com/reviews/ati/r520/

http://www.driverheaven.net/reviews/r520reviewxvxv /

http://www.extremetech.com/article2/0,1697,1867116 ,00.asp

http://www.firingsquad.com/hardware/ati_radeon_x18 00_xt_xl/

http://www.guru3d.com/article/Videocards/262/

http://www.hardocp.com/article.html?art=ODIy

http://www.hexus.net/content/item.php?item=3603

http://www.hothardware.com/viewarticle.cfm?article id=734&cid=2

http://www.neoseeker.com/Articles/Hardware/Reviews /ati_radeon_x1800_x1600preview

http://www.pcper.com/article.php?aid=172

http://www.tbreak.com/reviews/article.php?id=407

http://www.techreport.com/onearticle.x/8864

and check out the wicked new 3d tech demos... both are very impressive but the toystore demo is jawdropping.

http://www.ati.com/designpartners/media/edudemos/R adeonX1k.html

wmv9 hi def format but plays fine in mplayer or VLC.

We measured total system power consumption at the wall socket using a watt meter. The monitor was plugged into a separate outlet, so its power draw was not part of our measurement. The idle measurements were taken at the Windows desktop, and cards were tested under load running a loop of 3DMark05's "Firefly Forest" test at 1280x1024 resolution.

Listed alphabetically so no preference to which site is good or not.

http://www.beyond3d.com/reviews/ati/r520/
http://www.driverheaven.net/reviews/r520reviewxvxv /
http://www.guru3d.com/article/Videocards/262/
http://www.hardocp.com/article.html?art=ODIy
http://www.hexus.net/content/item.php?item=3603
http://www.neoseeker.com/Articles/Hardware/Reviews /ati_radeon_x1800_x1600preview
http://www.noticias3d.com/articulo.asp?idarticulo= 527
http://www.pcper.com/article.php?aid=172
http://www.tbreak.com/reviews/article.php?id=407
http://www.techreport.com/onearticle.x/8864

Many other sites will do reviews. Techreport, for example.

http://www.pcper.com/article.php?aid=172

http://www.techreport.com/

http://www.hardocp.com/article.html?art=ODIy

http://www.beyond3d.com/reviews/ati/r520/

About 6 months ago, one of my cheap free-with-case PSU's died. I went on a mission to find a good bang for the buck PSU to replace all 3 of my systems with high quality PSUs. I read several tests and reviews, including the one on Toms and another http://www.overclockers.com.au/article.php?id=3598 67 and another http://techreport.com/reviews/2004q4/psus/index.x? pg=1 and not to mention http://www.pcper.com/article.php?aid=-270.

The conclusion? I went with FSP power supplies labeled as Fortron or Sparkle Power. Yes, you can get nicer more expensive ones, but for a very low price, these beasts couldn't be beat for price/performance. I opened up 300w Sparkle, a cheapie 400watt free with case, and my friends ThermalTake which was fairly pricey rated at 350watt. In terms of size of capacitors, mosfets, heat sinks, etc, the Sparkle definately had the edge in size (size does matter with this stuff!). I replaced all 3 of my systems with these and they are humming along great.

http://www.beyond3d.com/reviews/ati/crossfire/
http://www.bit-tech.net/hardware/2005/09/26/ati_cr ossfire_detail/1.html
http://www.driverheaven.net/reviews/crossfireatire viewxxx/
http://www.hardocp.com/article.html?art=ODE1
http://www.hothardware.com/viewarticle.cfm?article id=730&cid=2
http://www.pcper.com/article.php?aid=168
http://www.tbreak.com/reviews/article.php?id=404
http://techreport.com/reviews/2005q3/ati-crossfire /index.x?pg=1
http://graphics.tomshardware.com/graphic/20050926/ index.html

The summary is wrong, it (the RE2) does have NCQ.

See a real review like TechReport instead of that amateur crap they posted.

Proper TechReport's review here.

Go read. Now!