GTX 260 gets a tune-up
WE WERE S(TROLLING) through our massive list of feeds today when (at least) three sites made our eyebrows raise. Three reviews pitting the GTX 260 Core 216 against an HD 4870 1GB… wow, what a coincidence.
Sarcasm aside, at least Bjorn3D did a good job at not making it look like a build-to-order review. In this particular article, they take five games you’re likely to pick up at Christmas and run them on both a GTX 260 Core 216 and an HD 4870 1GB. Ah... nuthin' like a reality check…
Fudzilla, also tested the GTX 260 Core 216 with the new 180 drivers and saw just how much of a spanking it delivers. Amazing what a driver tune-up will do to your graphics card.
What’s that? Hardware Secrets has the same review? Ah… let’s see, same games, same hardware… reviews off the peg anyone?
Tom’s Hardware has a quintet of DTRs taking up most of the real-estate at the office. You’ve got anything from 17-inch to 20.1-inch DTRs there, so just how multimedia mad are you? Sony loses all-round, while Acer seems to come out smelling like roses. Sounds a bit useless, but they did test battery life, mind you… get it here.
Hardware Logic is doing the RAM tour once again, only this time it’s G.Skill’s PI Black 2x2GB series DDR2-800. Ultra-cheap, high performance, that’s all we have to say about this. We’re sure you’ll agree once you read the review…
Walton-Chaintech (yes, yes, we know it sounds like a brokerage firm, but it’s still good ol’ Chaintech), has started retailing APOGEE-branded SSD discs, according to TweakTown, who happens to have a first review. This is a MLC-based drive with 64GB of storage that features very high read and write speeds. Price is in line with its competitors, says Chris.
Xbit Labs has something on its mind... literally. Oleg is testing the OCZ Neural Impulse Actuator – or NIA for short. This product has been talked about for quite a while, at least a year now, and this is the first serious test we see. It works - sort of - says Oleg. While doing simple tasks was quite rewarding, more complex things like FPS games become impossible. Still a gadget, he thinks. Read it here.
Just to tease your fanboi tastebuds, we’re remanding you to PC Perps as a final roundup link. Ryan was invited to an AMD shindig where dessert was served at -196C. Apparently hacks were treated to a 6GHz Phenom II on liquid nitrogen, just to prove that AMD has some headroom for its 45nm marchitecture. Tasty. µ
Share this:
ShareBusiness Intelligence Software
Customer Relationship Management (CRM)
Enterprise Accounting Software
Enterprise Resource Planning (ERP)
Enterprise Systems Management
Why do you gents keep linking to "reviews" that start bar graphs at numbers other than zero, so as to mislead the punters that read them?
I am, of course, talking about bjorn3d. The graphs made it look like the nvidia card has double the performance of the ati card, when the real difference in performance was much, much smaller...
I can't help but think if it was an ATI card spanking NV then you'd be singing praises. Not looking for any reason to poke holes.
I sold my GTX280 off when the 4870X2s arrived. And boy was that a mistake, with websites like these who needs to check the facts about ATI cards. And now with new drivers from nvidia they take the crown right back with their cards. When Nvidia releases their new cards in 2009, i'm buying one.
Agreed. While the Nvidia card did have a decent lead there was no purpose to set up the graphs like that other than to mislead.

Why don't they just make the graphs much smaller and start them in the -100 fps range? Makes about as much sense.

I don't think I'll trust another review by those guys.
Agreed the graphs alone may look like an inpressive lead but there is also accompanying text which explains quite clearly the margins. To be blunt, the graphs also have things called axis which clearly have the numbers marked. Using "0" as an origin is pointless. The larger the value the more abusrd the graph looks with a "0" origin. It's also about logic, a bar graph demonstrates absolute values whereas a "0" origin curvilinear graph is better for logarithmic non-linear relationships. And one more thing, it is about time there was a gracious nod from the 'inq' about the performance of the 260's. 4870's are very good cards but so is the 260 and people have lost track of that simple fact in the red/green war of fanboyishness.
I like how the article is all about NVDAs new drivers giving better performance. Make's u wonder why fuDzilla ends the article with :

"It's worth mentioning that ATI will have some new drivers of its own in the second week of December, but we are not sure if this driver can change the odds to Nvidia's advantage."

lol
Both the 260GTX and 280GTX are single chip cards. ATI had to put two of their best together the even compete at about twice the power consumption and heat. Nvidia is kIng just Like intel.
The 260 gives the best bang for the buck falling only slightly behind the the monster 280GTX by only a few FPS sometimes.
You want good graphs and honest comparisons go to Tomshardware.com.
Isn't the "Neural Impulse Actuator" the thingy in lieutenant Data's positronic matrix that allows him to get a boner at the site of conuselor Troi?
Displaying benchmark graphs without the origin clearly stated, or without showing the whole bar, is something that has been done since probably before 3Dfx lost it to Nvidia.
The problem is not actually the size of the gap, or the width of the bar, it is that people will not engage the brain when looking at the graphs. Instead, most will just see a humongous lead on the chart in favor of one of the competitors, and blindly decide that that card is the best, even if the lead represents nothing but a 1 or 2% difference.

It's the ability to think critically, to analyze, that is needed.
Having an explicit origin on the graph is nothing but eye candy.