First 4870 review on the web

byusinger84 said:
I don't know where you checked, but I just checked with the local place here and they said that they can get them for me for $299 like they've been saying all along.
Click to expand...

Distributors and some selected shops are getting some Radeon HD 4870 cards but it doesn’t look that there will be too many to begin with.

Radeon HD 4870 will remain in short supply for the next few weeks mainly due to the limited availability of GDDR5 memory, but the situation is likely to get better in the weeks to come. As we've said many times before, the price wars have started and we are sure that the prices will change many times before the end of this hot and interesting summer.
Click to expand...

Any distributor that ACTUALLY has them in stock, you won't be seeing them near the MSRP for at least a couple of weeks.

If you know otherwise, let me know where they are being sold this week for MSRP of 299 :)
 
LstOfTheBrunnenG said:
LOL!

That was a fast about-face from one of the biggest nVidia proponents I've seen for this round of GPU wars!
Click to expand...

That's because in the end we are all looking for a good deal and ATI can most likely provide just that with the 4870 enthusiast/value card. We will know for sure tomorrow when the [H]ard review gets posted.
 
Glad I waited. If [H]ard's review tomorrow shows a similar trend in performance, this may be my next card. Strange, this will be my first ATI card since the x800xt I had a few years ago. Haven't seriously looked at an ATI product since then. I almost feel dirty :)
 
MrE said:
Glad I waited. If [H]ard's review tomorrow shows a similar trend in performance, this may be my next card. Strange, this will be my first ATI card since the x800xt I had a few years ago. Haven't seriously looked at an ATI product since then. I almost feel dirty :)
Click to expand...

Haha same here for me. It does kinda feel a bit dirty doesn't it?

@sekira05 - I'll let you know, but I don't expect them to run THAT high overall. Once this place has them in stock and for sale, I'll let you know how much they were.
 
sekira05 said:
Any distributor that ACTUALLY has them in stock, you won't be seeing them near the MSRP for at least a couple of weeks.

If you know otherwise, let me know where they are being sold this week for MSRP of 299 :)
Click to expand...

My educated guess as well. Retailers will be marking up the very few cards they have like no tomorrow. I wouldn't be suprised if we see prices at or around $400 or more.

If you're not a deficient and wait a couple weeks, you should be able to get them for MSRP and maybe even with rebates.
 
Wow looks like ATI finally made a comeback. Just shows that we definitely need competition, Im sure NVidia is going to adjust their ridiculous 260/280 prices now.
 
LstOfTheBrunnenG said:
LOL!

That was a fast about-face from one of the biggest nVidia proponents I've seen for this round of GPU wars!
Click to expand...

I'm not a proponent of either one, I am for whatever's fastest :D while not being grossly overpriced. At 90% of the performance for 50% of the price, I think this card fits the bill ;).
 
GoldenTiger said:
I'm not a proponent of either one, I am for whatever's fastest :D while not being grossly overpriced. At 90% of the performance for 50% of the price, I think this card fits the bill ;).
Click to expand...

It's ok, he is not an nvidia proponent he just expected big gains from the 280 and another round of mediocre performance from ati gpus.
 
With all this fanboy noise, i'm just happy to be upgrading from a mobility radeon 7500 (Its partially fossilized). Picked up a 4850 the other day and if these upcoming 4000 series cards perform as well, I might have to treat myself to another upgrade. :D
 
If this holds true then I'll be picking up a HD4870 fo-sho'. I don't trust these guy's power/temp measurements. The [H] crew are the only ones on the whole net that can correctly measure IMO. Everyone else just acts dumb and post whatever looks good. Just like how the GT200 series' assumed power requirements were completely wrong. At 55nm there's no way the card should be pulling as much as they claim. They even have the GT280/260s numbers wrong too... I'm waiting on the [H] review anyway just to make sure as far as performance is concerned. I've seen benchmarks done wrong plenty of times, but this one looks to give us an idea of the truth at the least.

If it truly can match the GTX260 then I'm all for it. Even though I won't be ready to buy for another month or two lol. You others need to give up this red vs. green stuff; straight up for 6th graders. Debating facts is one thing, but bias... isn't called for. The R600 sucked, people had a hard time with that. The 4870 (mainly) looks to be the best gpu choice this year so far, lets accept the facts (after the [H] review is in).
 
GoldenTiger said:
I tried adding to cart a minute after you posted but it just said No longer available from the manufacturer. Luckily I found one elsehwere and have it ordered, my 280's on the 'Bay now :).

EDIT: And it's sold :D $100 profit too!
Click to expand...

nice job GoldenTiger. congrats on the new card. make sure to give us your comparison benches as soon as you get your 4870. i've read your 280 benchmark thread so i'm looking forward to a similar thread for the 4870.
 
Miscommunication said:
If this holds true then I'll be picking up a HD4870 fo-sho'. I don't trust these guy's power/temp measurements. The [H] crew are the only ones on the whole net that can correctly measure IMO. Everyone else just acts dumb and post whatever looks good. Just like how the GT200 series' assumed power requirements were completely wrong. At 55nm there's no way the card should be pulling as much as they claim. They even have the GT280/260s numbers wrong too... I'm waiting on the [H] review anyway just to make sure as far as performance is concerned. I've seen benchmarks done wrong plenty of times, but this one looks to give us an idea of the truth at the least.

If it truly can match the GTX260 then I'm all for it. Even though I won't be ready to buy for another month or two lol. You others need to give up this red vs. green stuff; straight up for 6th graders. Debating facts is one thing, but bias... isn't called for. The R600 sucked, people had a hard time with that. The 4870 (mainly) looks to be the best gpu choice this year so far, lets accept the facts (after the [H] review is in).
Click to expand...

Why are there so many suck-ups here?

[H] is not the only place that can produce great reviews.

If [H] was the god of reviews, then all other reviewing businesses would go bankrupt.

We don't need to hear about the quality of [H] reviews in every single post on this forum.
 
Salavat23 said:
Why are there so many suck-ups here?

[H] is not the only place that can produce great reviews.

If [H] was the god of reviews, then all other reviewing businesses would go bankrupt.

We don't need to hear about the quality of [H] reviews in every single post on this forum.
Click to expand...

I'm not sucking up to the [H]. I genuinely think their reviews are really something else, not to mention how they go about them. This is the [H] forums, don't read the posts if they're so concerning to you. Lighten up on the emotions?
 
Salavat23 said:
Why are there so many suck-ups here?

[H] is not the only place that can produce great reviews.

If [H] was the god of reviews, then all other reviewing businesses would go bankrupt.

We don't need to hear about the quality of [H] reviews in every single post on this forum.
Click to expand...

The Problem with any other sites is the info is all static. It averaged this, it maxed this, its the greatest thing ever, [H] uses real world uses and gives the whole monitor to you and tells you how they ran it. The fact that they aren't even at that point using captured run, but active CPU processing and such just makes it even better. Its nice to see that not only did card A average 40 frames per second but that at one point it dipped to 10 FPS and then stay at 25 fps for 4 minutes. Then card B averaged 38 frames per second dipped to 9 FPS and stayed at 32 frames per second 6 minutes.

Most of the time you would see that Card A averaged 40 FPS and Minimum was 10 FPS at the same cost Card B Averaged 38 FPS and had a 9FPS minimum. With that info you would be stupid to get Card B because it was always worse. But when you see the status monitor and you see that actually even though Card A had an average frame rate higher it spent quite abit of real play time at below 30 FPS while card B while card B sits for a longer period of time at a lower FPS its consistently above that magic 30. What was at one time an easy decision just became more complicated and real judgement just had to be made by the purchaser.

Some people like it simple, here though people like it [H]ard.
 
We've just gotten the okay from our AMD rep and we should be posting some benches of the HD4870 soon. :D
 
GoldenTiger said:
I tried adding to cart a minute after you posted but it just said No longer available from the manufacturer. Luckily I found one elsehwere and have it ordered, my 280's on the 'Bay now :).

EDIT: And it's sold :D $100 profit too!
Click to expand...

Good for you! I thought about eBaying my 280's but I'll just get a pair of 4870's next month for my other rig. Can't really go wrong having the best of both!:D
 
You can also (probably?) just hold onto the GTX 280 and step up to their next revision.
 
Topweasel said:
The Problem with any other sites is the info is all static. It averaged this, it maxed this, its the greatest thing ever, [H] uses real world uses and gives the whole monitor to you and tells you how they ran it. The fact that they aren't even at that point using captured run, but active CPU processing and such just makes it even better. Its nice to see that not only did card A average 40 frames per second but that at one point it dipped to 10 FPS and then stay at 25 fps for 4 minutes. Then card B averaged 38 frames per second dipped to 9 FPS and stayed at 32 frames per second 6 minutes.

Most of the time you would see that Card A averaged 40 FPS and Minimum was 10 FPS at the same cost Card B Averaged 38 FPS and had a 9FPS minimum. With that info you would be stupid to get Card B because it was always worse. But when you see the status monitor and you see that actually even though Card A had an average frame rate higher it spent quite abit of real play time at below 30 FPS while card B while card B sits for a longer period of time at a lower FPS its consistently above that magic 30. What was at one time an easy decision just became more complicated and real judgement just had to be made by the purchaser.

Some people like it simple, here though people like it [H]ard.
Click to expand...


I like [H] reviews but what happens when you don't have a 30" monitor with 2560x1600 resolution, then how will the [H] reviews help some one decide on these new GPUs?
 
My guess, if it has playable fps @ 2560x1600 they will increase as the res decreases. ;)
I get 2K fps on my Dick Tracy wrist watch TV. :p
 
Kowan said:
My guess, if it has playable fps @ 2560x1600 they will increase as the res decreases. ;)
I get 2K fps on my Dick Tracy wrist watch TV. :p
Click to expand...

Yea but some cards scale differently than others, some hit a wall at higher resolutions and AA/AF settings, but as soon as you lower them they can become competitive.

Example:
etqw.jpg
:source anandtech.com

At 2560x1600 the GTX280 is a beast and beats all others, but as the res goes down to 1920x1080 the 8800gt sli and 980gx2 start to catch up. Once you hit 1680x1050 they surpass the gtx280, being more cost effective too.
 
Doodies said:
I like [H] reviews but what happens when you don't have a 30" monitor with 2560x1600 resolution, then how will the [H] reviews help some one decide on these new GPUs?
Click to expand...

The point of the [H] review isn't to tell you what the framerate is at 1280x1024, or 1680x1050, even though that's what a lot of people are using for resolution. The [H] review tries to tell you what the highest playable settings are. And those highest playable settings are what looks best (subjectively) to the reviewer. So if 1920x1200 with low AA looks better to [H] than 1680x1050 with high AA then they will post the 1920x1200 numbers.

I personally wouldn't mind at all if they decided to include some 1680x1050 results anyway, but I think I'm smart enough to get a very good impression of performance even if the resolution is higher than what I have. edit: just read your newer post. Good point.

Honestly, if Highest Playable Settings are up in the insane-resolutions-range then it stands to reason the video card in question is more than you likely need for your 17" LCD.
 
In a world of fugly video card coolers, that carbon fiber look = massive win.
 
them bioshock numbers are tastey.

I think I can say with certainty the HD4870 X2 will be the performance king until Nvidia can shrink they're GT200 and get two on in a single package.

Good stuff.
 
rls669 said:
In a world of fugly video card coolers, that carbon fiber look = massive win.
Click to expand...

These are engineering samples with beta drivers and generic ATI stickers, but the cooler sticker design will be totally up to the OEM resellers. Take that as you will, their designs may appeal or appall you. :eek:

But yes, these cards seem very promising. ;)
 
I will get the 4870X2, it will beat the crap out of g200 and g200b for sure with a good margin of about 50%, perhaps even more if cf works in many games or they have some magic at work. I wonder if the engineers at nvidia are on steroids now to retake the performance crown ASAP, becuase of the partial failure GTX280.
 
I have to say, having moved to Nvidia after many years of ATi, if any of preliminary numbers are even remotely accurate, as long as the visual/performance combination is there on a single 4870, especially for the price, I might be moving back to ATi.

So far, though they're just "generic benches", the 4870 is falling right between the GTX 260 and 280, sometimes matching the 280, and for half the cost.

Considering so many gamers run at 22" (1680 res), and ATi's history of much more reliable drivers and slightly sharper visuals, the 4870 looks like the way to go for the new gen of GPUs.

Anticipating greatly [H]'s review, and I'm definitely leaning towards snagging this beast, should it turn out as well as it's looking. If you could hit the performance even between the GTX 260 and 280, for half the cost, it would almost seem like lunacy to me to pay $650 for the 280, no matter how you look at it, as it will not offer that much more performance (projection).

Of course, that's just a projection based on the generic numbers we're seeing so far, but it's looking like (as with many others) I might just be moving back to ATi, depending on how well it does being put through [H]'s testing.

Either way, it should pound my current stock 8800GTX into dust, and allow for max settings in all games without any lag at all, increased performance and perhaps even slightly better visuals.

I'm not "rooting" for anyone, though it would be great to see ATi make this come-back, and offer the consumer more choices, most certainly. I'm just dying to upgrade from my 8800GTX without having to sell a kidney on the black market.

I'm very excited for the review tomorrow!
 
iBP_James said:
Our AMD rep gave us the green light to post some preliminary benchmarks of the HD4870s in single and CrossFireX setups. We have the stats on our forums, check it out!

It certainly looks promising for the Reds considering the pricepoint they are offering their HD4800 series cards. :D
Click to expand...

Something isn't right with the CF WIC results....

Single:
1600 x 1200 (4x AA, 4x AF) @ Very High: 40.6 avg, 23.6 min, 69.3 max
Click to expand...

CF:
[1600 x 1200 (4x AA, 4x AF) @ Very High: 28 avg, 26 min, 112 max
Click to expand...

How can both the min and max be higher in CF but it's average is lower then a single? :confused:
 
You must log in or register to reply here.
Back
Top