Intel & ATI naming schemes?

Dario D.

Gawd
Joined
Dec 8, 2004
Messages
582
Hi all. This has been bugging me for some time. I can understand nVidia and AMD naming schemes easily. It's so direct. nVidia is just 5000 series, 6000 series, etc, and AMD has the easy to follow 2000+ thing.

But I can't get a clue how to follow ATI video cards and Intel CPU's. When I see each new generation, the product names look to me like randomized math problems or something, like ATI 1x21800, or the Intel Another-New-Series E12345-DP or something.

The above examples are probably unfair, as I'm sure there is logic to their naming schemes, but I simply can't figure out how to follow them. Can someone enlighten me as to how their naming schemes work?

Thanks. There are also other people near me who have no clue, but can folow nVidia and AMD perfectly (and we are regular nVidia and AMD customers for that reason - we just don't understand ATI or Intel)
 
:rolleyes:

ATI:
x300
x600
x800
x1300
x1600
x1650
x1800
x1900
x1950

The higher the number the better, i dont know why its so hard to follow compared to nvidia's naming scheme, They are very close to the same thing. And of course you stick on all the letters at the end just like nvidia, XTX is atis flagship model such as GTX is nvidia's

Intel:
Core 2 Duo E4300
" E4400
" E6300
" E6320
" E6400
" E6420
" E6600
" E6700
" X6800

As with amd, nvidia and ati the higher number the better...

btw wrong forum?
 
I agree with much of what you're saying. AMD got lots of criticism when they used "PR" ratings instead of clockspeed numbers, but at least it was an accurate measure of performance. Intel abandoned that, to my dismay. How am I supposed to know how the 6600 compares to the 6400 without any other specs? And the alphabet soup that goes with video cards (of both flavors) drives me nuts. You have GS, GT, Pro, XT, XTX, GTX, GX2, GTS, LE, SE, in combination with the numbers, which aren't too bad themselves (7300/7600/7900 or X1300/X1600/X1900).
 
How am I supposed to know how the 6600 compares to the 6400 without any other specs?
That is the whole point of it. Transparency is bad for the seller, good for the buyer. If Intel obscures the performance, it makes it harder for the buyer to make an "informed" decision.
And the alphabet soup that goes with video cards (of both flavors) drives me nuts. You have GS, GT, Pro, XT, XTX, GTX, GX2, GTS, LE, SE, in combination with the numbers, which aren't too bad themselves (7300/7600/7900 or X1300/X1600/X1900).
I get the numbers, but I have no idea how fast an x7850GTSEProX2LE is. Heck at times I don't even know if it is better to get an LE edition "top end" card over a Pro edition mid-range card. Which goes to prove my point about transparency being good for the consumer.
 
This is why we read reviews and news articles. If you don't care to research the parts you're buying, then just go for the higher numbers, lol. This way, you spend more than you should, and get performance you most likely don't need.
 
This is why we read reviews and news articles. If you don't care to research the parts you're buying, then just go for the higher numbers, lol. This way, you spend more than you should, and get performance you most likely don't need.

i think that is the point of this thread :)


and im with you, i dont understand intel's naming scheme for shit. ATI is good, its just when they went to the x series. See, first it was radeon 8xxx, then 9xxxx and the Xxxx where 'X' equals 10.
 
Intel uses a similar scheme to BMW:
First digit denotes series, higher series is better
3xx is Celeron
5xx was Single core (Prescott)
6xx was Single core v2 (Prescott, Cedar Mill)
8xx was Dual core (Smithfield)
9xx was Dual Core v2 (Presler)

9>8, 6>5, Anything>3

Second (and third, in multiples of five for Celeron) denotes rank within the family:
335 > 330, 34x > 33x
54x > 52x
84x > 83x

Third digit denotes revision, features
332 (64-bit 65nm) > 331 (64-bit 90nm) > 330 (90nm)
651 > 650

Now it's the same with Core2
'e' means mainstream dual-core
'Q' means mainstream quad-core
'X' means Extreme, Enthusiast, etc

First digit denotes family:
4xxx is low-end,
6xxx is high(er) end

Second digit is rank within family
44xx > 43xx
67xx > 66xx

Third digit is revision
6350 (1.8'ish GHz, 1333FSB, rev2) > 6300 (1.8'ish GHz, 1066FSB, rev1)
 
And seventh digit is me buying an AMD :D

(TY for the enlightenment)

That is the whole point of it. Transparency is bad for the seller, good for the buyer.
I don't know how it helps. There are 3 significant types of consumers that I'm aware of- 1) people who just need a PC, so they buy one, 2) people who know what they're doing, and buy the video card they intend, 3) people who THINK they know what they're doing, and buy the wrong video, then return it when it ends up sucking. Of the three choices, I think marketing transparency only benefits from the third, and probably not by that much.

I remember being at Fry's one day back when the GF 5000 series came out, and overhearing a few guys saying they should get a 5200 LE because it was supposedly "way better" than their GF4 Ti 4600. BS marketing tactics won that day, but I'm sure they were back within minutes of realizing that performance was worse with that card. I think most LE's and other BS-marked cards suffer similar fates when purchased separate from a computer. When someone picks one up just because the number is high, I'm sure it's a matter of minutes before they return it, seeing marginal or no improvement (after perhaps reformatting their hard drive, as some tend to do at the slightest hint of trouble).

This marketing transparency problem has also resulted in about 15 AMD/nVidia purchases among computer people near me, and not a single ATI or Intel. When it comes to making a choice, people want to buy what they can understand, right? Looking at my stash of receipts, and calling the mind the specs of those near me, that's at least *my* experience.
 
The higher the number, the better the performance, as long as you stay in the same class... which holds true for almost everything: GF 7200 > GF 7100, Radeon X1600 > Radeon X1300, C2D E4400 > C2D E4320, AX2 4200+ > AX2 3800+... its all relative.

So you're saying AMD's naming convention makes more sense than Intel's? The funny thing is, AMD's naming convention is based on an obsolete Intel naming convention (no longer used), rendering the AMD naming convention somewhat useless, lol. Intel's naming is simply a make/model now; but as outlined above, its not that hard to figure out which is better than what.

ATI's is still the same, they're just using "X" to represent "10" for everything after the 9000 series. So, the old 9800's successor is the X800... and it's successor is the X1800... its not much different from NVIDIAs. It makes me wonder what NVIDIA will do 2 generations from now, hehe.

If you want AMD, then go ahead and buy AMD. You'll get good performance for your money. However, if you want the best performance for your money, you'll have to do a little more research. Currently, any Intel Core2 CPU is better than its AMD counterpart. Did you need some good review sites so you could do more research?
 
So you're saying AMD's naming convention makes more sense than Intel's? The funny thing is, AMD's naming convention is based on an obsolete Intel naming convention (no longer used), rendering the AMD naming convention somewhat useless, lol. Intel's naming is simply a make/model now; but as outlined above, its not that hard to figure out which is better than what.
I, for one, prefer the current AMD numbering scheme. So the original basis for it is now outdated, so what? So is the basis for most of the scientific units we use today. The point is that it is a standard that can apply for a good long while, regardless of architecture. Joe Sixpack can tell that "4400" is roughly 10% faster than "4000" without knowing that the first digit means one thing, the second means something else entirely, and the third is something that he doesn't really care about anyway.
 
Back
Top