What is Digg?103 Comments
- Lane, on 10/11/2007, -1/+35now for the real question, where the hell do i get that bitchin Valve Source engine particle simulation?
- inspecality, on 10/11/2007, -3/+29I just bought an Intel e4300 1.8ghz and overclocked it to 2.4ghz on stock cooling, no sweat. This CPU cost me $117. If you want to save money, overclock.
- pureevilmatt, on 10/11/2007, -3/+19Conclusion: If you are a gamer, get the cheapest Dual Core CPU you can, and spend the big bux on GPU.
- Motodog, on 10/11/2007, -2/+13E6300 1.8 ghz overclocked to 3.01ghz easily and safely on air. That's $120 chip running at a clock speed of a $1000 chip. You want value, then educate yourself a little about computers. Unless you're content on emptying your wallet and learning nothing.
- alwaysthere, on 10/11/2007, -2/+12The tests in the article are great and the numbers are useful, but the value of the CPU's as reported are completely wrong. You have to consider total system cost to establish a realistic value based on price versus performance. Basing your value entirely off the price of the processor will lead you to buy the wrong system.
I frequently buy dozens of computer systems at a time for a 3D render farm. When I make my buying decisions I look at how many frames per hour I can render with $10,000.00 worth of computers. That includes CPU, motherboard, RAM, HDD, power supply, and rack mount shelf. I've found I can render a lot more frames per dollar with more cheap computers than I can with a few bleeding edge systems. Heat and electricity are not a significant cost factor for me. Right now (for a render farm) the "sweet spot" is with the AMD Athlon 64 X2 5600+, 4gb RAM, a cheap motherboard with on-board video, a cheap small hard drive and a good quality power supply. I can build 18 of these systems for $10k. - mitrovarr, on 10/11/2007, -0/+9Well, if you want to, you can add the benchmark results for those CPUs into their data set. They scored a zero in every single test for failing to boot the testing operating system or run the benchmarking application.
- betterth, on 10/11/2007, -2/+11Riiiight, because the E6300 at 3.0GHz is the same as a Quad Core QX running at 2.93. Everyone knows that clock speeds are everything and manufacturers just ship the same chip underclocked! Wait, *****, it's not 2001 anymore.
You can clock your processor at 3GHz all day long, but I'm willing bet money that it still doesn't hold a candle to a processor stock at that speed. The other processor will have different architecture, better cache, better assembly, probably better power usage at that level, better voltage levels (for safer/more reliable operation) and to boot, will still be under warranty.
Overclocking helps, but it's not better and it doesn't make you smarter. - inspecality, on 10/11/2007, -0/+9With a non-stock heatsink.
- maninblac1, on 10/11/2007, -9/+17The testing is solid, the results are undeniable, but the conclusions are jank.
The question ultimately isn't, what is the best performance per dollar, it is, what is the best performance per dollar a real person is going to be able to live with. A 21min encode time isn't exactly on the list of optimal times.
I mean, to be honost, the graphs don't lie, but it's hard to say that the chip that lies the furthest to the left is best, ultimately, if you think about what you the end user would actually go out an by, it's going to be the top of the first ridge, whether it's AMD or Intel. - geminitojanus, on 10/11/2007, -0/+7Microsoft already patched their OSes, Apple's OS isn't affected due to the engineers already having access to the TLB specs directly from Intel, and Linus is banging up the two required patches for Linux.
So all-in-all, total non-issue. But you can go on about you FUD now. - kingkilr, on 10/11/2007, -2/+8One thing I would love to see is a performance per dollar after overclocking, because all the top chips reviewed are known overclockers.
- PayneX, on 10/11/2007, -0/+6Someone tell the guy playing Oblivion that he needs to sleep in order to level up.
- computerguydk, on 10/11/2007, -2/+8If this is a list of current CPU's, then why don't they have the Intel Core 2 Duo e6x20's listed? The offer the best bang for the buck in my opinion and are great overclockers.
- computerguydk, on 10/11/2007, -0/+6I see the AMD fan boys are digging me down. Look, I have nothing against AMD, hell I run AMD chips on my other computers, but lets be honest here, core 2 duo basically beats the crap out of most of the AMD chips right now. End of story.
- Stephenishere, on 10/11/2007, -0/+5the quad cores are dropping in price too, the q6600 will only be about 266$. Thats when i am buying a whole new pc :D
- Rcdriver, on 10/11/2007, -0/+5My Firefox window just froze from all that ignorance you posted.
- endersadvocate, on 10/11/2007, -0/+5name one.
- FDL1, on 10/11/2007, -0/+4The Q6600 is going to be almost half the price it is now ($480 to ~$260) in a month's time. If they had done a performance per dollar, it would practically double by then.
- betterth, on 10/11/2007, -0/+4... It has been since it came out. Unless you mean, it's the only game in a list of rendering/computational benchmarks, in which case, it's always good to test game performance when testing processors, since gaming is a processor intensive application.
- geminitojanus, on 10/11/2007, -0/+4Both companies have serious bugs throughout their chips. You just never hear about AMD's errata.
http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/25759.pdf
Somehow it's always a big deal when Intel releases chip errata, but it's absolutely a non-issue that AMD has to do the exact same. These engineers don't do a perfect job, but they also allow their chips to be patched in software either by a BIOS upgrade, or by the OS loading in a microcode patch, or by simply telling the OS developers to avoid putting the processor in a certain state which will cause it to crash. - endersadvocate, on 10/11/2007, -0/+4The Thermaltake Big Typhoon http://www.newegg.com/Product/Product.asp?Item=N82E16835106061
costs less and offers vastly superior cooling, i have my e4300 at 3.4ghz and it tops out at like 55c. with orthos. - cbailster, on 10/11/2007, -0/+4Interesting article. But the scatter plots are the more informative areas of the piece. What you actually see is what looks like a log relationship of score to price when looking at the whole spectrum of processors. However, if you look just at the cheapest processors then the distribution appears to be more linear. On this basis you'd have to say that the best option would be to pick a processor which sits at the top of the linear section of the results, that way you get the best performance possible without having to pay over the odds for a 'bleading-edge' processor
- GentooRiverRat, on 10/11/2007, -4/+8I wonder if they took into account all the problems with the Core 2 Duo chips that are going to be exploited as soon as people know more about them: http://marc.info/?l=openbsd-misc&m=118296441702631
- palehorse864, on 10/11/2007, -0/+4Maybe you overclocked too much.
- Ramble, on 10/11/2007, -0/+3Are you stupid? The G5's suck ass. Compare a Mac Pro to a G5 Mac, see what beats what.
Also, games on PPC? Are you having a laugh? - Xyleene, on 10/11/2007, -0/+3So much hate for the overclockers from some? To each their own, but mine's faster :-))
- cliffhanger407, on 10/11/2007, -0/+3Wait, so performance and price aren't linearly related?! I never would have guessed!
Seriously... Yes, you get better performance per dollar for an old X2. Is it going to suck if you're trying to do anything CPU Intensive? Yes. - computerguydk, on 10/11/2007, -1/+4Dude, you can't benchmark the same programs on a Mac. A lot of the tests are from games that only run on windows. Where's your logic? What are you going to benchmark, photoshop? The powerpc processors that you are referring to are very fast, when you are doing photo/video editing, not for gaming.
- Massif, on 10/11/2007, -0/+3I was so close to buying a Q6600 recently. I would have been a fool. I'm definitely waiting until July 22nd.
- carl25, on 10/11/2007, -1/+4actually i can get my e6300 on 3.0 on the stock heatsink, stock grease and a only 2 fans in the case
- Xyleene, on 10/11/2007, -0/+3Ya, mine is rock solid for the last 2 months @ 3ghz (8x375). The cache limitation is about a 5-10% performance drop.. but at 3ghz that's 3-8% less than a Core 2 Extreme x6800 (2.93ghz) which is valued at 999$!! And @ lastrite's speed it actually beats the x6800 at stock! Mind you, you can OC the 6800 to 4ghz and beyond..
The e4300 is hands down, no questions, the best bang for your buck in the current market after that last price drop. - tomato3017, on 10/11/2007, -1/+4Ya really, give us one or STFU fanboy
- PATSCRU, on 10/11/2007, -0/+3yup, got my 2.6 e6600 (part of the top 5) up to 3.4 on air......could go more (that's what they all say, huh?) but talk about value. It'd be great to see overclockability worked into their tests somehow.
- jon26l, on 10/11/2007, -0/+2I've got it. It's also on the NGs
- AaronCompNetSys, on 10/11/2007, -0/+2Skip the same old CPU testing crap. Strait to the interesting part. (Power Efficiency)
http://techreport.com/reviews/2007q2/pricevperf/index.x?pg=10 - betterth, on 10/11/2007, -1/+3For the same reason Motor Trend doesn't remap the timings of the cars they're testing. Sure, you can tweak a product out, change how it operates to remove reliability/longevity and increase performance, but that misses the point of testing the product itself as the normal consumer will use it.
- mok000, on 10/11/2007, -0/+2I agree with your analysis, but not with your conclusion. In most cases, the E6700 seems to fall right on the curve, meaning that you get exactly the performance you pay for. E6600, on the other hand, consistently (well, almost) lies _above_ the curve, meaning that you get more performance than you pay for. Along the same lines (looking at the aggregate curve) X6800 is a bad buy, since it is below the curve.
However, as it's pointed out in the review, you should seek out your typical application and look at that data for that. Thus, for gaming, it makes sense to save money on the CPU (go for the 3600+) and spend the cash on the graphics card. - lastrite, on 10/11/2007, -0/+2Did the same but overclocked to 3.2ghz on stock cooling! :D
- vrillco, on 10/11/2007, -2/+4Like the last two posters said, Intel's prices are being aggressively reduced on July 22nd so this whole article is 3 weeks away from being obsolete.
- antiver, on 10/11/2007, -2/+4Here's what the article should have concluded with-
If you're going to mathematically analyze the results based on price vs performance, the E6700 scores the best in almost all tests. Draw an curve to interpolate the results (line of best fit). Based on the 10 or so cheapest processors, the performance curve should continue to increase rocket as it moves to the right (performance should increase as expected when you spend more, not quite linearly obviously, but a curve should compensate and show some realistic expected performance for your dollar. The curve compensates for diminishing returns and shows you the FPS you *should* get at a given price point.). Again, if you do this, all the "high end" processors fall way short of the mark. Yes, they're faster in many tests, but their not getting you the extra performance that you're paying extra for. In summary, the higher end processors either need to be cheaper or perform better in order to "win." Ergo, E6700 wins. I wish I had one. - carl25, on 10/11/2007, -2/+4the amd 6000 is great for the price. I've seen a few benchmarks where the amd 6000 beats the 6700 on stock speeds. And price wise the amd 6000 is friggin awesome
- Giga, on 10/11/2007, -0/+2So your box is built except for everything that matters? Wow, I must say that I am impressed. How is that empty case treating you these days?
- inactive, on 10/11/2007, -0/+2im just waiting on the quadcore intel chip to come down around the end of july its at 400 bux, once it breaks less than 300, its on!
- init100, on 10/11/2007, -0/+2Well, that must be the reason Apple still uses PowerPC processors in their products. /sarcasm.
- inactive, on 10/11/2007, -0/+2Wow, I didn't realise the mac fanboys were quite as deluded as opiophile seems to be!
The G4s stank compared to the A64 line, and even the G5s were trounced as well, especially by the dual core chips.
I do however, remember a few articles that thoroughly debunked apple's speed claims about their G4 and G5 processor based systems when they contracted VeriTest to do a comparison.
Apple/VeriTest only ran benchmarks which favored their own chips, and actually disabled some of the settings which would have allowed the AMD and Intel chips to run correctly, all in aid of making the PPC chips look far far better than they actually were (to say nothing of using compilers which are known to not do x86 code very well/efficiently).
http://www.theregister.co.uk/2003/06/24/apple_accused_of_cheating_over/
Then you get this ***** where apple fanboys would LOVE people to believe that these tests are fair, despite the video cards being 3 years apart in age and speed.
http://www.barefeats.com/voo5b.htm - inactive, on 10/11/2007, -0/+2God I love those Big Typhoon heatsinks!! Completely quiet, and phenomenal cooling!
- init100, on 10/11/2007, -0/+2No? Do you even know what Moore's Law says?
- inactive, on 10/11/2007, -0/+2Up to a point I'd agree, however you have to remember that all the game tests were performed at 800x600 (or close) and more gamers these days simply won't put up with such crappy resolutions. A better video card will certainly help, but you also have to have a reasonable CPU to get the best results.
I recently went from a X2 4400+ to an E6600 system but keeping the same video card (640Mb 8800GTS) and my framerates went up a noticeable amount. - harshbarj, on 10/11/2007, -0/+1Problem with overclocking is it's hit or miss. Not every CPU can be overclocked the same. The only fair / consistent comparison is at stock speeds (who wants to buy a CPU expecting to hit 3ghz and barley get over 2ghz).
That said I have an e6300 than can hit 3ghz easy on stock cooling but only do that when I encode video. in everything else I do the extra speed is wasted. - harshbarj, on 10/11/2007, -0/+1I don't see how. I went from a 3500+ to a c2c e6300 and kept my video card and could only see an improvement with benchmarks. Now in video rendering I saw a huge bump (from taking 5-6 hours to only 30min - 1hour). nearly every dual core CPU today will run today's games just fine. (as my old a64 3500+ with a x850xt runs bf2, hl2, coh at 1280x1024 max settings with no problems, i.e. staying above 30fps. just don't use it for rendering)
-
Show 51 - 100 of 103 discussions
Digg is coming to a city (and computer) near you! Check out all the details on our 
© Digg Inc. 2009
— Content created and posted by Digg users is