Whats very strange here is that looking around on another ANAND article, for FEAR 16x12 4xAA (this vs. newer article):
the X1900XTX got 51 fps vs. 43 (wow, ATi got downgraded in the future? hmm)
the 7800GTX512 got 31 vs. 40 (hmm it later got a boost? odd)
the 7800GT got 22 vs. 28 (wow, nVidia keeps getting much better all of a sudden)
the 7800GS got 13 vs. 27 (NOW THATS IMPRESSIVE!)
Seems this article was garbage as it really downplays nvidia h/w and is on the verge of craptastic fantasy for ATi.
comparing a middle of the line card against high end cards is pointless and quite frankly a waste of time.
this card is meant for those of us that are running middle of the line systems and aren't quite ready to put down the change to scrap the whole system.
i.e. compare this to the 6800 agp line (and ati's equivalent if you want).
AND, use a realistic processor for this graphics card range, i.e. intel 2.4's to 3.2's (and AMD's equivalents).
haven't been visiting as much lately and i'm seeing the quality going down...give me a reason to come back guys.
I can not stand Best Buy, and the fact that Nvidia is getting into 'bed' with them brings them down a notch in my book. As for the card, does not look like a reason to upgrade from my 6800 GT AGP, card.
It's only been a day...if they do update it, it may take a few days
I think the general concensus is
1) Add some benches of AGP cards to show it's true competitors as well as give an indication of what end AGP users currently have
2) Overclock the 7800GS and show that benchmark
3) Discuss the potential of softmodding the card, unlocking pixel pipelines, shader units, ROPs
How about it Derek, any chance that this review can be updated or are we going to have to wait a while until your next AGP face-off that you alluded to in the review ?
I too am looking forward to an update to this review. I do have one question though. Isn't the X850XT PE just a factory oc'd X850XT? I mean besides the clock speeds there really isn't any difference between the two. So that being the case didn't he in fact underclock those two 7800GS's and bench them against a oc'd X850XT??
There is a difference, the X850 XT PE is an official SKU from ATI, even though it's only a bit faster then the X850 XT.
This is the same situation that is occuring with the X1900 XT and X1900 XTX now.
Derek underclocked the 7800 GS to the point of Nvidia's minimum specs for this SKU 375/1200, but this doesn't represent the performance of the product we will be getting from certain manufacturers, such as the eVGA Superclock edition.
That's a pretty weak argument. The fact is he under clocked those cards below there off the shelf speeds and tested them against what is essentially an overclocked X850XT. He could have redeemed himself if he had just included the benches for the 7800GS's at their factory oc'd stock speeds. As it is, he misrepresented the performance of those cards.
You misunderstand me.
I am not arguing that this review is good, quite the contrary is what I believe as it stands now.
The 7800 GS should have been compared at Nvidia's reference design speeds and using eVGA's 7800 GS SuperClock Edition speeds as well as that represents the product the consumer will get.
Using Nvidia's reference designs speeds only represents the worse case scenario which isn't good enough.
This card will spark alot of interest for people like myself who want to get as much life out of their AGP system as is practical before deciding that it's time to do a more complete system upgrade.
In my case, I'm still running an Athlon XP at 2.3GHz, and since upgrading from a Ti4600 to a 6800GT, when Battlefield 2 came out, I have a viable, mid-range game system (it plays Battlefield 2 at 1024x768 with mostly high settings). This has allowed me to postpone that upgrade to a new PCI-E based system, for now.
So, the question is for me whether any new AGP card will provide yet another intermediate upgrade step before the inevitable upgrade to a PCI-E system. This means that I do want to know how the peformance from this new AGP card stacks up against the kind of PCI-E cards that I would likely to looking at (such as the 7800GT, probably the best bang for your buck PCI-E card, and the X1900XT, an example of what the top end cards have to offer). I also want to know how much faster this new AGP card is compared to the other top end AGP cards, such as the X850XT PE that you tested, as well as other cards that people are likely to be using now, such as the 6800GT/GS, X800XL, 6800 Ultra, etc. Not all of these cards need to be included, of course, but enough to give a feel for how the new card stacks up against the old guard.
Finally, but not least, I want to know about what sort of value this new card has to offer. How much is it selling for? Does this card seem to have alot of overclocking headroom, and if so, when running it at an overclocked speed, what cards is it comparable to?
Assuming that this new AGP card has enough horsepower to make it a significant upgrade from what someone is already running in their legacy system, how does the cost of this single component upgrade compare to the cost of a more complete overhaul to a PCI-E system of at least comparable speed (i.e., PCI-E motherboard, processor, and video card)?
Those are the sort of questions that I will be looking for answers to when I investigate what a card like the 7800GS AGP has to offer.
Nice review...
It's not complete because we don't have comparisons with other AGP cards (6800/6800GT), but we can deduce it ourselves :) Maybe later AT will add these tests !
I want to keep my A64 3400+ but change my 6800GT to a more powerfull card without eating bread for 3 months... I was considering buying a 7800GT + nforce 4 socket 754 board or a 7800GS AGP.
I think I'll wait a few months to see what happens : the 7800GT + mobo costs nearly the same price than a 7800GS here in France, for more performance... But the 7800GT is already struggling in 1280 with recent games (ie Fear).
Wait and see...
You underclock the boards from what is actually available....
You leave out MOST of its direct competitors.....
You rig 20% of your tests to make the board lose to the only direct competitor you bothered to test.....
What is happening? If a part has issues then let them come out on their own- too high a price and not enough performance- OK, but let it come through on an honest basis. This review is something considerably below you ignoring it all together.
The 6800GT was for a long time the AGP card of choice, and the 6800GS has replaced it recently as it is cheaper and only slightly slower. Given how either of those two cards would have been what AGP users might have considered, or still be considering, the lack of a 6800GS in your benchmarks was unforgivable. For what it's worth I'll say a stock 7800GS is slightly faster than a stock 6800Ultra, and therefore has a decent margin over a 6800GT, and especially a 6800GS. The real strength of the 7800GS is in its overclocking headroom though, as you pointed out. That doesn't necessarily make it a good purchase though.
Anyone who wants to hang on to an AGP based system could use a 7800GS to upgrade from the likes of a 6600GT or lower, but I really think their money would be better spent on moving to a PCIe platform and graphics-card.
The interesting thing is that AGP 6800 GS cards actually use NV40 chips, so they can unlock to 16 pipes and 6 vertex units (plus 16 ROPs) if you're somewhat lucky. They're basically like the X800 GTO in OC/unlock potential. Problem is getting software that will allow that. They *don't* OC hardly at all, in comparison to the NV42 6800 GS.
If you want my opinion, this 7800 GS AGP stuff is too little too late. NVIDIA and ATI are both trying to get users to upgrade to PCIe for several reasons. First, it potentially makes them more money, since they also sell a lot of chipsets. Second, and probably more importantly, supporting older technology just makes more work for them. They would rather focus on one platform and forget about AGP. The conspiracy theorist in me says that 7800 GS was intentionally crippled just to help AGP die out.
FWIW, 6800 Ultra is overall slower, though in a few games it can tie/slightly lead the 7800 GS. With overclocking, 6800 GT/GS/Ultra AGP cards can't touch the 7800 GS - 450+ MHz on the core seems readily achievable. Overclocked, the 7800 GS easily comes out as the fastest AGP card, which makes it more palatable. Losing half the potential fillrate still makes the card less than perfect.
In the end though what matters is performance, and the G7x Shader enahncements plus 1.2GHZ memory speed, help it beat out the 6800 Ultra, not to mention be cooler running. Even with the disadvanatge of 8 ROP at 375MHZ. It just beats out the 6800 Ultra, and brings the 7 Series feature set to AGP, which is perfect as it doesn't hurt the 7800 GT. So far though I haven't seen any situations where 6800 Ultra wins...
Well given that the 7800 GT PCI-E has a MSRP of 449US and have been on the market for nearly 6 months and the 6800 GS PCI-E has a MSRP of 249US and been on the market for nearly 3 months, we have to give a few months for this part to settle in pricing.
2-3 months should be ample time for this part to fall to correct pricing for the AGP, and remember it's competitors are 6800 GS AGP, as the 6800 GT and Ultra line for AGP supplies should be dwindling to basically nothing. Same thing with the X850 Series, only a competitor assuming you can find it at a good street price.
There are a few instances where 6800U wins out, but not by a huge margin. They're not shown in our benches, but http://www.firingsquad.com/hardware/bfg_geforce_78...">FiringSquad has them. The thing is, if you have a 7800GT in the charts, you get a much larger difference between 6800GS and 7800GT, likely due to the higher clock speed, additional pipelines, and twice as many ROPs. The 6800U never even gets close to 7800GT performance, I don't think.
I wouldn't call the instances where the 6800 Ultra wins out as significant, as that is the worse case scenario, most games do fine with a 2:1 ratio on Pipes/TMU vs ROPs as the 6600 GT has shown.
Of course not for the 7800 GT, that is very obvious, G7x Shader technology coupled with a 25% higher bilinear texel fillrate and 16% higher vertex power, equal ROP power, gurantees a victory for the 7800 GT, typically I would say Shader Power > Memory Bandwidth in determining performance, unless you cripple memory bandwidth to crazy low levels.
Considering how crippled the 7800 GS is I am very impressed that it typically defeats the 6800 Ultra. G7x Shader technology helps out quitea bit as it has a core clock disadvanatge too.
Why only show underclocked results when the cards are retail parts that run faster out of the box?
Why not compare to other AGP cards?
I'm very confused as to what AT was thinking when they decided to do the review this way. Unless the intent was to force your readers to look at other websites as part of some AT master plan.
Although "hard launch" is important, does every review have to dance around with definitions of it? The 7300GS was available to buy when it was announced (I saw one in a store the same week the announcement hit the web) and it looks like the 7800GS is available in retail stores in the USA right when the card was announced as well. Seems to me that both cards were properly "launched" and that both of the main card companies are putting "paper" launches behind them. How about making a longer comment if there is a real "paper" launch by either one and giving it a rest otherwise. In the case of something like the 7300GS, saying that the card will be available in different regions as per X schedule should be enough. Even with a full on "hard launch" there are always places in the world where cards don't make it day one. The real question is always the real availability of the cards - do they really exist to be bought in some sort of base line quantity? If yes, then leave it at that. Releasing a review on Thursday and then complaining that a card will only be at a major retail store 3 days later seems like a very faint complaint to me.
To close off that point - cell phones often are launched with an exclusive for one region or one carrier when they first come out. That is both marketing and matching initial production with demand. It also doesn't result in ranting and raving all over the place.
quote: Someone from AT should probably say this, but perhaps they think it's obvious:
This article is about the launch of the NVidia GeForce 7800GS GPU. It is NOT about a particular vendor's board. Vendor boards are tested in vendor roundups or occaisionally individual board reviews.
While I agree it would have been *interesting* to see the vendor cards tested at their shipping clock speeds, that is really for another article. A GPU launch article should test either reference hardware or vendor hardware set to reference clocks, and this article does exactly that. Period. Quit complaining.
People should also keep in mind that neither AT nor most other sites re-run every benchmark on every card for every test. They test the new card on their standard hardware platform using current benchmarks, and then use benchmark numbers for other cards from recent tests. Older cards were last tested on an older platform using older benchmarks, so no comparable benchmark numbers are available.
AT isn't going to re-benchmark a pile of older cards unless doing so is the POINT of the article. This article, again, is about the launch of a new GPU, not a 'Mid-range to High-end AGP Comparo' or '7800GS Vendor Roundup'.
I only have a couple of thoughts on why this was done:
1) For the non-conspiracy theorists, these are just G70 cores that had some defective ROPs and pixel pipelines. Alternately, maybe the HSI chip somehow has issues with the heavier load?
2) Option two is that this card is a bone thrown to the AGP crowd, and unfortunately the bone is already dry and picked clean. 8 ROPs basically relegates this card to equaling what was already available - at best. I would have much rather seen a 7800 GT or GTX 256MB (or even a 512MB - though those seen to be in tight supply due to chip yields).
3) Realistically, it may be a combination of both factors. They don't want *too* much power available for an "old" platform, but they also had a lot of 16 pixel pipeline chips available. I really doubt that all of the chips had 8 defective ROPs, but maybe some where 8, some were 12, and some were 16 and it made some sense to just release them all as lowest common denominator parts? Be interesting to see if anyone can figure out a way to attempt to reenable the extra pipelines and ROPs.... (Go RivaTuner!)
Considering that the 6800 GS is a 12 PP/8 ROPs chip and it only costs $200, I really wish this had been at least a 16 PP/12 ROPs release. That would have maybe made it worth the 50% price increase. Overclocking and possibly unlocking the 6800 GS AGP cards will get you really close to 7800 GS performance. Hmmm... Derek probably should have included some 6800 GT/GS/Ultra numbers as well. Oh well - maybe in a follow-up article.
I hate to repeat what the others have said, but this review was borderline useless to the very people who would come here to find out how well the card performs...
My other half has a 6600GT, and I have a 9800 Pro. How does the 7800 GS compare as an upgrade. NO IDEA is the conclusion I reached after reading your review, which has forced me to look elsewhere.
I did however find the article interesting to see how it compared with PCI-E cards, but thats about it.
I think you need to add in some more cards to THIS review, rather than letting everyone wait for another AGP comparison in a few weeks time.
Is everyone overlooking the fact that this is going to be a part with a $399 MSRP?, and why no 6800GT AGP, or 6800GS AGP(same thing) an ommision like that is preaty bad.
the MSRP of the 7800 GS AGP is indeed 349US, the OC parts are about MSRP 379US, however we will have to wait to see how this translate to street pricing.
The X850 XT PCI-E/AGP is MSRP 499US, Street 200-250
The 7800 GT PCI-E is MSRP 449US Street 275-300
The 6800 GS PCI-E/AGP is MSRP 249US Street ~200
Someone from AT should probably say this, but perhaps they think it's obvious:
This article is about the launch of the NVidia GeForce 7800GS GPU. It is NOT about a particular vendor's board. Vendor boards are tested in vendor roundups or occaisionally individual board reviews.
While I agree it would have been *interesting* to see the vendor cards tested at their shipping clock speeds, that is really for another article. A GPU launch article should test either reference hardware or vendor hardware set to reference clocks, and this article does exactly that. Period. Quit complaining.
People should also keep in mind that neither AT nor most other sites re-run every benchmark on every card for every test. They test the new card on their standard hardware platform using current benchmarks, and then use benchmark numbers for other cards from recent tests. Older cards were last tested on an older platform using older benchmarks, so no comparable benchmark numbers are available.
AT isn't going to re-benchmark a pile of older cards unless doing so is the POINT of the article. This article, again, is about the launch of a new GPU, not a 'Mid-range to High-end AGP Comparo' or '7800GS Vendor Roundup'.
Then why doesn't AT just have a Video Chart that can pull similiar Min/Max/Avg FPS for each type of test at each resolution / AA / AF setting at-a-glance? Instead we have to muck around dozens of articles and have no generic yardstick. Why include the 7800GTX then if not to compare? Stop being an apologist.
Your kidding right? Ok 1) You do realize that more than one person does reviews for Anandtech. 2) Due to 1 (see above), it would be a weebit difficult to ship the standardized review rig to each persons house. 3) The tests done on games change over time. When 6800 Ultra's were being reviewed, there was a totally different set of games being played. 4) As with 3 (see above), hardware changes over time. If they wanted to test on the same system for every graphics card YEARS apart, they would still be using a 486.
This isnt Toms. If you want the VGA Charts, go get them.
Im sure this being a *technical* site, one can factor in plain vanilla CPU/mobo/chipset and formulize those effects vs. the video card. I guess Futuremark's database is closer to where I'm going, where its a composite scoring system, and then an overall, but not divergant and possibly too specialized and even misses the market's techniques. Tom's chart is a bit fuzzy to read it makes my eyes blur at times ;) I mean I could try to scour thru AT and do my own chart in excel (and factor in platform %), but who has time
First, I enjoyed this review as it has relevance to me (NF2 AGP user). However, as already mentioned, it would help to compare this card to other AGP cards (6600GT, 6800GT/Ultra, X800XT/PE, X850XT/PE) because AGP systems use these cards. That way, it'll help to access any inherent value (if any) to purchasing. Also, while most reviewers have chosen A64 AGP systems, benchmarking on other platforms (like Socket A) would be extremely appreciated as many (including me) still use these systems.
As I said in my earlier post. Head over to Tom's HW for comparison with a 6800 Ultra, the 7800 GS beats the 6800 Ultra AGP in almost all the latest games. The X850XT is still tops, but can't do the eye candy.
quote: We've got full retail boxes from both BFG and EVGA.
...
For the purposes of these tests, we underclocked a card to the minimum speed NVIDIA is endorsing.
I'm a little shocked at the decision to ignore the actual speed these cards run at out of the box, and instead show an artificially limited card -- exclusively.
I wouldn't have minded seeing the "published spec" score, but not at the expense of knowing what you get if you walk into a store and walk out with one of these cards. Without that, this article is glaringly incomplete. I'm left little more informed about how an actual retail product performs than I was before I read it. That doesn't sound like AnandTech.
I agree that the reviews are subpar. I think anand should not be delegating as many reviews as he has been. You can ALWAYS tell which reviews he does from the moment you start reading it (a good thing). I guess he is just to busy these days to do more of them.
Where is the 6800 Ulra or 6800GT in that review? How about a 6600GT? Come on guys... these are the cards that people who would consider buying a 7800GS own right now. Not including a 6800 or 6600 series in that review makes no sense at all.
On a side note... AT has changed since I started visiting... I used to be able to get all the info I could ask for right here... now I have to visit 5 other forums to get the info I want, and that's not a good thing.
here here.. this article is weak; run the cards at their OC speed, and then LABEL the REFERENCE Speed 7800 GS on the charts, since this is MISLEADING and *NOT* intuitive in comparisons when putting money down on the table.
From a competition performance-cost standpoint, let's talk about bang for the buck here. This isn't a PREVIEW, this is about a card coming out in days.
How does the X1600XT 256MB, 6660GT 128MB, 6800 128MB rate in the $130 range.
Then its the sub 200$ range with the 6600GT or 6800 256MB versions.
We can all go thank Tom's HW site for the 2005 VGA charts to get a rough ball estimate due to the X850XT PE (which was not 210$ but rather in the $400+ range (from using pricegrabber). Not only that but they do compare the 7800GS against the 6800U and the BFG card is modestly beating the old tech. For the 300$ target price of the 7800GS AGP, there is a vast wasteland and no competition pricewise, but I feel its not justified to say it will drop 6800GT/U prices since it is no longer being manufactured (still going to be rare, and same performance, so why even bother getting an old large hot stinker of a card on E-bay-like services when you can get brand-spanking new tech?)
I'm still leering of why nVidia cannot see to get HDR and AA/AF to play nice, so I personally don't see much advantage in upgrading from a 6600GT / 6800 GT / U part just for SM 3.0. If the 7800GS is ~30% faster than a 6600GT, ~11% faster than a 6800 GT, and barely ekes out victory over a 6800 U, it doesn't do justice to upgrade unless you need to make the leap from TurboCache or integrated. I do hardware (& sys building) recommendations for many people and want to steer them in the right direction and make them feel they got the best value. Will this make people put off buying an M2 & PCI-e mobo for another year and stick with AGP? I'm not in the business of making predictions, but for me personally, I would rather get an X2 chip, mobo, and X1900 or 7800GTX (or maybe even the refresh to choose from), possibly an Creative X-Fi entry (but highly doubtful, shrug to the 5.1 audiophiles because there is a lack of a stand out digital spdif 5.1 headset - this boggles my mind and is another side topic) and 2GB of DDR2-667 RAM right about xmas 2006. I never find a truly good integrated PC builder that satisfied my DIY urges.
The 6800 Ultra looses the lead in comparison to 7800 GS. Head over to toms hardware to see benchmark scores. I have an Opteron 146 w/ socket 940 and 2GB ECC-DDR400 that I received from work, but it has an ATI 9600SE. The 7800 GS is exactly what I have been waiting for since upgrading to a 939 with PCI-E is useless and will cost me more than a 7800 GS. The 7800 GS should give my system usefull life for a year at least, maybe more and it will be half the price of a 6800 Ultra.
I'd be interested in a comparison between the XT1600 and the 7800 GS AGP cards. They are my 2 candidates for a low power, high performance AGP card for my HTPC. I think many other HTPC owners would be interested in this as well, especially how the HDTV performance and quality compares on older CPUs.
Yeah i cannot see why Anandtech decided not to show default factory o/c scores from evga. Sure if AT wanted benchmark with 7800gs underclocked but ALSO include 7800gs standard retail overclock benchmark.
I mean the consumer who buys the evga 7800gs is not going to underclock the card by 20% from 460/1.35 to 375/1.2 and lose 20% performance.
I think you missed the point of this article. It was not meant as a review of the particular brands, but of the card nvidia just released. Those are the reference clocks and that's what nvidia is putting out right now, it is the baseline you can expect all the 7800gs cards to do, not what some brands may do. That they just happened to use an evga card to show it doesn't really have anything to do with the matter.
I'm pretty sure there will be a followup article discussing the differences between the brands and their clocks, with some lower end comparison cards everyone is craving for thrown in for good measure. I hope you can leave your pants on till then.
If the context is seen as those cards, for that price, we can reasonably assume only someone ignorant of clock speeds would pay same price for a lower clocked stock card, else they'd planned to really o'c it themselves.
So, we do care what some brands do because it's not that we're buying a "7800GS" if we do, rather we're buying a 7800GS for 7800GS performance levels which are not a conceptual thing but an actual product.
No kidding, I was actually VERY upset and confused as to why you downclocked the retail parts, or at least didn't run tests both ways. VERY poor reviewing on this one IMO, not typical of Anands norm. IF I go out and purchase a eVGA 7800GS for my FX-51 AGP system, I"m sure as hell not going to underclock it the second I get it. I would most likely just leave it at stock and call it a day.
PLZ for the love of God, show us the true out of the box performance numbers, and compare it to other cards like my AGP x800XL, or the 6800GT/Ultra AGP parts.
This is my idea of a hard launch --> A buyer has to be able to purchase the part on the day of the launch, or it is not hard. Not ALL buyers, but enough stock so that parts are available to purchase nationally for at least 24 hours. Nationally meaning the part would have to be at a national retail chain (Best Buy, CompUSA, Fry's, etc.) or online where a price search engine can find it.
So my definition of hard launch is, parts nationally available for 24 hours, or it is soft.
The AGP version of the X850xt is way higher than $210, more like around $400. IMO, its pointless to get it. You can just get a new mobo for the price difference, especially if you already have a good processor.
When I went to PCIE, I bought a new mobo, Sempron 2800+, and X800GTO2, all for ~$350 $350. Then then sold my old setup for $150, so my net cost was $200, and I have a much faster cpu as well as a faster video card. Way better than paying $300 or so dollars just for a new video card.
Well to be more accurate, the X850 XT AGP can be had for <250. Heck I bought an X800 XT PE last summer for $249 shipped, and some people are now getting the 850's for less than that. You don't see these deals widely advertised, as it would compete with the 800GTO's.
Now to be fair, dgniel compared street prices of the X850 to MSRP of the 7800GS which is apples-to-oranges. But I don't think too many people pay $400 for X850 XT's anymore, seeing as a 7800GT will match its performance for 100 less and better feature set.
If the numbers are correct it shows an overclocked 7800GS with stock cooler getting a higher score than the 7800GT.
I have read in several places that people have been able to unlock the unused pipelines and shaders using RivaTuner.
If all this is true then Its a must buy for all of us still using AGP
I agree with all above. I would also like to see the card compared to the AGP 6800GT and Ultra with OC scores.
This article needs to be taken with a grain of salt since the 7800GS in the article is a PCI-E engineering sample. In this article the 7800GS scales very well when overclocked, hopefully the AGP version will show similar gains when overclocked.
I have x800 Pro o/c. I really want to upgrade but buying a new mobo + video card is a bit too expansive. If only I could get a better video on the level with 7800GT and x1800XL, I would be happy. Unfortunatelly, 7800GS just doesn't cut it (you've seen the tests). I wish they come out with AGP versions of 1800XL and 7800GT, I would have gotten one. But then again, ATi and nVidia both make chipset with PIC-E, thus it is profitable for them to have people buy new mobos with PCI-E. :(
it seems rediculous to me that with an agp part, you compared...not other AGP parts, but high end pci-express cards. i have an agp system, i want to know how much better it is than other AGP cards out there, not a top of the line card, that would require me to upgrade my processor and mainboard.
again, you have done so in the past, so i request it once again. many people buying this product run either 10x7 or 12x10 with and without 4x 8x. these were left out in the line graphs, i request them once again. otherwise without both these things, i think it leaves us people who would considering to buy this product, needing to go elsewhere to find out how it compares, and what is a good buy.
Agreed. Not everyone is as interested in how a card does at all kinds of maxxed out settings and features as we are at what measures would be needed to use the card for modern games, playable rates. If a feature entails a significant performance penalty, but one has an AGP board, of course they'd disable the feature when necessary.
I also fail to see the logic in lowering card clock speeds when (all) 2 samples were shipped at a higher rate. While one may not want to manually raise the speeds for the tests, the out-of-box speed a card uses should be considered stock no matter what nVidia spec'd. IOW, I don't see a lot of people underclocking their cards, it would be a scenario never seen in gaming use, IMO.
I agree with others that you should have picked other cards for a comparison. Specialy considering your introduction words in the article.
The point is that it's about users who still use up-to-date AGP systems (like me) that would like to know if this card is worth the money and wait a little longer to upgrade to PCIe.
Therefore the comparison should have been between this card and other AGP cards at the same price point, or more expensive ones. Now you only tested the X850. A card that most people reading the test will probaly not even consider as an upgrade. And comparing it to the X1900XTX and 7800 GTX 512 Mb ? Come on ! Like these cards would be viable upgrades considered by the people who are intrested in this card..
Besides, if I want PCIe numbers I can read any of your other tests.
There are also no noise, power consumption, and OC results (I bet the eVGA would beat the X850 hands down). And the test was done on a medicore ULi based board instead of a high-end S939 board.
Sorry for the harsch comment, but I simply had to say this ;-)
No. It could be considered a contradictory statement. An oxymoron is a phrase, not a sentence. ;)
But I still disagree with you. Since AGP is still a viable platform for playing current games. In fact, the 7800 GS keeps AGP systems relevant, because it puts an AGP card on the market that utilizes the most advanced kinds of mainstream video technology.
I would be interested in seeing the scores from the eVGA version of this card. It has significantly higher clock speeds. I would expect that it gives quite a bit better scores. As it stands now though, if I were looking for an AGP card, I would look toward the X850 XT. The non-PE version can be found for $250 or less. The stock-clocked 7800 GS is a nice card, but not a very good deal at $300 or more.
So, to conclude, looking forward to some numbers from the over-clocked eVGA 7800 GS. :-)
Unfortunately, comparing this card only to cards that are PCI Express only or you won't be upgrading from (like the X850XT PE) is kind of pointless...
I say this because I think the question worth asking is "Is this card a viable upgrade for AGP users", so you would want to compare it to cards like the 6800GT, 6800, 6600GT, and perhaps those Radeon counterparts. Then, having the X850 line in there for comparison would give you a good idea as to which you'd want to upgrade to.
I second this, why not put in a 6800GT or Ultra? if you have a 6800 card it would be nice to know if this card is faster. I stopped reading after I found out there were no cards I found usefull to compare it against..
The PCIe numbers are useful to AGP users in that it shows where the cutoff between the highest possible AGP performance is and the next step up in performance possible on a PCIe platform.
That being said, we are planning on looking at the AGP version of the X1600 as compared to the AGP 6800 GS, so we will certainly revisit this issue.
Huh? If you're using AGP and looking to upgrade the video card, what does it matter what the PCIe cards are scoring? If you'd decided to go PCIe, pretty much any of the cards out would be better in one way or another. Maybe just throw in one 7800 GT or such for that comparison. What's "useful" to someone looking to uprade an AGP card is how it will perform compared to what he has, a 6600GT, 6800GT, 6800GS (AGP version), a 97-9800 and a few of the XXX800XLXXLX X's. That's what they need to be able to see if the card's worth it, or maybe it is time for them to switch to PCIe. Speaking of which:
quote: The cost of keeping up with the pace of technology is certainly steep, so upgrade cycles for some people stretch quite a bit longer than others.
AT's been on this jag for a while now. It was just 18 months ago you were recommending an AGP board as your "Editors Choice" to readers. While some upgrade more frequently, I don't see the guy who buys a top-of-the-line mobo and expects to get two years out of it as in the same category as the guy trying to nurse his Sempron system along. Oh well, at least it's better than when PCIe was available for less than a year and offering zero performance benefits compared to AGP, yet AT was insisting that avid gamers shouldn't expect to see any new AGP products.
/rant
Anyone know what the numbers are for the same tests for a 6800 Ultra?
I think I am happy to see that there is a 7800 part for the AGP users out there like me, but I am not sure if I should get it because I can never find comparisons to the 6800 Ultra.
I know I would get the better shader model and more pixel pipelines, but what does that really mean in a head to head comparison?
did you compare SC with SM1.1 path or did you run it on 3.0 for the Geforce?
Didn't see it written somewhere so I thought I ask, because it would make a big difference....
"Our 7800 GS was tested with SM3.0 enabled and all the options while the X850 XT PE does not support these features.
The price the 7800 GS pays for enabling HDR and other SM3.0 eye candy is that it is limited in playability to lower resolutions. Disabling the SM3.0 features would give performance quite a boost without AA (SM3.0 features are not available with AA enabled). "
Im a little bit dissapointed, overall. While I guess the 850XTPE is a pretty lofty card to try to beat, i was hoping for a little more power from this piece, to keep my AGP rig alive a little longer. :D
Might have been a good idea to compare AGP and PCI-Ex versions of this card on the same board with a ASRock 939Dual-SATA2. I know weve gone over that terriroty before, but it would have still been interesting to me to see it again using this higher end part that likely would take a little bigger hit going through the HSI bridge. Just an idea.
To be very clear -- in case the title didn't sink in :-)
The 7800 GS will not have a PCI Express counterpart. This is an AGP only card.
The engineering sample we tested last year was simply that and will never be anything more. There are a few differences as well, so a direct comparison isn't really possible.
Doh, didnt get that the PCI-Ex version was NEVER coming out. Just thought it was coming out later. I read the title more as "NVIDIA's AGP Only VERSION of the 6800 GS" Im a jerk-face. My bad. :D
I think its great that nvidia and ati are still releasing AGP parts. I've beeen planning on replacing my 9700pro so I like having so choices even if its very limited.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
93 Comments
Back to Article
spinportal - Tuesday, April 18, 2006 - link
Whats very strange here is that looking around on another ANAND article, for FEAR 16x12 4xAA (this vs. newer article):the X1900XTX got 51 fps vs. 43 (wow, ATi got downgraded in the future? hmm)
the 7800GTX512 got 31 vs. 40 (hmm it later got a boost? odd)
the 7800GT got 22 vs. 28 (wow, nVidia keeps getting much better all of a sudden)
the 7800GS got 13 vs. 27 (NOW THATS IMPRESSIVE!)
Seems this article was garbage as it really downplays nvidia h/w and is on the verge of craptastic fantasy for ATi.
boa49 - Friday, February 10, 2006 - link
There is definitely interest!!!!!!!
yacoub - Tuesday, February 7, 2006 - link
For those who wanted a review that shows how the stock-overclocked 7800GS compares both to a stock-clocked 7800GS and all the other AGP cards out there, check out this review:http://www.firingsquad.com/hardware/evga_e-geforce...">http://www.firingsquad.com/hardware/evg...00_gs_co...
Very comprehensive and what I think a lot of us were looking for.
araczynski - Tuesday, February 7, 2006 - link
thanks, that's EXACTLY what i was after.i haven't been to firingsquad in ages, i guess i better start going there more often now.
Scarceas - Saturday, February 4, 2006 - link
I disagree with those who said it was pointless to compare to the x850xtpe. This let's me know which card I can top out the AGP slot with.I would have liked this a bit better if the benchmark results included something using the Source engine.
Why no overclocking results? Heck, we didn't really even get to see mfgr defaults... I'd like to see how 460 MHz scales.
araczynski - Friday, February 3, 2006 - link
comparing a middle of the line card against high end cards is pointless and quite frankly a waste of time.this card is meant for those of us that are running middle of the line systems and aren't quite ready to put down the change to scrap the whole system.
i.e. compare this to the 6800 agp line (and ati's equivalent if you want).
AND, use a realistic processor for this graphics card range, i.e. intel 2.4's to 3.2's (and AMD's equivalents).
haven't been visiting as much lately and i'm seeing the quality going down...give me a reason to come back guys.
jzander - Friday, February 3, 2006 - link
I can not stand Best Buy, and the fact that Nvidia is getting into 'bed' with them brings them down a notch in my book. As for the card, does not look like a reason to upgrade from my 6800 GT AGP, card.Alphafox78 - Friday, February 3, 2006 - link
50+ people have complained and the article STILL hasnt been updated with more cards! what gives?!stephenbrooks - Friday, February 3, 2006 - link
--[50+ people have complained and the article STILL hasnt been updated with more cards! what gives?!]--Hey, I guess if you offered them some money they might do it. :)
puffpio - Friday, February 3, 2006 - link
It's only been a day...if they do update it, it may take a few daysI think the general concensus is
1) Add some benches of AGP cards to show it's true competitors as well as give an indication of what end AGP users currently have
2) Overclock the 7800GS and show that benchmark
3) Discuss the potential of softmodding the card, unlocking pixel pipelines, shader units, ROPs
sandman74 - Friday, February 3, 2006 - link
Agreed.
How about it Derek, any chance that this review can be updated or are we going to have to wait a while until your next AGP face-off that you alluded to in the review ?
Thanks,
sm74
JamesDax - Friday, February 3, 2006 - link
I too am looking forward to an update to this review. I do have one question though. Isn't the X850XT PE just a factory oc'd X850XT? I mean besides the clock speeds there really isn't any difference between the two. So that being the case didn't he in fact underclock those two 7800GS's and bench them against a oc'd X850XT??coldpower27 - Friday, February 3, 2006 - link
There is a difference, the X850 XT PE is an official SKU from ATI, even though it's only a bit faster then the X850 XT.This is the same situation that is occuring with the X1900 XT and X1900 XTX now.
Derek underclocked the 7800 GS to the point of Nvidia's minimum specs for this SKU 375/1200, but this doesn't represent the performance of the product we will be getting from certain manufacturers, such as the eVGA Superclock edition.
JamesDax - Friday, February 3, 2006 - link
That's a pretty weak argument. The fact is he under clocked those cards below there off the shelf speeds and tested them against what is essentially an overclocked X850XT. He could have redeemed himself if he had just included the benches for the 7800GS's at their factory oc'd stock speeds. As it is, he misrepresented the performance of those cards.coldpower27 - Saturday, February 4, 2006 - link
You misunderstand me.I am not arguing that this review is good, quite the contrary is what I believe as it stands now.
The 7800 GS should have been compared at Nvidia's reference design speeds and using eVGA's 7800 GS SuperClock Edition speeds as well as that represents the product the consumer will get.
Using Nvidia's reference designs speeds only represents the worse case scenario which isn't good enough.
unclebud - Friday, February 3, 2006 - link
just a quick thanks for reviewing an agp card solution -- rather thoroughly imo, i might add... peace.Spacecomber - Friday, February 3, 2006 - link
This card will spark alot of interest for people like myself who want to get as much life out of their AGP system as is practical before deciding that it's time to do a more complete system upgrade.In my case, I'm still running an Athlon XP at 2.3GHz, and since upgrading from a Ti4600 to a 6800GT, when Battlefield 2 came out, I have a viable, mid-range game system (it plays Battlefield 2 at 1024x768 with mostly high settings). This has allowed me to postpone that upgrade to a new PCI-E based system, for now.
So, the question is for me whether any new AGP card will provide yet another intermediate upgrade step before the inevitable upgrade to a PCI-E system. This means that I do want to know how the peformance from this new AGP card stacks up against the kind of PCI-E cards that I would likely to looking at (such as the 7800GT, probably the best bang for your buck PCI-E card, and the X1900XT, an example of what the top end cards have to offer). I also want to know how much faster this new AGP card is compared to the other top end AGP cards, such as the X850XT PE that you tested, as well as other cards that people are likely to be using now, such as the 6800GT/GS, X800XL, 6800 Ultra, etc. Not all of these cards need to be included, of course, but enough to give a feel for how the new card stacks up against the old guard.
Finally, but not least, I want to know about what sort of value this new card has to offer. How much is it selling for? Does this card seem to have alot of overclocking headroom, and if so, when running it at an overclocked speed, what cards is it comparable to?
Assuming that this new AGP card has enough horsepower to make it a significant upgrade from what someone is already running in their legacy system, how does the cost of this single component upgrade compare to the cost of a more complete overhaul to a PCI-E system of at least comparable speed (i.e., PCI-E motherboard, processor, and video card)?
Those are the sort of questions that I will be looking for answers to when I investigate what a card like the 7800GS AGP has to offer.
Space
bobbyto34 - Friday, February 3, 2006 - link
Nice review...It's not complete because we don't have comparisons with other AGP cards (6800/6800GT), but we can deduce it ourselves :) Maybe later AT will add these tests !
I want to keep my A64 3400+ but change my 6800GT to a more powerfull card without eating bread for 3 months... I was considering buying a 7800GT + nforce 4 socket 754 board or a 7800GS AGP.
I think I'll wait a few months to see what happens : the 7800GT + mobo costs nearly the same price than a 7800GS here in France, for more performance... But the 7800GT is already struggling in 1280 with recent games (ie Fear).
Wait and see...
Sorry for my English, i'm from france :p
BenSkywalker - Friday, February 3, 2006 - link
You underclock the boards from what is actually available....You leave out MOST of its direct competitors.....
You rig 20% of your tests to make the board lose to the only direct competitor you bothered to test.....
What is happening? If a part has issues then let them come out on their own- too high a price and not enough performance- OK, but let it come through on an honest basis. This review is something considerably below you ignoring it all together.
PrinceGaz - Thursday, February 2, 2006 - link
The 6800GT was for a long time the AGP card of choice, and the 6800GS has replaced it recently as it is cheaper and only slightly slower. Given how either of those two cards would have been what AGP users might have considered, or still be considering, the lack of a 6800GS in your benchmarks was unforgivable. For what it's worth I'll say a stock 7800GS is slightly faster than a stock 6800Ultra, and therefore has a decent margin over a 6800GT, and especially a 6800GS. The real strength of the 7800GS is in its overclocking headroom though, as you pointed out. That doesn't necessarily make it a good purchase though.Anyone who wants to hang on to an AGP based system could use a 7800GS to upgrade from the likes of a 6600GT or lower, but I really think their money would be better spent on moving to a PCIe platform and graphics-card.
JarredWalton - Friday, February 3, 2006 - link
The interesting thing is that AGP 6800 GS cards actually use NV40 chips, so they can unlock to 16 pipes and 6 vertex units (plus 16 ROPs) if you're somewhat lucky. They're basically like the X800 GTO in OC/unlock potential. Problem is getting software that will allow that. They *don't* OC hardly at all, in comparison to the NV42 6800 GS.If you want my opinion, this 7800 GS AGP stuff is too little too late. NVIDIA and ATI are both trying to get users to upgrade to PCIe for several reasons. First, it potentially makes them more money, since they also sell a lot of chipsets. Second, and probably more importantly, supporting older technology just makes more work for them. They would rather focus on one platform and forget about AGP. The conspiracy theorist in me says that 7800 GS was intentionally crippled just to help AGP die out.
FWIW, 6800 Ultra is overall slower, though in a few games it can tie/slightly lead the 7800 GS. With overclocking, 6800 GT/GS/Ultra AGP cards can't touch the 7800 GS - 450+ MHz on the core seems readily achievable. Overclocked, the 7800 GS easily comes out as the fastest AGP card, which makes it more palatable. Losing half the potential fillrate still makes the card less than perfect.
coldpower27 - Friday, February 3, 2006 - link
In the end though what matters is performance, and the G7x Shader enahncements plus 1.2GHZ memory speed, help it beat out the 6800 Ultra, not to mention be cooler running. Even with the disadvanatge of 8 ROP at 375MHZ. It just beats out the 6800 Ultra, and brings the 7 Series feature set to AGP, which is perfect as it doesn't hurt the 7800 GT. So far though I haven't seen any situations where 6800 Ultra wins...BlvdKing - Friday, February 3, 2006 - link
The price of this part is way too high IMO.coldpower27 - Friday, February 3, 2006 - link
Well given that the 7800 GT PCI-E has a MSRP of 449US and have been on the market for nearly 6 months and the 6800 GS PCI-E has a MSRP of 249US and been on the market for nearly 3 months, we have to give a few months for this part to settle in pricing.2-3 months should be ample time for this part to fall to correct pricing for the AGP, and remember it's competitors are 6800 GS AGP, as the 6800 GT and Ultra line for AGP supplies should be dwindling to basically nothing. Same thing with the X850 Series, only a competitor assuming you can find it at a good street price.
JarredWalton - Friday, February 3, 2006 - link
There are a few instances where 6800U wins out, but not by a huge margin. They're not shown in our benches, but http://www.firingsquad.com/hardware/bfg_geforce_78...">FiringSquad has them. The thing is, if you have a 7800GT in the charts, you get a much larger difference between 6800GS and 7800GT, likely due to the higher clock speed, additional pipelines, and twice as many ROPs. The 6800U never even gets close to 7800GT performance, I don't think.coldpower27 - Saturday, February 4, 2006 - link
I wouldn't call the instances where the 6800 Ultra wins out as significant, as that is the worse case scenario, most games do fine with a 2:1 ratio on Pipes/TMU vs ROPs as the 6600 GT has shown.Of course not for the 7800 GT, that is very obvious, G7x Shader technology coupled with a 25% higher bilinear texel fillrate and 16% higher vertex power, equal ROP power, gurantees a victory for the 7800 GT, typically I would say Shader Power > Memory Bandwidth in determining performance, unless you cripple memory bandwidth to crazy low levels.
Considering how crippled the 7800 GS is I am very impressed that it typically defeats the 6800 Ultra. G7x Shader technology helps out quitea bit as it has a core clock disadvanatge too.
Michael - Thursday, February 2, 2006 - link
Why only show underclocked results when the cards are retail parts that run faster out of the box?Why not compare to other AGP cards?
I'm very confused as to what AT was thinking when they decided to do the review this way. Unless the intent was to force your readers to look at other websites as part of some AT master plan.
Although "hard launch" is important, does every review have to dance around with definitions of it? The 7300GS was available to buy when it was announced (I saw one in a store the same week the announcement hit the web) and it looks like the 7800GS is available in retail stores in the USA right when the card was announced as well. Seems to me that both cards were properly "launched" and that both of the main card companies are putting "paper" launches behind them. How about making a longer comment if there is a real "paper" launch by either one and giving it a rest otherwise. In the case of something like the 7300GS, saying that the card will be available in different regions as per X schedule should be enough. Even with a full on "hard launch" there are always places in the world where cards don't make it day one. The real question is always the real availability of the cards - do they really exist to be bought in some sort of base line quantity? If yes, then leave it at that. Releasing a review on Thursday and then complaining that a card will only be at a major retail store 3 days later seems like a very faint complaint to me.
To close off that point - cell phones often are launched with an exclusive for one region or one carrier when they first come out. That is both marketing and matching initial production with demand. It also doesn't result in ranting and raving all over the place.
Michael
Cygni - Thursday, February 2, 2006 - link
From above:shabby - Thursday, February 2, 2006 - link
Whoa whoa whoa, wtf is nvidia thinking? The 7800gs has only 8 rops and the msrp is 350 bucks? F#$% that.JarredWalton - Thursday, February 2, 2006 - link
Hear, hear!I only have a couple of thoughts on why this was done:
1) For the non-conspiracy theorists, these are just G70 cores that had some defective ROPs and pixel pipelines. Alternately, maybe the HSI chip somehow has issues with the heavier load?
2) Option two is that this card is a bone thrown to the AGP crowd, and unfortunately the bone is already dry and picked clean. 8 ROPs basically relegates this card to equaling what was already available - at best. I would have much rather seen a 7800 GT or GTX 256MB (or even a 512MB - though those seen to be in tight supply due to chip yields).
3) Realistically, it may be a combination of both factors. They don't want *too* much power available for an "old" platform, but they also had a lot of 16 pixel pipeline chips available. I really doubt that all of the chips had 8 defective ROPs, but maybe some where 8, some were 12, and some were 16 and it made some sense to just release them all as lowest common denominator parts? Be interesting to see if anyone can figure out a way to attempt to reenable the extra pipelines and ROPs.... (Go RivaTuner!)
Considering that the 6800 GS is a 12 PP/8 ROPs chip and it only costs $200, I really wish this had been at least a 16 PP/12 ROPs release. That would have maybe made it worth the 50% price increase. Overclocking and possibly unlocking the 6800 GS AGP cards will get you really close to 7800 GS performance. Hmmm... Derek probably should have included some 6800 GT/GS/Ultra numbers as well. Oh well - maybe in a follow-up article.
sandman74 - Thursday, February 2, 2006 - link
I hate to repeat what the others have said, but this review was borderline useless to the very people who would come here to find out how well the card performs...
My other half has a 6600GT, and I have a 9800 Pro. How does the 7800 GS compare as an upgrade. NO IDEA is the conclusion I reached after reading your review, which has forced me to look elsewhere.
I did however find the article interesting to see how it compared with PCI-E cards, but thats about it.
I think you need to add in some more cards to THIS review, rather than letting everyone wait for another AGP comparison in a few weeks time.
manno - Thursday, February 2, 2006 - link
Is everyone overlooking the fact that this is going to be a part with a $399 MSRP?, and why no 6800GT AGP, or 6800GS AGP(same thing) an ommision like that is preaty bad.spinportal - Thursday, February 2, 2006 - link
So far the guesses have been between 300 and 350 USD for a plain vanilla or OC part.coldpower27 - Thursday, February 2, 2006 - link
the MSRP of the 7800 GS AGP is indeed 349US, the OC parts are about MSRP 379US, however we will have to wait to see how this translate to street pricing.The X850 XT PCI-E/AGP is MSRP 499US, Street 200-250
The 7800 GT PCI-E is MSRP 449US Street 275-300
The 6800 GS PCI-E/AGP is MSRP 249US Street ~200
johnsonx - Thursday, February 2, 2006 - link
Someone from AT should probably say this, but perhaps they think it's obvious:This article is about the launch of the NVidia GeForce 7800GS GPU. It is NOT about a particular vendor's board. Vendor boards are tested in vendor roundups or occaisionally individual board reviews.
While I agree it would have been *interesting* to see the vendor cards tested at their shipping clock speeds, that is really for another article. A GPU launch article should test either reference hardware or vendor hardware set to reference clocks, and this article does exactly that. Period. Quit complaining.
People should also keep in mind that neither AT nor most other sites re-run every benchmark on every card for every test. They test the new card on their standard hardware platform using current benchmarks, and then use benchmark numbers for other cards from recent tests. Older cards were last tested on an older platform using older benchmarks, so no comparable benchmark numbers are available.
AT isn't going to re-benchmark a pile of older cards unless doing so is the POINT of the article. This article, again, is about the launch of a new GPU, not a 'Mid-range to High-end AGP Comparo' or '7800GS Vendor Roundup'.
spinportal - Thursday, February 2, 2006 - link
Then why doesn't AT just have a Video Chart that can pull similiar Min/Max/Avg FPS for each type of test at each resolution / AA / AF setting at-a-glance? Instead we have to muck around dozens of articles and have no generic yardstick. Why include the 7800GTX then if not to compare? Stop being an apologist.spinportal - Thursday, February 2, 2006 - link
Let me be more clear..Why can't AT build an result aggregator database / spreadsheet?
Cygni - Thursday, February 2, 2006 - link
Your kidding right? Ok 1) You do realize that more than one person does reviews for Anandtech. 2) Due to 1 (see above), it would be a weebit difficult to ship the standardized review rig to each persons house. 3) The tests done on games change over time. When 6800 Ultra's were being reviewed, there was a totally different set of games being played. 4) As with 3 (see above), hardware changes over time. If they wanted to test on the same system for every graphics card YEARS apart, they would still be using a 486.This isnt Toms. If you want the VGA Charts, go get them.
spinportal - Friday, February 3, 2006 - link
Im sure this being a *technical* site, one can factor in plain vanilla CPU/mobo/chipset and formulize those effects vs. the video card. I guess Futuremark's database is closer to where I'm going, where its a composite scoring system, and then an overall, but not divergant and possibly too specialized and even misses the market's techniques. Tom's chart is a bit fuzzy to read it makes my eyes blur at times ;) I mean I could try to scour thru AT and do my own chart in excel (and factor in platform %), but who has timeSharptooth - Thursday, February 2, 2006 - link
First, I enjoyed this review as it has relevance to me (NF2 AGP user). However, as already mentioned, it would help to compare this card to other AGP cards (6600GT, 6800GT/Ultra, X800XT/PE, X850XT/PE) because AGP systems use these cards. That way, it'll help to access any inherent value (if any) to purchasing. Also, while most reviewers have chosen A64 AGP systems, benchmarking on other platforms (like Socket A) would be extremely appreciated as many (including me) still use these systems.--Fernando
mrkkbb007 - Thursday, February 2, 2006 - link
As I said in my earlier post. Head over to Tom's HW for comparison with a 6800 Ultra, the 7800 GS beats the 6800 Ultra AGP in almost all the latest games. The X850XT is still tops, but can't do the eye candy.ciparis - Thursday, February 2, 2006 - link
I'm a little shocked at the decision to ignore the actual speed these cards run at out of the box, and instead show an artificially limited card -- exclusively.
I wouldn't have minded seeing the "published spec" score, but not at the expense of knowing what you get if you walk into a store and walk out with one of these cards. Without that, this article is glaringly incomplete. I'm left little more informed about how an actual retail product performs than I was before I read it. That doesn't sound like AnandTech.
formulav8 - Thursday, February 2, 2006 - link
I agree that the reviews are subpar. I think anand should not be delegating as many reviews as he has been. You can ALWAYS tell which reviews he does from the moment you start reading it (a good thing). I guess he is just to busy these days to do more of them.Jeff7181 - Thursday, February 2, 2006 - link
Where is the 6800 Ulra or 6800GT in that review? How about a 6600GT? Come on guys... these are the cards that people who would consider buying a 7800GS own right now. Not including a 6800 or 6600 series in that review makes no sense at all.On a side note... AT has changed since I started visiting... I used to be able to get all the info I could ask for right here... now I have to visit 5 other forums to get the info I want, and that's not a good thing.
spinportal - Thursday, February 2, 2006 - link
here here.. this article is weak; run the cards at their OC speed, and then LABEL the REFERENCE Speed 7800 GS on the charts, since this is MISLEADING and *NOT* intuitive in comparisons when putting money down on the table.From a competition performance-cost standpoint, let's talk about bang for the buck here. This isn't a PREVIEW, this is about a card coming out in days.
How does the X1600XT 256MB, 6660GT 128MB, 6800 128MB rate in the $130 range.
Then its the sub 200$ range with the 6600GT or 6800 256MB versions.
We can all go thank Tom's HW site for the 2005 VGA charts to get a rough ball estimate due to the X850XT PE (which was not 210$ but rather in the $400+ range (from using pricegrabber). Not only that but they do compare the 7800GS against the 6800U and the BFG card is modestly beating the old tech. For the 300$ target price of the 7800GS AGP, there is a vast wasteland and no competition pricewise, but I feel its not justified to say it will drop 6800GT/U prices since it is no longer being manufactured (still going to be rare, and same performance, so why even bother getting an old large hot stinker of a card on E-bay-like services when you can get brand-spanking new tech?)
I'm still leering of why nVidia cannot see to get HDR and AA/AF to play nice, so I personally don't see much advantage in upgrading from a 6600GT / 6800 GT / U part just for SM 3.0. If the 7800GS is ~30% faster than a 6600GT, ~11% faster than a 6800 GT, and barely ekes out victory over a 6800 U, it doesn't do justice to upgrade unless you need to make the leap from TurboCache or integrated. I do hardware (& sys building) recommendations for many people and want to steer them in the right direction and make them feel they got the best value. Will this make people put off buying an M2 & PCI-e mobo for another year and stick with AGP? I'm not in the business of making predictions, but for me personally, I would rather get an X2 chip, mobo, and X1900 or 7800GTX (or maybe even the refresh to choose from), possibly an Creative X-Fi entry (but highly doubtful, shrug to the 5.1 audiophiles because there is a lack of a stand out digital spdif 5.1 headset - this boggles my mind and is another side topic) and 2GB of DDR2-667 RAM right about xmas 2006. I never find a truly good integrated PC builder that satisfied my DIY urges.
mrkkbb007 - Thursday, February 2, 2006 - link
The 6800 Ultra looses the lead in comparison to 7800 GS. Head over to toms hardware to see benchmark scores. I have an Opteron 146 w/ socket 940 and 2GB ECC-DDR400 that I received from work, but it has an ATI 9600SE. The 7800 GS is exactly what I have been waiting for since upgrading to a 939 with PCI-E is useless and will cost me more than a 7800 GS. The 7800 GS should give my system usefull life for a year at least, maybe more and it will be half the price of a 6800 Ultra.Cheers
phusg - Thursday, February 2, 2006 - link
I'd be interested in a comparison between the XT1600 and the 7800 GS AGP cards. They are my 2 candidates for a low power, high performance AGP card for my HTPC. I think many other HTPC owners would be interested in this as well, especially how the HDTV performance and quality compares on older CPUs.Mithan - Thursday, February 2, 2006 - link
Wish you guys had included a 6800GT or Ultra in here.puffpio - Thursday, February 2, 2006 - link
Could you add one more line showing how the card benches at the EVGA default overclock? Or maybe do your own overclock and test?Since this is the last bastion of AGP performance, I don't think any of us AGP users will want to stock clock this..
DrZoidberg - Thursday, February 2, 2006 - link
Yeah i cannot see why Anandtech decided not to show default factory o/c scores from evga. Sure if AT wanted benchmark with 7800gs underclocked but ALSO include 7800gs standard retail overclock benchmark.I mean the consumer who buys the evga 7800gs is not going to underclock the card by 20% from 460/1.35 to 375/1.2 and lose 20% performance.
Spoelie - Friday, February 3, 2006 - link
I think you missed the point of this article. It was not meant as a review of the particular brands, but of the card nvidia just released. Those are the reference clocks and that's what nvidia is putting out right now, it is the baseline you can expect all the 7800gs cards to do, not what some brands may do. That they just happened to use an evga card to show it doesn't really have anything to do with the matter.I'm pretty sure there will be a followup article discussing the differences between the brands and their clocks, with some lower end comparison cards everyone is craving for thrown in for good measure. I hope you can leave your pants on till then.
mindless1 - Friday, February 3, 2006 - link
If the context is seen as those cards, for that price, we can reasonably assume only someone ignorant of clock speeds would pay same price for a lower clocked stock card, else they'd planned to really o'c it themselves.So, we do care what some brands do because it's not that we're buying a "7800GS" if we do, rather we're buying a 7800GS for 7800GS performance levels which are not a conceptual thing but an actual product.
Who wears pants? ;)
phaxmohdem - Thursday, February 2, 2006 - link
No kidding, I was actually VERY upset and confused as to why you downclocked the retail parts, or at least didn't run tests both ways. VERY poor reviewing on this one IMO, not typical of Anands norm. IF I go out and purchase a eVGA 7800GS for my FX-51 AGP system, I"m sure as hell not going to underclock it the second I get it. I would most likely just leave it at stock and call it a day.PLZ for the love of God, show us the true out of the box performance numbers, and compare it to other cards like my AGP x800XL, or the 6800GT/Ultra AGP parts.
bldckstark - Thursday, February 2, 2006 - link
This is my idea of a hard launch --> A buyer has to be able to purchase the part on the day of the launch, or it is not hard. Not ALL buyers, but enough stock so that parts are available to purchase nationally for at least 24 hours. Nationally meaning the part would have to be at a national retail chain (Best Buy, CompUSA, Fry's, etc.) or online where a price search engine can find it.So my definition of hard launch is, parts nationally available for 24 hours, or it is soft.
What do the rest of you think? Fair or not?
dqniel - Thursday, February 2, 2006 - link
Very fair.dqniel - Thursday, February 2, 2006 - link
Very disappointing. This card at stock is $300 yet has worse performance than the x850xt in most cases which is about $210 now...Perhaps I need to see some results with this thing overclocked and the pipes unlocked (if they truely can be)
kmmatney - Thursday, February 2, 2006 - link
The AGP version of the X850xt is way higher than $210, more like around $400. IMO, its pointless to get it. You can just get a new mobo for the price difference, especially if you already have a good processor.When I went to PCIE, I bought a new mobo, Sempron 2800+, and X800GTO2, all for ~$350 $350. Then then sold my old setup for $150, so my net cost was $200, and I have a much faster cpu as well as a faster video card. Way better than paying $300 or so dollars just for a new video card.
OvErHeAtInG - Thursday, February 2, 2006 - link
Well to be more accurate, the X850 XT AGP can be had for <250. Heck I bought an X800 XT PE last summer for $249 shipped, and some people are now getting the 850's for less than that. You don't see these deals widely advertised, as it would compete with the 800GTO's.Now to be fair, dgniel compared street prices of the X850 to MSRP of the 7800GS which is apples-to-oranges. But I don't think too many people pay $400 for X850 XT's anymore, seeing as a 7800GT will match its performance for 100 less and better feature set.
Avalon - Thursday, February 2, 2006 - link
So how many pixel pipelines does the card have, 16? Let's see some benches of it overclocked.rcxplane - Thursday, February 2, 2006 - link
Check out this reviewhttp://www.ngohq.com/home.php?page=articles&go...">http://www.ngohq.com/home.php?page=articles&go...
If the numbers are correct it shows an overclocked 7800GS with stock cooler getting a higher score than the 7800GT.
I have read in several places that people have been able to unlock the unused pipelines and shaders using RivaTuner.
If all this is true then Its a must buy for all of us still using AGP
I agree with all above. I would also like to see the card compared to the AGP 6800GT and Ultra with OC scores.
Ezza - Saturday, February 4, 2006 - link
This article needs to be taken with a grain of salt since the 7800GS in the article is a PCI-E engineering sample. In this article the 7800GS scales very well when overclocked, hopefully the AGP version will show similar gains when overclocked.Stas - Thursday, February 2, 2006 - link
I have x800 Pro o/c. I really want to upgrade but buying a new mobo + video card is a bit too expansive. If only I could get a better video on the level with 7800GT and x1800XL, I would be happy. Unfortunatelly, 7800GS just doesn't cut it (you've seen the tests). I wish they come out with AGP versions of 1800XL and 7800GT, I would have gotten one. But then again, ATi and nVidia both make chipset with PIC-E, thus it is profitable for them to have people buy new mobos with PCI-E. :(NullSubroutine - Thursday, February 2, 2006 - link
it seems rediculous to me that with an agp part, you compared...not other AGP parts, but high end pci-express cards. i have an agp system, i want to know how much better it is than other AGP cards out there, not a top of the line card, that would require me to upgrade my processor and mainboard.again, you have done so in the past, so i request it once again. many people buying this product run either 10x7 or 12x10 with and without 4x 8x. these were left out in the line graphs, i request them once again. otherwise without both these things, i think it leaves us people who would considering to buy this product, needing to go elsewhere to find out how it compares, and what is a good buy.
mindless1 - Thursday, February 2, 2006 - link
Agreed. Not everyone is as interested in how a card does at all kinds of maxxed out settings and features as we are at what measures would be needed to use the card for modern games, playable rates. If a feature entails a significant performance penalty, but one has an AGP board, of course they'd disable the feature when necessary.I also fail to see the logic in lowering card clock speeds when (all) 2 samples were shipped at a higher rate. While one may not want to manually raise the speeds for the tests, the out-of-box speed a card uses should be considered stock no matter what nVidia spec'd. IOW, I don't see a lot of people underclocking their cards, it would be a scenario never seen in gaming use, IMO.
coldpower27 - Thursday, February 2, 2006 - link
X850 XT non PE
6800 GS (AGP)
6800 GS (PCI-E)
EVGA 7800 GS at it's factory OC Settings.
6800 Ultra
6800 GT
coldpower27 - Thursday, February 2, 2006 - link
In addition to the aboe now that I think of it I would like these 2 to be included as well.6800 AGP
6600 GT AGP
DerekWilson - Thursday, February 2, 2006 - link
Thanks for the suggestions, we will certainly consider including these cards in future AGP performance comparisons.Thanks,
Derek Wilson
Boushh - Thursday, February 2, 2006 - link
I agree with others that you should have picked other cards for a comparison. Specialy considering your introduction words in the article.The point is that it's about users who still use up-to-date AGP systems (like me) that would like to know if this card is worth the money and wait a little longer to upgrade to PCIe.
Therefore the comparison should have been between this card and other AGP cards at the same price point, or more expensive ones. Now you only tested the X850. A card that most people reading the test will probaly not even consider as an upgrade. And comparing it to the X1900XTX and 7800 GTX 512 Mb ? Come on ! Like these cards would be viable upgrades considered by the people who are intrested in this card..
Besides, if I want PCIe numbers I can read any of your other tests.
There are also no noise, power consumption, and OC results (I bet the eVGA would beat the X850 hands down). And the test was done on a medicore ULi based board instead of a high-end S939 board.
Sorry for the harsch comment, but I simply had to say this ;-)
MrKaz - Friday, February 3, 2006 - link
80% agree.You only failed on one post.
"And the test was done on a medicore ULi based board instead of a high-end S939 board."
You forget that there arent no longer high eng AGP systems.
The existing ones are "outdated".
bob661 - Thursday, February 2, 2006 - link
Isn't this an oxymoron?APKasten - Thursday, February 2, 2006 - link
No. It could be considered a contradictory statement. An oxymoron is a phrase, not a sentence. ;)But I still disagree with you. Since AGP is still a viable platform for playing current games. In fact, the 7800 GS keeps AGP systems relevant, because it puts an AGP card on the market that utilizes the most advanced kinds of mainstream video technology.
Alphafox78 - Thursday, February 2, 2006 - link
dittoJWalk - Thursday, February 2, 2006 - link
I would be interested in seeing the scores from the eVGA version of this card. It has significantly higher clock speeds. I would expect that it gives quite a bit better scores. As it stands now though, if I were looking for an AGP card, I would look toward the X850 XT. The non-PE version can be found for $250 or less. The stock-clocked 7800 GS is a nice card, but not a very good deal at $300 or more.So, to conclude, looking forward to some numbers from the over-clocked eVGA 7800 GS. :-)
neogodless - Thursday, February 2, 2006 - link
Unfortunately, comparing this card only to cards that are PCI Express only or you won't be upgrading from (like the X850XT PE) is kind of pointless...I say this because I think the question worth asking is "Is this card a viable upgrade for AGP users", so you would want to compare it to cards like the 6800GT, 6800, 6600GT, and perhaps those Radeon counterparts. Then, having the X850 line in there for comparison would give you a good idea as to which you'd want to upgrade to.
Alphafox78 - Thursday, February 2, 2006 - link
I second this, why not put in a 6800GT or Ultra? if you have a 6800 card it would be nice to know if this card is faster. I stopped reading after I found out there were no cards I found usefull to compare it against..OvErHeAtInG - Thursday, February 2, 2006 - link
Also, both the X850 XT and X800 XT were available in AGP. I think the poster was looking for a way to get SM3.0 by swapping out his X850 XT AGP ?DerekWilson - Thursday, February 2, 2006 - link
The PCIe numbers are useful to AGP users in that it shows where the cutoff between the highest possible AGP performance is and the next step up in performance possible on a PCIe platform.That being said, we are planning on looking at the AGP version of the X1600 as compared to the AGP 6800 GS, so we will certainly revisit this issue.
Thanks,
Derek Wilson
fishbits - Thursday, February 2, 2006 - link
Huh? If you're using AGP and looking to upgrade the video card, what does it matter what the PCIe cards are scoring? If you'd decided to go PCIe, pretty much any of the cards out would be better in one way or another. Maybe just throw in one 7800 GT or such for that comparison. What's "useful" to someone looking to uprade an AGP card is how it will perform compared to what he has, a 6600GT, 6800GT, 6800GS (AGP version), a 97-9800 and a few of the XXX800XLXXLX X's. That's what they need to be able to see if the card's worth it, or maybe it is time for them to switch to PCIe. Speaking of which:AT's been on this jag for a while now. It was just 18 months ago you were recommending an AGP board as your "Editors Choice" to readers. While some upgrade more frequently, I don't see the guy who buys a top-of-the-line mobo and expects to get two years out of it as in the same category as the guy trying to nurse his Sempron system along. Oh well, at least it's better than when PCIe was available for less than a year and offering zero performance benefits compared to AGP, yet AT was insisting that avid gamers shouldn't expect to see any new AGP products.
/rant
GoatMonkey - Thursday, February 2, 2006 - link
You forgot a few "T"s.
hshendon - Thursday, February 2, 2006 - link
Anyone know what the numbers are for the same tests for a 6800 Ultra?I think I am happy to see that there is a 7800 part for the AGP users out there like me, but I am not sure if I should get it because I can never find comparisons to the 6800 Ultra.
I know I would get the better shader model and more pixel pipelines, but what does that really mean in a head to head comparison?
Jojo7 - Thursday, February 2, 2006 - link
I agree.We need AGP 6800U thrown into the mix.
fl0w - Thursday, February 2, 2006 - link
did you compare SC with SM1.1 path or did you run it on 3.0 for the Geforce?Didn't see it written somewhere so I thought I ask, because it would make a big difference....
RamarC - Thursday, February 2, 2006 - link
"Our 7800 GS was tested with SM3.0 enabled and all the options while the X850 XT PE does not support these features.The price the 7800 GS pays for enabling HDR and other SM3.0 eye candy is that it is limited in playability to lower resolutions. Disabling the SM3.0 features would give performance quite a boost without AA (SM3.0 features are not available with AA enabled). "
fl0w - Thursday, February 2, 2006 - link
thanks didn't see that... still can't find it btw, but at least now I know ;)Cygni - Thursday, February 2, 2006 - link
Im a little bit dissapointed, overall. While I guess the 850XTPE is a pretty lofty card to try to beat, i was hoping for a little more power from this piece, to keep my AGP rig alive a little longer. :DMight have been a good idea to compare AGP and PCI-Ex versions of this card on the same board with a ASRock 939Dual-SATA2. I know weve gone over that terriroty before, but it would have still been interesting to me to see it again using this higher end part that likely would take a little bigger hit going through the HSI bridge. Just an idea.
DerekWilson - Thursday, February 2, 2006 - link
To be very clear -- in case the title didn't sink in :-)The 7800 GS will not have a PCI Express counterpart. This is an AGP only card.
The engineering sample we tested last year was simply that and will never be anything more. There are a few differences as well, so a direct comparison isn't really possible.
Thanks,
Derek Wilson
Cygni - Thursday, February 2, 2006 - link
Doh, didnt get that the PCI-Ex version was NEVER coming out. Just thought it was coming out later. I read the title more as "NVIDIA's AGP Only VERSION of the 6800 GS" Im a jerk-face. My bad. :DCygni - Thursday, February 2, 2006 - link
PS, throwing in the engineering sample would have still been pretty interesting, for me at least. :)gooser - Thursday, February 2, 2006 - link
Or maybe, better yet....the 6800 GS agp? Those two would be the one of the top choices for an AGP upgrade.kalaap - Thursday, February 2, 2006 - link
I think its great that nvidia and ati are still releasing AGP parts. I've beeen planning on replacing my 9700pro so I like having so choices even if its very limited.artifex - Thursday, February 2, 2006 - link
Maybe; do those chips come with DIPs? :)
Cygni - Thursday, February 2, 2006 - link
Beat me to it. :DDerekWilson - Thursday, February 2, 2006 - link
Gotta love those crazy typos that OpenOffice.org doesn't catch for me. Fixed.