Slashdot Mirror
The Mother of All CPU Charts
||Plazm|| writes "Tom's Hardware has an entertaining read on the latest offerings from processor makers Intel and AMD. Not only does it contain a plethora of benchmarks on the latest Dual core CPU's, but it also includes benchmarks from over 60 other legacy processors. Better yet, they let the benchmarks speak for themselves and let you draw your own conclusions. You may want to fill up your 44oz mug before sifting through this one, though."
I'd like to see a comparison of average cost against the speed, since the real question is what's the fastest speed I can get for the money.
KeepTrackOfIt.com - Find the lowest gas prices in your area graphically
I didn't say that clock speed = performance, however, an increase in clock speed on the same architecture does indicate a probable increase in performance. I'm also sure you know that it is arguable at best that hyperthreading increases performance.
I'm also not sure where you got the impression that the article equated speed with clock rate - if they did, why did they bother with all those benchmarks?
As an example of the stagnation of the past few years, I have some code whose critical loop is unparallelizable (each instruction relies on the result of the previous instruction). My dual core Opteron running in 64 bit mode performs only about 15% better than my dual Athlon MP system from about 3 years ago. Synthetic benchmarks show a little more improvement, but largely only when memory access becomes a factor.
I'm not talking about servers, where parallelism is a necessity, or even general computing, I'm talking about unparallelizable, single threaded code. In this area, progress has been very slow. I'll grant you that the market is not as important in the scheme of things, but it is still there. Given how obsessed Intel in particular has been with clock speed to this point, it makes me wonder if they have gone to dual cores and such because they couldn't get more clock speed, which raises the question of whether we are hitting the physical limits of miniaturization.
are you kidding me?! Tom's actually got blackballed by Intel before for publishing stuff that didn't make Intel look good, and Tom's was pretty much the guys which demonstrated the flaws in the first 1+Ghz P3s, leading to Intel being forced to recall the chips!
Not only that, Toms were the first guys I read that actually went to TIME the FSB frequencies on various motherboards, thereby discovering some manufacturers were cheating by "pre-overclocking" their boards so they'd look better on benchmarks against others!
"Does anyone else find it odd that in practically all the benchmarks, the single core processors beat out the dual core processors?"
No more odd than say a pickup truck beating a sports car in a race to move from one apartment to another.
"Derp de derp."
You seem to have fallen prey to the alpha geek mentality. The implication of your post amounts to:
"It works for me. If it doesn't work for you, you must be stupid".
There are people with drives out there that have had to install hard disk drivers using a floppy drive. A floppy drive is a very cheap piece of equipment to add to your PC, even if it's just as a contingency. It makes no sense to save $10 on a computer that might cost anywhere from $400 to $4000, when there might be situations you might need that $10 piece of equipment.
For the record I've also seen mouse drivers and other hardware come only on a floppy disk. Granted most of the time you can download from the net, but compared to pushing a disk that came with your product into a drive it's more work.
What I'd like to see is all hardware and software manufacturers stop distribution of their drivers and software on floppy. It really is an ancient piece of technology that should be dead by now. Sadly it's not.
These posts express my own personal views, not those of my employer
his chart even does not cover all AMD and Intel CPUs. For example, processors like the AMD 29K, Intel i860, i960 and the Intel Itanium are missing.
You missed a big one: The Celeron. How come not one single processor from Intel's budget line was tested?
One word -- Dell. Dell sells nothing but Intel chips, and they're the biggest PC manufacturer. Plenty of manufacturers got a bad taste in their mouth after the K5, K6, and K6-2. The chips/chipsets were not stable, and prompted a lot of returns. Some of them are starting to sell AMD chips again, but it took a long time to regain their confidence.
I used up all my sick days, so I'm calling in dead.
You laugh, but this is how I keep my dorm room warm....
To be or not to be: There is no maybe.
My guess is, it is probably compiled with the Intel ICC compiler...
- intel.html
http://www.swallowtail.org/naughty
I'm probably not the only one on slashdot who tends to accumulate old hardware. As I think many of the collectors here might sympathize, often the pieces are obscure/generic brands for which not much information is available. In fact, even some of the brand names often no longer have any documentation. Locating drivers has become an easier task, but performance features and such are still hard to find. I have an old 486 class motherboard which I found out (only from talking to older folks) was very well regarded by those who used it. But it was very hard work trying to find anything to coroborate those casual conversations on the web. Charts and tables like the ones compiled by THG can give a good idea about which pieces should be paired up for an appropriate system that avoids unnecessary bottlenecks. I really wish this kind of information exists for soundcards, video cards, and modems in the form of a giant database of hardware products from the past as well as the present.
Stay sentient. Don't drink bad milk.
I was looking at this very recently. First, most of the tests are single threaded, so the dual core is of limited use. Secondly, and more importantly, the individual cores run slower on the X2 processors than the single core does. For example, the AMD64-3800+ runs at 2.4GHz; the X2-3800 runs two at 2Ghz. For single threaded apps, like current games or benchmarks, the single core chip will beat the dual core hands down. The AMD64 3800+ is cheaper too.
For general heavy use on a non-gaming rig, you're still better off with an X2, as the dual core really comes into it's own as one core can concentrate on say, rendering or encoding, while the other core is available for other high-CPU apps.
Remember kids, it's all fun and games until someone commits wholesale galactic genocide.
When I read that word, I thought there would be comparisons old CPU's such as 486s, then I started reading TFA and it talked about the evolution from 1993 till today, so I though "so it will be from Pentiums on, ok". Then came the benchmarks and the only one "legacy" processor is an Athlon Thunderbird 1400, from 4 or 5 years ago. The rest are still sold today (well, some cores might be difficult to find, but Semprons and Athlon64 3000's are definitively still sold).
I think the article should better read "from over 60 other recent processors".
While there are a lot of problems with the DMCA, I don't see how it can be used to suppress publishing of benchmarks.
I know that some software EULAs have contained a 'do not benchmark' clause, but whether such a restriction would stand up in court is, as far as I know, not been tested yet.
If J.K.R wrote Windows: Puteulanus fenestra mortalis!
Tom's hardware makes the list.
It's a massive undertaking to create it.
That's news.
Is it really news every time they update the list?
Yes, this is news. This is the time of year that many people, myself included, plan on buying computer upgrades. Based on Tom's charts, I can see my (older) CPU, compare it to newer CPUs using the video card, memory, etc. as a control, and decide if the upgrade is worth it. Although Tom talks about the latest and greatest all the time, only once or twice a year does he put things in perspective with older hardware. Personally, I want to see the same thing but with video cards, because Tom's article showed me that upgrading my CPU isn't worth the money.
Besides, Tom is talking about computer hardware. Nerds (myself included) love this stuff. So yes, this is news for nerds. And it does matter.
24 beers in a case, 24 hours in a day. Coincidence? I think not!