VisionTek CryoVenom R9 290 stock sold out in an hour

[c123456]

Not sure how many people noticed, but VisionTek had these come back in stock today at 11a est, and by noon they were sold out. I managed to get one of these, and eagerly await being able to use it. For reference, it was $557 to buy it.

Their site claims it out performs a 780 Ti and seeing as it was at least $300 cheaper than a water cooled 780 Ti I think it's well worth it. That said, anyone know of benchmarks for the card though?

 

[Brantyn Gerik]

http://hexus.net/tech/reviews/graphics/65789-amd-mantle-battlefield-4-performance-ruminations/

It doesnt out perform a 780ti unless you have a crappy gpu, or a VERY amd optimized game.

Now, for the price however, thier 290x just simply can't be beat as they are underselling the 290 with a water cooler (i read the article on it but dont remember if it was 290 or 290x)

so in all, ITSAGREATDEAL!! but still not better than a 780ti. For the price though, it's about as close as you can get and in metro last light for example, you will either tie the 780ti or beat it by 1 or 2 fps. this is however, a heavily amd optimized game, and in something like bf4 or say tomb raider, you will still see higher numbers from the 780ti, not to mention the 780ti is one of the best overclocking cards that has been made to date according to most reviewers, while the 290 and 290x seem to be hard pressed to overclock even under water cooling.

 

[Brantyn Gerik]

The benchmarks will be on par with any other 290 with a waterblock on it, as it is still a reference pcb underneath the waterblock

 

[hcforde]

The benchmarks will be on par with any other 290 with a waterblock on it, as it is still a reference pcb underneath the waterblock

^That is like saying every card with the same chip from every vendor is the same. They are not!. Until the benchmarks are out, broad based judgements should be withheld. Yes, this card is waterblocked, but it is a custom waterblock that covers all the critical components and not a one size fits all type block, it is also overclocked.

Maybe it is not faster, but let's allow the numbers to speak for themselves. I really do not thing Visiontek would run benchmarks with a different and much lower performing CPU to make their claim.

 

[Brantyn Gerik]

The 290x's in general are not great overclockers, and the board underneath the cooler is still a reference board. if you want to see similar performance numbers, find benchmarks of a 290x with a full cover waterblock and you'll see roughly the same numbers (this company does not bin gpu's so you wont see anything like you would say from a binned chip from asus or toxic under the same level of cooling)

Also you really think companies are above inflating their benchmarks? Both companies do it. hell the 290x was released as a titan killer, when most people don't realize that the 780 when fully overclocked still beats both 290x and titan (when both overclocked as well) in almost all games at 1080[ and 1440p (i have run these benchmarks for myself, as well as collected averages across multiple review sites for reference cards, and for best performing aftermarket cards). the exception being now, multi card amd configs with mantle, as mantle does significantly increase crossfire performance, but single card performance gains are negligible with a good processor.

also, if you look at the benchmarks between the slowest aftermarket 290x, and the fastest, there is no more than 10% difference. it's still more of a silicon lottery draw than good cooling vs great cooling.

Update: I have found benchmarks for this card on youtube, and they are pretty much on par with my initial assumptions. while yes temps are low as to be expected on water, it still only overclock roughly as well as any other 290x reference board on water cooling (which surprisingly still isn't great) usually with 1150 on core clock being rare but achievable, and an average overclock on the memory being possible as well.

 

[c123456]

@Bryanty: Perhaps you missed something. This is a 290 (not a 290X). If you're telling me it performs as well as a 290X with a waterblock attached then I saved myself quite a bit of money. Seeing as a 780 with a waterblock preinstalled is going to run north of $700. For performance/$ analysis it'd probably be a good comparison between a 770 Hydro Copper, being as its only $30 more than I paid.

That said, I plan on running the card at larger resolutions (either 1440p, or tri-1080p... can't decide). If I recall correctly, AMD is extremely competitive in these ranges.

 

[Brantyn Gerik]

I was basically just ranting to the guy above my last post. I fully agree that this company has undersold EVERYONE with this specific card, and compared to anything else it is a great deal, as basically they are making less money from the card. That being said, if it's a 290 (only read the article once) and is roughly $500 something, then it's not as big of an undercut as you may think. You also have to take into account the between the 290x and 290, there is very little performance difference for the greater price tag on the 290x.

As far as performance is concerned however, a 780 on air when overclocked will still outperform this card, but they will end up being roughly the same price instead of just a little cheaper (the 290 being cheaper than 780 i mean). considering it is a reference board underneath that watercooler, the only thing impressive about it is the current price vs the price of every other north american 290 version card. also at current pricing, a 780 with a waterblock can be had for around the 600-650$ mark, and will outperform a single 290 at all resolutions. It currently stands that the only way to get more performance out of the amd side than the nvidia side with dual cards combined with mantle. single cards with mantle just don't really do it if you already have a strong processor, and even up to 4k the 780 holds it's own, and the 780ti just blows them away (though being a higher priced card this makes sense). Mantle on the other hand, gives a HUGE benefit to crossfire in such a way that dual 290's will likely outperform dual 780's in every mantle application by a wide margin.

Conclusion: this card is unbelievably priced, but only because it is from a private seller, and even given the price and what you get, the performance is still roughly the same of a 780 on air before overclocking, if you are comparing single card to single card and the 780 is roughly the same price.

 

[hcforde]

The 290x's in general are not great overclockers, and the board underneath the cooler is still a reference board. if you want to see similar performance numbers, find benchmarks of a 290x with a full cover waterblock and you'll see roughly the same numbers (this company does not bin gpu's so you wont see anything like you would say from a binned chip from asus or toxic under the same level of cooling)

Also you really think companies are above inflating their benchmarks? Both companies do it. hell the 290x was released as a titan killer, when most people don't realize that the 780 when fully overclocked still beats both 290x and titan (when both overclocked as well) in almost all games at 1080[ and 1440p (i have run these benchmarks for myself, as well as collected averages across multiple review sites for reference cards, and for best performing aftermarket cards). the exception being now, multi card amd configs with mantle, as mantle does significantly increase crossfire performance, but single card performance gains are negligible with a good processor.

also, if you look at the benchmarks between the slowest aftermarket 290x, and the fastest, there is no more than 10% difference. it's still more of a silicon lottery draw than good cooling vs great cooling.

Update: I have found benchmarks for this card on youtube, and they are pretty much on par with my initial assumptions. while yes temps are low as to be expected on water, it still only overclock roughly as well as any other 290x reference board on water cooling (which surprisingly still isn't great) usually with 1150 on core clock being rare but achievable, and an average overclock on the memory being possible as well.

How do we know the person on Youtube did not use a crappy CPU with the CryoVenom card? LOL! I looked and did not see a CryoVenom review. I did see (1)a review that had a 290 under liquid cooling but it (2)did not give the resolutions of the review. Also the (3)Visiontek cooling block is a custom CNC machined block that covers ALL major heat producing components while the block in the review was a generic block. Also, (4)they only did 1 game and difference was 2 frames in Nvidia's favor in "FarCry 3".

(5)Another issue is whether Visiontek uses Elpdia or Hynix memory chips, the latter being very good in the overclock area. If you are just trying to win an argument to satisfy your personal ego online then continue with your rant. If you like honest results then by all means let's wait for a proper review of the actual product under acceptable conditions. here are at least 5 areas that need to be addressed before a proper assessment can be made about the CryoVenom.

 

[c123456]

@hcforde: The VisionTek Cryovenom card uses a normal EK Waterblock with the backplate. I know this because I own it, and they didn't even bother to remove the EK branding (granted they added their own logo to it) in addition to the EK coupon they gave me in the box (which for some reason doesn't work).

That said, 2 frames in Farcry 3 is hardly noteworthy as you've pointed out.

 

[hcforde]

Thank You, and that has been basically my point from the beginning. We should not be judging an unknown with a blanket statement. I put my name on VisionTek's email list after seeing your post. I had no clue these were being offered. I like clearer information on a product(especially a unique product) before I make up my mind about it. Bashing an unknown is counter-productive. I had some XFX DD's on order until I heard about their overly hot VRM's.

I look forward to your response after using the card.

 

[c123456]

It may be awhile if you're waiting on me. I'm doing some pretty time intensive case mods with it and various other things. As of right now I've only got about 40% of the computer purchased (in terms of cost). I'd probably estimate towards mid to late summer at this point.

 

[Brantyn Gerik]

@hcforde I have been in contact with the owner of a cryovenom card, and he confirms it is a reference board with elipda memory modules. The waterblock on the other hand appears to be a replica EK fullcover block, though his had the branding removed.

He also has overclocked it to 1157 on the core and only managed to get 6100 on the memory. For benchmarks, he has run the metro LL at 1080p and 1440p, getting 47.8 at the maxed 1080p and 40.2 at 1440p. this is now roughly identical with a 780, for the unigine valley benchmarks he got 69.1 at 1080p and 61 at 1440, this being about 1-2% slower than some of the aftermarket 780s when overclocked (so within margin of error)

for cpu, he uses a 4770k @ 4.2, 16gb corsair dominator plat @1866, psu- corsair rm 850, msi z87 gd-65 gaming motherboard, and crucial m4 ssd 120gb for benchmarks and games, windows 8 OS

Just because a card has a better cooler, or similar cooler does not mean great performance gains. Yes, his sample is clearly highly binned (or he just got lucky), but the reference board can only do so much without below ambient cooling, and even the chip itself is only capable of so much under air or water.

The lightning reviews have just come out, and with its' air cooling, it beats this card (to be expected though).

My WHOLE argument is that the performance promises by the company are downright misleading. This single card will beat no single 780ti in a game and will just barely edge past an overclocked 780 in some cases. With mantle, the 5% is enough to push it just past the 780 in single card solutions. Without mantle in sli, it does beat the sli 780, but does not beat the 780ti in sli. With mantle however, crossfire 290 will beat sli 780ti...without mentioning these specific circumstances, they are completely misleading consumers (i will admit, no matter what they say or promise for this card, the price is just plain awesome in comparison to anything else).

Now even to compete with mantle, microsoft has rumoured a new version of direct-x, so we may not even see mantle as a deciding factor in the gpu wars anymore (WAY too early to say though).

 

[hcforde]

It may be awhile if you're waiting on me. I'm doing some pretty time intensive case mods with it and various other things. As of right now I've only got about 40% of the computer purchased (in terms of cost). I'd probably estimate towards mid to late summer at this point.

No Problem, Just send me the card and I will return it to you when you are ready for it. I'll post the benchmarks for you so everybody can see how generous you are. Look forward to your reply. LOL!!!

I am waiting on VisionTek to get back to me. Tech guy there said they are suppose to get a big shipment in soon. They get the boards bare without heatsinks on them and go from there. I have another Gigabyte coming fairly soon (windforce R9 290) via Amazon to match up to my lonely Gigabyte reference board(with Hynix memory, I hope the Windforce has Hynix). If so, it will be a hard choice of whether to keep the Hynix based boards or get the liquid-cooled Visiontek's.

Brantyn:
I talked with a tech at Visiontek yesterday, He is going to get back to me on whether all the CyroVenoms have elpida memory. If so then that is the biggest drawback for the board. It would seem that the 290X boards should all have Hynix and the R9 290 should have elpida UNLESS, they are a specialty board like this one so it can take full advantage of what the board as a whole has to offer. I will be gaming on 3 2560*1440 monitors and a 4K monitor which is where the R9 cards are the happiest. Those are the only number I am really interested in. That is why I like [H]ardocp's reviews so much. I believe that 4K is the future of enthusiast gamers especially when the HDMI2.0 panels start to appear.

I will admit it is strange that Visiontek did not shop the card around to be benched by the different sites. but they don't usually do that. They are very low key.

 

[Brantyn Gerik]

well based on the memory overclock, being a little bit higher than the stock one is capable of (they say 1250 but most people can get 1350 out of the reference board) i'd say they are pretty set on elipda modules, and from the anandtech overview, they seem to believe that there have been no modifications to the board itself other than maybe a binning process.

i totally agree that 4k is the future, but is still not consumer available enough (still $800 for the cheapest consumer model) to worry about with the current generation of cards. perhaps by the next amd release we will see 4k being more popular, and that is when you may see 4k become really popular, but other than that I don't expect 4k to take off until maybe next year sometime (when 120hz 4k tech comes out it will make 60hz 4k much more easily affordable).

Sidenote: you do need to try and find more than just one reviewer, as the silicon lottery does favor some over others (linus has a shit 290x, but a king of a 780, TTL had a shitty 290x as well, but a pretty good 780. It seems like the only reviewers i can find that actually DID get a good 290 or 290x were anandtech and (H)ardocp, while virtually all others tended to have poor overclocking samples of the 290 and 290x. Were i a reviewer myself, and had access to both cards. Icurrently own an asus 780 that i fully admit to being the holy grail of the 780 chips @1256 on the core and a +486 mem offset ((yes, samsung modules ftw)), but when doing comparison numbers from other review sites, none of the 290x's even come close (ok well they come close, but 72 in unigine valley vs 70.2 or so on the highest 290x i could find) for the price difference found in north america right now. At base prices, the 290x wins, absolutely hands down for what you get vs what you pay; but the nvidia price drop, combined with the amd price hike, has made the amd side a MUCH less favourable option in north america.

With the performance gain in single card setups with the 780, i may not win at 3x 1440p, but at a single 4k it's neck and neck; and I'm fine with that because i make enough money to live comfortable, but am not about to spend over $2000 on a personal computer, and then another +$1000 on monitors, because that simply isnt sensible to me. when a 4k monitor is under $500 for a decent one, the hardware can run modern games at the level today's hardware can run today's modern games, then this will be an option. Considering my single 780 i5 rig scores within the 90th percentile of computers in 3dmark, i would expect the masses would be relatively in the same or worse boat than I am.
(quick edit) the 3dmark benchmark was run over the christmas break with 3570k@ 4.8ghz for time of the statement's sake

 

[hcforde]

Order of preference in memory
Samsung
Hynix
anything but Elpida
Elpida

My reference Gigabyte R9 290 - Hynix
2 Asus R9 290X DCU2 OC cards - Elpida ( I thought for sure these would come with Hynix)
Gigabyte R9-290 Windforce - ?? will be here by the 15th

I am under the impression that since they are now selling all they can make they are using what ever memory they can to meet minimum requirements. Getting cards out is first job1. The memory makers are probably up against the manufacturing wall also.

I just hope there is no plant fire of natural catastrophe that occurs.

 

[c123456]

It'd seem that because of the manufacturing wall we (as in the consumer) would be the ones to benefit. If the only thing available is Samsung memory at the time of production what are they going to use? The same is true for Elpida, but given the opportunity to buy Elpida memory they probably would because of price (assuming here).

That said, I plan on gaming at a higher resolution too (3x 1080... not sure if portrait or landscape, but probably landscape).

 

[hcforde]

Way too many cards are using Elpida rather than Hynix or Samsung on R9 290X cards except the MSI Lightning. The benchmarks are amazingly better than boards with Elpida with much better headroom for someone that wishes to overclock. The limitations seen in the 290 Chip is more than the chip itself, but the memory used. IT WOULD SEEM that on the 290X's they would give it a significant edge with better memory.

http://www.guru3d.com/articles_pages/msi_radeon_r9_290x_lightning_review,30.html
https://www.dropbox.com/s/9kkxjuw2a0n421v/R9%20290X%20LIGHTNING%20Infokit%201_0%20-%20PDF%20Version.pdf

This may be the 'holy grail' of 290X boards. They are committing to anything other than Eldida, LOL! and that is the way it should be IMHO for 290X's

 

[Brantyn Gerik]

"significant edge" isn't the word i would use...if you look closely, the overclocking results from the card are, in reality, actually quite poor for the price you are paying, and i stress this statement about the price vs. performance . Considering the price point, the only way this card is really going to be worth it over say one of the cheaper 290x's or even a 290, is if it can actually hit over 1200mhz on the core or does extemely well under water and l2n (a lot of records have been set with the 290x's so it is very likely) (also we all know core clock isn't everything, but from comparisons, the 290x would need a core clock of that 1200mhz to compete with a 780 at 1256 in game applications, firestrike makes use of a few of the amd side's compute capabilities, but is not overly representative of gaming performance)

I truly am sorry to be such a pessimist towards the new cards from AMD, but the truth is that compared to the promised performance and promised features (what we expected the cards to be vs what they are) they have let me down personally. First it was ambiguous marketing (pre ordering without telling you the price) then misleading advertising (claiming to be titan killer when the titan wasn't even the competing card upon release) and near scam-like promises from mantle without releasing full details on the types of gains we could see from each type of system. Were it not for the mining craze, i fully believe amd would have been in quite a bad situation with these cards, just as they were with, don't quote me on this, but the new am3+ cpu's and how they promised huge gains, released the new motherboards 6 months early, and then completely let down the consumers.


I will admit a personal bias against them, even though i do my best to NOT include these biases in my response. I personally have been burned by the company in terms of promised performance vs actual delivery of that promise. On the other hand, i feel that this aggression towards amd has allowed me to further my research into the subject and find the true performance differences between the products rather than follow popular belief and marketing schemes. Half of my spare time is spent either reading reviews, or borrowing hardware from friends to test numbers of my own. Nothing i say is without backing