Can you do some idle power testing in the review to show what the consumption would be -IF- the powerplay was working? I just wanna know how low it can go :D
Printable View
would be interesting if there will be non reference card with just 4 RAM chips, maybe power consumption would be better and there would be less of those repeating post about it :)
My guess is that the 4750 is the low power consumption alternative since it got GDDR3.
What if the 4750 matched the 9600GT in cases where bandwidth limitations weren't that severe?
I heard that was their target previously, 9800 for the highend version and 9600 for the GDDR3. Granted, the lake of salt is right... there.
The HD4850 would be about 10-15% slower if it had GDDR5 instead of GDDR3 - i'd say it's relevant
erm... no.
Those graphs are comparing GDDR3 with GDDR5 at half the speed of GDDR3. For example, in that case it's 993Mhz GDDR3 versus 497Mhz GDDR5. However, it would be pretty stupid for the HD4850 to have 500Mhz GDDR5. If the HD4850 had GDDR5, it would be 750Mhz minimum.
No, it's given the same.. let's call it the same "single data-rate clockspeed".
It's always been like that. DDR was already slower than SDR if DDR was underclocked to the half of the speed of SDR.
Well.. ATI has launched their share of highly bottlenecked cards before, like the HD4650 DDR2 (phail...).
Or even worse: the FirePro V3750, which is a 320sp RV730 with a 64bit memory controller (yeah, the original 320sp R600 had 8x more bandwidth).
That's exactly what the dude is saying. GDDR5 will be slower than GDDR3 at the same *effective* clockspeed. Ergo, while the 4770 can get in the same ballpark as GTS 250 in terms of bandwidth, it's memory subsystem will still probably be slower on the whole. Of course, take all this speculation with a grain of salt as the GTS 250 is a lot older, the memory controller in the HD 4770 is tweaked vs. the 4800 series, and the less complex 128-bit interface might result in some kind of lower latency (whether or not it would be significant, I have no idea)
Not exactly. That is true given the same bandwidth but maintaining the timings required for 1000Mhz (GDDR5-4000) operation. Running at 500Mhz will tolerate better timings thus increasing the performance.
Anyway, at the same BW GDDR5 will be slower, but it's far from 10% with adequate timings.
In an upcoming article, yes. That will be coupled with some additional cooling tests since right now there is no way to accurately test the heat output of the 40nm core against any other card due to the oddball offset of the heatsink mount. After some preliminary tests with some modded heatsinks, I am certain the results will shock many people who are defending the move to 40nm.
Did you get 2 cards to do some crossfire testing SKYMTL?
As I stated above, getting even one card was a stretch this time. Without ATI's support, websites have to go either to board partners (who didn't actually get the cards until last week) or to our contacts in Asia who may or may not get it to us in time. As it stands, I have one and will probably receive another a bit later in the week.
Sorry asking something you already posted :(
I'm really curious as to how the crossfire will perform. If a single card perform almost on par with a 4850, in crossfire they surely be close aswell, right? I'm either getting two of these or two 4850's. Not sure which will be better. I't s pity the stuff is so hellishly expensive in South Africa compared to the USA/UK
The only people who should be defending 40nm at this point is ATI, because they can put more cores into each waffer and start testing and tweaking this new process.
No "new" process has ever been instantaneously beneficial for power consumption and overclocking capacity, as far as I know.
I think you got things backwards.
AMD doesn't target GPUs, it targets price-points. What chip is in what graphics card and how old the architecture is, is completely irrelevant for the end user. What matters is what performance+features you get for how much money, period.
And the HD4830 is quite successfull at its price point. Where I live, the HD4830 is priced at the level of a 9600GT.
He did say bandwidth, not clockspeed. If he "meant" clockspeed, all is well :cool:.
You can't really draw that conclusion from the data we have. First of all, Damien's test underclocked the GDDR5 way below its normal operating parameters. Secondly there are a host of other variables that affect the efficiency including the sizing of buffers and command protocols used by the memory controllers.
So simply underclocking GDDR5 to half its speed and saying "see it's slower than GDDR3 at the same speed" isn't a very scientific test. Or in other words there's no way we can say that RV740's 72GB/s isn't as good as G92's 72GB/s.
I agree but they are marketing it as efficiency personafied.
Upon its release the HD 4830 was in the same price point as the 9800 GT. To the dollar here in Canada. It was only through the latest round of price reductions that the HD 4830 is now able to compete price-point wise with the 9600 GT.Quote:
AMD doesn't target GPUs, it targets price-points. What chip is in what graphics card and how old the architecture is, is completely irrelevant for the end user. What matters is what performance+features you get for how much money, period.
And the HD4830 is quite successfull at its price point. Where I live, the HD4830 is priced at the level of a 9600GT.
Pricing based on a feature list is completely understandable but there comes a price point where people start looking less and less at features and more about gaming capacity. To me, that is right around the $100 - $120 price points and above since below that you can get the same features (albeit less gaming potential) for much less money.
You mean the marketing guys are trying to deceive people?!?? Now that's a first!
;)
So you agree with me when I say the HD4830 is, right now, a successfull product in its price range? Of course, a RV770, even a salvaged one, should be more expensive to make than the G94b, but that's one of the main reasons for the RV740 appearance anyways.
The point I am trying to make is that it is up to reviewers to bridge the gap between marketing mumbo jumbo and reality.
Right now; definately. Especially considering the fact that Nvidia has been caught with their pants down with overly expensive 9800 GT and 9600 GT cards. Back when the HD 4830 was retailing for the same price as the 9800 GT, they had a reason to keep the price at ~$120 but for some reason there has been no reactionary price cut by Nvidia.Quote:
So you agree with me when I say the HD4830 is, right now, a successfull product in its price range? Of course, a RV770, even a salvaged one, should be more expensive to make than the G94b, but that's one of the main reasons for the RV740 appearance anyways
My retailer contacts continue to tell me that the 9800 GT sells SIGNIFICANTLY more units than the HD 4830...regardless of the near $30 difference between the two cards. This tells me that Nvidia knew their card would continue to outsell the HD 4830 at a higher price and didn't bother with a price cut in reaction to ATI cutting prices.
We all hear about the high-end price cuts but it is the mid-range that is really suffering price-wise these days.
Not only my retailer but people at retail level I am in contact with worldwide. That goes for the EU, USA, Asia and Canada. Even the board partners who support both Nvidia and ATI back this up off the record.
The simple matter is that Nvidia seems to have better market penetration through a very aggressive sales / marketing team. I am not debating which card is "better" even though for its price, the HD 4830 wins hands down. It is just a problem with people's perception of a certain product that will have them holding back purchasing a less expensive yet just as capable alternative. We see this with everything from running shoes to GPUs to the whole Mac versus PC war. The fact that this same thing is happening between ATI and Nvidia should surprise no one.
It probably has something to do with the huge splash the 8800/9800GT made when it was first introduced (beating the old GTS cards at a much lower price) vs. the 4830 being more of an afterthought designed to improve yields. If the 4830 was the first RV770 product to come out I bet it would have made a much bigger splash and get more mindshare. Still a pity that a (slightly debatable) better performing card that sells for less money is getting the shaft from the consumer. Same thing happened through most of the P4 generation too.
Thatīs simply impossible. Your contact is 0.0001% of the world. Don't make up numbers that anyone know about.
I can come here and say the same but on the oposite. The diference is that I don't say because I know that those numbers are impossible to know except if you count them on TSMC and UMC fabs.
Probably used the same method the linked to review uses.
"Catalyst Control Center is the only utility to adjust clocks for now. The core/memory clock limit is 830/850MHz, which doesnt meet our demand for overclocking. If you want more, just open the Rivatuner.cfg, find RV770 = 9440h-9443h,944Ch, and add the ID 94B3h after."
So I'm the one that has to prove it? :rolleyes::confused:
Someone come here and Say:
- I have 10000000000000001 contacts and in every world "X" happen and "Y" don't happen.
I Say:
- That's simple impossible. The only way to know if in this case "X" or "Y" happen is if you have direct numbers from TSMC or UMC that are the source. If you don't know that you can't predict numbers because my local store say this or that.
I'm not even going to talk about the 100000000 of local stores around the world, and not even going to talk about thousands of OEM makers around the world and I'm not even going to talk about the dozen of different add-in-board and retail channels.
Anyone here grown-up yesterday, and this is takes you to the coffee talking and Fud talking.
IDC, Isupply and other spend thousands of $$$$ to count market share numbers around the world. And even them canīt tell you numbers with good accuracy. They follow some standard functions to calculate numbers.
Take console numbers as example (microsoft, nintendo, sony) that looks easy to count market share numbers, but in reality those company's spread fud around their sales. Again if you have access to manufactures numbers you could easily say who is selling more.
In the end he is only spreading fud in a way that fud does. How many times you read on Fudzilla site (X is selling realy well) and you open the link and you have a bunch of nothing....
It's simply really every gamer knows nvidia, but only computer enthusiasts know ATi for the most part. Also over here there's this myth floating around that ATi has bad drivers, which in my experience is not the case. I had convinced a friend to buy a 4850, despite the sales person telling him that ATi had bad drivers. Well he tried to play GTA IV where he had texture problems. Next thing you know he gives the card back and pays extra to get a OCed 9800GT
Not going to discuss "Coffey chat", like: I have a friend that know a friend that bought one FX 5200 instead of one HD 4870 because the number was bigger or bought 8600GT 1GB instead of GTX260 896Mb because it have more memory (both cases are real and the only thing they show is how ignorant or stupid humans can be). Not going to discuss also: I have a friend that know a friend that in his local store sell more "X" then "Y".
You should listen to yourself. It's ridiculous read those kind of posts here on XS. (same opinion about those posts in favor of Nvidia, intel, amd, microsoft, sony, watever)
im assuming these are ppl who had to buy a card cause their current card died, not because they were just upgrading.
when your upgrading, you have time to do your homework and stuff.
its really not too difficult with google.
I find it easy to believe the 8800 sells better. Nvidia has way better marketing and I can imagine alot of older 8800 gt owner recommending new computer owners to go 8800gt or 9800gt. TWIMTBP also probably has an effect too since it is in game advertising.
Yeah, 9800/8800GT selling better isn't too surprising, but I bet the 4770 will overtake it.
Yeah, also the fact that the HD4830 wasnt really advertised as much and was mostly overlooked
on the other hand, 4850 and 4870's hve sold admirably well
4770 is a very important GPU for AMD. whats the die size of this GPU?
Something like 136mm2
Still waiting for our 4770's, will do some CF & OC testing when they turn up.
One thing I've noticed recently is that some 9800GTs are showing up with no PCI-E power connector (slated GTS240 spec?) so for power efficiency and as an upgrade path it is a good product and can probably remain competitive. But the systems that cards in this range suit are more often CF than SLi compatible which might give 4770 the edge it needs.
I can't wait for these cards :D:D
Already figured out a voltmod (the controller looks VERY moddable :D)
I'm just hoping this won't be a paper launch, and that they sell the reference design with the reference sink here. Shame about the mounting holes though.
You can't come and say that he is wrong, without showing anything to the contrary. Down here its pretty much the same story, stores sell more nvidia cards because nvidia is well known. I would be honestly surprised if ANYONE claims that they work in a regular PC store and they sell more ATI than nvidia.
You don't have to go to the source and ask "Hey, TSMC, how many G92B/RV770 did you sell this month?"
Talking about retail channel, it is correct to take random samples of retailers ask them if "X" sells more units than "Y" and conclude wich one is selling more units. That would be correct in a statistics point of view, the more samples you get the more accurate the result will be.
It is not impossible what SKYMTL said, here in Argentina we have the same tendency.
You can buy these on newegg now.
http://www.newegg.com/Product/Produc...n=4770&x=0&y=0
Ordering from the egg right now!
I wish they hadn't junked the better looking better performing stock cooler AND raised the price $10 which is 10% for this segment.
If it goes down after release or a little later that won't be bad but it's gonna be butting heads with 4830/4850 a little on price at this point.
God those coolers look hideous to me....do they even have ram cooling on? or mosfet cooling?
What's the deal with the coolers anyway? Is this just the first batch that AMD wanted to get out the door? I've seen other pics with much better looking hardware on there.
The price increase is likely Newegg's typical +$10 on release markup
No paper launch. Initial batch will be with these coolers-
http://en.expreview.com/img/2009/04/24/HD4770_14.jpg
Suppose to have a $20MIR once AIBs get their own designs out.
Yep.
Other than the other cooler exhausting air out the back of the case what is so good about it? We don't know how it performs.
NewEgg's marked up price does make the cards value suck. They have a 4830 for $86 and a 4850 for $119
It IS efficient for THEM (mm^2). It WILL be efficent for us later on. (4750 + DDR3 // Mobility Radeon 4830).
The GPUs themselve use less power than ever. (though the Furmark load test conducted by Expreview that power usage doesn't overflow TDP this time round) It's just that they're unfortunately saddled with standard-voltaged GDDR5 (easily overclockable back to spec 1250Mhz) instead of nVidia's favorite high speed GDDR3.
But fine, since you consider having 9600M GT (9500/8600GT) performance the pinnacle of what a notebook can have and no further improvments are needed, so RV740 is pointless. :rolleyes:
You seem to have big issues with ATI making new chips for predominantly different (OEM, Mobile) markets with desktop as a secondary target while having literally zero issues with nVidia making new, overlapping SKUs. Whatever, your personal vendettas are up to you.
p/s: I don't agree that the 4770 is the perfect chip especially with current market conditions or the pricing on it. I don't like the fact that the stupid GDDR5 memory controller retrain/flicker issue still persists. But at least ATI is doing something instead of just shipping their restickered SKUs to reviewers ASAP before launch day so they can get hits. Ultimately it's still up to the product itself.
http://enthusiast.hardocp.com/articl...50aHVzaWFzdA==
http://www.guru3d.com/article/ati-ra...d-4770-review/
http://www.hardwarecanucks.com/forum...r5-review.html
http://www.hexus.net/content/item.php?item=18202
http://hothardware.com/Articles/ATI-...4770-40nm-GPU/
http://www.legionhardware.com/document.php?id=826
http://www.pcper.com/article.php?aid=700
http://www.techpowerup.com/reviews/ATI/HD_4770/
http://xbitlabs.com/articles/video/d...on-hd4770.html
http://www.neoseeker.com/Articles/Ha.../hd4770launch/
http://www.tomshardware.com/reviews/...4770,2281.html
http://www.pcgameshardware.com/aid,6...00-GT/Reviews/
http://benchmarkreviews.com/index.ph...=295&Itemid=72
http://www.extremetech.com/article2/...2346048,00.asp
Looks like a lot of site got the old cooler. Hope for us yet, or is ATI merely pulling an Nvidia with launch samples?
Radeon HD4770 Crossfire VS. HD4890
http://www.expreview.com/img/review/4770CF/score_02.png
http://www.expreview.com/review/2009...9d12291_3.html
Is it worth it selling my 4850 for one of these? Or should I just keep it?
From TPU:
Quote:
We asked AMD to clarify why there was a discrepancy between our cooler and the one we saw on the leaked images and it seems that AIBs were given a choice which cooler to use. All AIBs went for the cheaper one, so initially there will be no HD 4770s on store shelves that use the same cooler as our review sample.
damnQuote:
The HD 4770 supports CrossFire configurations with two cards. I did confirm with AMD that there is no support for CrossFire triple or quad. It wouldn't make sense on such a product anyway.
Funny how the cheaper one also seems to cool a good bit better...
10W shaved with 800 -> 450.Quote:
Originally Posted by PCGamesHardware
OMG, i think i'm going to have a baby! I'm a man btw. Those crossfire results is looking GOOD! I'm so getting two of these!
Onetreehill: Your the man! Thanks for all the info & links every 3 seconds :)
Looks like we are comparing a HD4770 to a 260-216 now :ROTF:
http://i3.techpowerup.com/reviews/AT...es/perf_oc.gifQuote:
The final overclocks of our card are 825 MHz core (10% overclock) and 1120 MHz Memory (38% overclock). ... The actual 3D performance gained is 16.8%.
on stock cooler
http://www.techpowerup.com/reviews/ATI/HD_4770/31.html
Weaker performance with the 'big' cooler probably has something to do with the smaller, slow spinning fan. At the same time the 'big' cooler also has that giant ramsink and should get better avg. memory overclocks. Maybe board partners will be able to offer the "small" cooler with a better memory cooling implementation?
pcgameshardware got 900MHz out of the "big" cooler, which kind of dents the theory that the cheaper cooler is better. Perhaps there's just a large variability in the amount of OC headroom in these chips? Would make sense seeing as these are the first chips on the new process.
Seems to be a good mid-end card.If it helps lower 4850 price even better....
4770CF has good scaling as usual, but I'd never go cf with 512mb cards
Why won't it support 3 and 4 way crossfire? It has two bridges..
Did someone try to do more than 2-way CF anyways?
How about a HD4770+HD4850 CF? Should be possible, right?
Seems a few reviews got good idle power consumption. The majority horrible. Wonder if its just a driverfix with powerplay.
But considering that review samples aint like retail is hmm hmm. Other cooler and different phase as mentioned above.
Need retail testers :p:
If it is like in the xbitlabs one. Then I'll take one.
http://anandtech.com/video/showdoc.aspx?i=3553
http://www.bit-tech.net/hardware/gra...d-4770-512mb/1
http://www.elitebastards.com/cms/ind...=711&Itemid=27
http://www.firingsquad.com/hardware/...0_performance/
http://www.legitreviews.com/article/963/1/
http://www.pcstats.com/articleview.cfm?articleID=2392
From Expreview regarding oc limit:
Quote:
If you want more, just open the Rivatuner.cfg, find “RV770 = 9440h-9443h,944Ch”, and add the ID “94B3h” after.
NCIX have better prices - http://www.ncixus.com/search/?categoryid=0&q=4770
Which reviews use retail cards?
I have never used crossfire/SLI before, will my E8500 @ 4.25GHz be fine with 4770's?
The cooler is actually the same as some 4850 have − http://www.msi.com/index.php?func=pr...7&prod_no=1688
Terrible idle consumption, GDDR5 strikes again... They should pack ATI Tray Tools with custom profiles with these cards :/
cc limit is 830/850Code:OC Limits
Site Gpu Ram Sample
Expreview 900 1100 Retail
Elitebastards 820 840 Retail
Guru3d 852 1004 Review
Xbitlabs 860 975 Review
Techpowerup 825 1105 Review
Hardware Canucks Cc Limit Retail
Pcstats Cc Limit Retail
Legit Reviews Cc Limit Review
Hardocp Cc Limit Review
Firingsquad Cc Limit Review
Toms Cc Limit Review
Neoseeker Cc Limit Review
Legion Hardware Cc Limit Review
Computerbase Cc Limit Review
I know quite a few well educated gamers that work in the "tech" field who still only buy nvidia based on misconceptions about ati drivers. These are software developers and server admins btw, and they still have no clue because they aren't what I'd consider "enthusiasts".
The points is not to say that every consumer feels one way or another but to share feedback from based on your experiences and conversations with others. The consumer bias favoring nvidia does exist whether you want to admit it or not. Have you ever heard someone who isn't an enthusiast say they only buy ati?
Can someone link me to Xfire results?
perkam
There you go :)
http://lab501.ro/placi-video/ati-radeon-hd-4770
You greedy bit-tech guys, just can't say no to Nvidia's money, can't you? :down:
GTS250 anywhere to 50%-100% faster than 9800GT according to them
http://www.bit-tech.net/hardware/gra...-4770-512mb/10