jmke

Your comments on this little piece of writing here

what piece? this one!

__________________

piotke

n/o, but you compare the slowest p3 with the fastest (widely available) p3. What about the p3 500, 550, 600, ....

there were also p4 2.4 533 and 3.06 533, so you could make also a bigger jump..

If you take a look at the impact of the upgade, you're correct...

450 --> 1 Ghz is +50%
2.4 --> 3.06 is less

But in general it's an interesting read !

__________________

jmke

sure there were a lot steps , but compare with the older series

P1 - 70 90 120 133 166 200 233
P2 - 233 250 300 333 366 433(?) 450
P3 - 500 550 600 650 700 750 800 866 900 1000

(there are more but don't remember them all)

now take the P4 - 1.5 1.6 1.8 2.0 2.2 2.26 2.4 2.53 2.6 2.66 2.8 3.0 3.06 3.2 3.4

and then I'm still forgetting some, just look at what they are cooking up for the new prescott series and their Pentium M series
over 20+ models, each differ only a few % in speed, it's a waste of marketing money


about the upgrade, I'm about going from a P4 1.5 to a 3.0

__________________

Sidney

This is what business is looking into when new computer purchase is being considered. People associate in IT and IT Purchase always argue these points you mentioned.

I am not talking about gaming in graphic card; rather 3D engineering app demands as much as gamers.

The trend you indicated has a lot to do with higher production yields in chip making then yesteryears; sales/marketing approach in promoting higher speed plays a key role in the rapid new CPU introduction for both Intel and AMD.

Bottom line; upgrading on an annual basis could never achieve the best ROI in business formula; but acceptable to enthusiasts who want the bragging rights. Once, people go over the ego trip; upgrading will slow down.

I am from the very old school, hence my graphic card is at least two years behind. I admire Dan Rutter's stand on his consistent view point (dansdata.com), CPU upgrade consideration should be at least 50% improvement over the previous one. For that matter, any hardware upgrade should meet the same parameter.

May be John could look into the parameters in acquiring upgrades. This will certainly be an interesting topics.

__________________
lazyman

Opteron 165 (2) @2.85 1.42 vcore AMD Stock HSF + Chill Vent II

jmke

thanks for the valuable comments Sidney!


that 50% speed increase rule is great, but as time goes buy, you'll have to wait longer and longer for that to happen


the only thing we really NEED know to take things to next level is very fast large storage media for everything BUT gaming.

very fast GPU action for gaming.

CPU and Memory is speedy enough at this time, it's the other two who are lagging behind. Sure the NV40 can show you that at 800x600 your system is "CPU" limited, but if you pump in Doom 3 like quality with 10001 shader effects to create a blistering image of super high quality with immense high resolution, then your GPU will sweat, even at 640x480 (like the Ti4600 which can be easily killed by the D3 beta leak)

__________________

piotke

yes, but they weren't made at the same time ...

you can in my humble opinion compare
1.6 1.8 2.0 2.2 2.4 2.5

then you got

2.4 2.53 2.66 2.8 3.06

and now

2.4 2.6 2.8 3.0 3.2 3.4

so it are actually always steps of 133 mhz of more. Your line of cpu's is correct, there's always about 8 months in between a news serie, while 2.4 (533)-->3.06 came ALMOST at one..

__________________

jmke

doesn't matter, speed increase between the revisions was at max 10% and then some.

10% is nothing, because then you're only talking CPU, if you throw in the rest, you still got a rig which has been barely made faster, although you just spend $400 on a new cpu and mobo, and in my case if I upgrade to DDR1, RAM.

Upgrading to DDR1 for me is only valuable if I'm going the AMD road, if the Pentium M marches forward and is able to pick up speed, then we'll have PCI-X and DDR2 boards. Something Intel should have done 9 months ago, instead of trying to PresHott us.

__________________

Bosw8er

I don't believe in investing in raw cpu-power :
1) too expensive (horrible price/gain ratio)
2) the amount of cpu power ain't necessary in 90% of daily pc-labour
3) you won't notice any difference at all (except from benchmarks) when you bought your latest pc within one year or so
4) worst investment you ever made because the cpu ain't worth half of the price you bought it for ... one day after you bought it

I wouldn't invest one cent in upgrading in cpu-power if i had a PIV or XP 2600 to 3000 coupled with PC2700 and a recent mobo. Mainly because i wouldn't notice the difference.

I'll stick to upgrading on a budget friendly way, according my needs.
Like an Ati 8500DV : expensive 500EUR card, but it has so much bells and whistles i think it was even cheap. I have had it for about 3 years ... no need to replace it ... for the next 2 years.
If you have the urge to upgrade (and the money) i would rather go for raptor disks (or go for raid), 17 or 19 inch LCD screens, a quality nicelooking case, upgrade your graca (if you're into gaming), invest in a second or third pc for backups or testing, ... or other valuable things that will keep you happy for several years ... in stead of several hours.

__________________
"Think of how stupid the average person is, and realize half of them are stupider than that."

Sidney

The difference between WANTS and NEEDS.

Perfectly makes sense in 19" LCD or 22" CRT; better looking case to please my eyes. As for FAST harddrive, I could see at work when 1 to 3 GB of 3D pipings data in the plant; Autocad drawings with GB data points at Boeing; MRP run with over 100,000 sku's in BOM explosion using level 1 MRPII planning where HDD speed would make a huge difference.

Office apps; forget it. BUT, hardware manufacturers know exactly the majority customers weakness ..... WANTS overshadow NEEDS.

Intel has us by the balls; if fast cars could sell; fast "speed" processor will too. It will be un-nature to go the other way. AMD marketing folks made a BIG mistakes after the success of K6 when they gave up the OEM market and turned to the enthusiasts market which by far small in SOM (share of market) in total by lowering the price point only creates the noise and not the wallet.

Bottom line; John has a good idea with this particular topic. It will become an endless thread. Or, a thread that goes timeless.

__________________
lazyman

Opteron 165 (2) @2.85 1.42 vcore AMD Stock HSF + Chill Vent II

jmke

end of this year we finally double the SATA speed
http://www.xbitlabs.com/news/storage...420164544.html

__________________