thilanliyan
Lifer
I could tell you where they pull it from butt it wouldn't be very nice. 😛
-
We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.
ATi 4870 & 4850 Price cuts
Page 4 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
I could tell you where they pull it from butt it wouldn't be very nice. 😛
Of course, you're just trying to create distinction where there is none by focusing on semantics.
Of course, but that was never in question as I was comparing identical parts with different amounts of VRAM. Obviously if you increase VRAM and its no longer the issue, other differences would show benefit.
If you were comparing a 448MB part you'd have more of a point but its obvious you want to compare apples to oranges to eggs instead. And I'm well aware of the differences that improve a GPU's performance, which is why I tend try to stick to a single variable in comparison, instead of multiple variables as you're showing.
ROFL, again, how much VRAM does the GTX 260 have?
And in no case have I said otherwise, in fact I specifically stated I was only comparing situations where performance was so limited by VRAM it suppressed all other factors.
Ya I was a step ahead and didn't explain I was comparing the 4870 and 4850 which put your bandwidth theory to rest, as the 4870 is 20% faster than the 4850 in core clock speed and the bench results were 20% faster. The 9800GTX+ comparisons also proved your assumptions wrong about bandwidth as the 1GB and 512MB parts showed a drastic difference even though bandwidth was the same.
Where did I discredit Anandtech's benches? As for being completely absurd, well I guess you haven't seen enough benches that post single digits at 2560 due to lack of VRAM. Similar is happening at lower resolutions with AA, its just not as pronounced.
Uh you clearly claimed the reason for the difference in performance was due to bandwidth on numerous occasions:
- I have? If it was "so limited by VRAM that it outweighs all other considerations" why is a 4870 512 MB still beating a 9800 GTX+ 512 MB?
Answer: Because bandwidth matters.
Where did I blame it on you? I said it was a typo, move on heh.
Again, its pretty obvious its the greatest limiting factor, ie bottleneck, as lack of VRAM clearly limits the performance of otherwise identical parts. I'm strictly comparing 9800GTX+ 512 to 1GB and 4870 512 to 1GB, that's all.
Ya what you said was much less intelligible, but you also discredited FiringSquad and a few others along with Guru3D, instead linking to Softpedia and Neoseeker as credible resources. And yes I do try and link multiple reviews when they're actually available, except for you because its a foregone conclusion you won't be able to make heads or tails of them anyways. 🙂
Where did I say the GTS 250 was faster than the 4870 512MB? I strictly said there would be instances where the additional 512MB frame buffer might allow the GTS 250 to surpass the 512MB 4870 even at playable resolutions and AA settings. The reviews today have clearly shown this to be the case although the 4870 512MB is still clearly the faster part. Now read my first reply in the thread and you'll see I state the above nearly verbatim.
AzN
Banned
Or it could be that you have an interpreting problem.
When there seems to be errors and biased in these reviews I point them out. you however use these errors and biased into your arsenal. That's just the nature with you.
Anyways all reviewers can make an error. It happens more often than you think when they rush it before dead line.
1920 and 2560 with 30+ averages are considered playable to many people.
Rofl, how would you be able to find an error when your conclusions are typically erroneous to begin with? Anyways, I got a chuckle out of that, good ole Guru3D is suddenly good enough for you now when you couldn't discredit it fast enough when linked benches from them clearly proved you wrong. Funny how that works isn't it? 🙂
AzN
Banned
And 4870 beats GTS 250 1gig @ 1920. At 2560x1600 who the hell would run GTS 250 to power their 30"? 30fps in a FPS game. Good luck.
Right and your conclusions have been retarded at best.
Let's see.
1. ROP makes the biggest difference in GPU performance
2. CPU makes huge impact in GPU lmited situations
3. GTS250 will beat 4870 because it has 1 gig of vram at resolutions and settings that's not playable
You seem to have a hard time understanding yet again. I should have expected it coming out of you. :laugh: I don't exactly remember what you are talking about with Guru3d. If I did discredit their benchmarks it was because I compared it with other benches on the web which there were errors or being biased. You however point to errors and praise for your ever failing Nvidia hack and slash marketing sham. :disgust:
Dadofamunky
Platinum Member
Nah, I totally agree with you there. Bang for the buck is the thing for me.
I just think rebranding a card is consumer abuse. It says to me, "hey, this is a failed product, let's gussy it up with some marketing and palm it off as something different to clear our inventory overload."
That is FREAKIN' insane. Even in my current employment state, I'm seriously tempted.
http://www.newegg.com/Product/...70&bop=And&Order=PRICE
a whole bunch of rebates just dropped in for 4870 cards!
a whole bunch of rebates just dropped in for 4870 cards!
MustangSVT
Lifer
wow , what an interesting thread 🙂 Ill wait until it hits below 100 to replace my 8800gs .
There's a few titles and siutations where the GTS 250 beats the 4870 512MB at 1920 with 4x and 8xAA, particularly FC2 and Fallout 3. And that's not including titles that favor Nvidia parts. People who might be interested would be those looking to SLI to cheap cards, and in that case it would clearly outperform parts that were frame buffer limited at 2560.
LMAO.
1. I've proven time and again with at least a dozen Nvidia parts isolating ROP have the greatest impact on performance since G80. You still can't explain away the differences which is why you try and discredit the results from sites like Guru3D and FiringSquad, simply because they prove you wrong or you don't understand them.
2. CPU clearly does have an impact even in historically GPU limited situations, there's plenty of reviews showing this with even single GPU. The fastest multi-GPU solutions absolutely require the fastest CPUs, a point I stressed repeatedly in the face of much skepticism 6-8 months ago. Core i7 and the proliferation of benches with highly overclocked CPUs since then have proven me undoubtedly right. 6-8 months ago, buzzwords like "CPU bottlenecking" was apparently taboo for hardware neophytes like yourself, now its once again a commonly accepted reality.
3. I never claimed the GTS 250 would be faster overall, I said it would beat the 4870 in VRAM limited situations and it clearly has.
I could list off all of your glaring errors, but what would be the point? That'd just come down to looking at your post history and randomly picking any post. :laugh:
MarcVenice
Moderator Emeritus <br>
You forgot your "/end sarcasm" on purpose, didn't you ?
I'd stay it's more common with FC2, Fallout 3 can have mixed results:
4850 512MB beating a GTS 250 2GB in Fallout 3 @ 1920x1200 8xAA
The GTS 250 2GB wins at 2560x1600, but is still not playable.
Now, that said, in FC2 more VRAM wins, but still doesn't get up to 30 frames under said settings. Even with 2GBs of VRAM available.
29 frames in FC2 @ 1920x1200 / 8xAA
This makes it more of a successful failure than an all out victory when discussing single GPU setups and extreme resolutions+AA.
I agree, once you add a little more processing power with SLI combinations, and you should see some nice scaling though.
Pneumothorax
Golden Member
I'm looking for a card in the $50-$100 range
Are the 47XX series going to be faster than the 4830? Should I just spring for the 4830 now?
The RV740 will be just a hair faster than a 4830. It pretty much falls between a 4850 and a 4830.
Here's a review of the upcoming card: RV740 preview
It looks like it'll have a very effective cooler on it if they stick with the one that we saw there. Only thing I don't like about it is you won't be able to adjust the fan (look at the connection).
But, it will be 40nm, so less power and possibly cooler.
If you choose to buy a 4830 now it won't be a huge difference. Likewise, if you wait for the RV740.
AzN
Banned
Riiiiiight. :roll:
1. You haven't proven jack. If anything you've proven ROP doesn't make the biggest impact when it comes to performance. There are so many examples of this. Here's 1. 8800gtx vs 9800gtx. 9800gtx wins with smaller ROP.
2. Right and there's like hundreds of articles on the web who is more qualified than you say other wise.
3. You claimed plenty. :laugh:
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 25K
-
-
-
