ASUS Radeon RX 5700 XT Strix Review

Wrong results

Once again, in battlefield 5, your results for the 5700xt reference and vega 7 are totaly wrong.
 
Once again, in battlefield 5, your results for the 5700xt reference and vega 7 are totaly wrong.

You say they are wrong, yet give nothing as to why they are wrong and/or don’t give any info/source as to backup your claim.
 
To be fair I don't think he needs to provide a source when the graph says Vega64 outperforms the RVII, but it's almost certainly down to the fact that trying to actually reliably benchmark BFV is almost impossible.
 
£500 is too expensive IMO. Great performing card, but I personally won't be getting this one.

Also, I've never found OC3D's GPU benchmarks overly reliable. They don't retest things and so results can appear skewed. I remember the 1070Ti was raved about here saying it was faster than a 1080, when the reality was, the 1080 had been tested over a year earlier and had many driver revisions since then. I would never take OC3D's numbers at face value. I'd always compare them to other sites.
 
£500 is too expensive IMO. Great performing card, but I personally won't be getting this one.

Also, I've never found OC3D's GPU benchmarks overly reliable. They don't retest things and so results can appear skewed. I remember the 1070Ti was raved about here saying it was faster than a 1080, when the reality was, the 1080 had been tested over a year earlier and had many driver revisions since then. I would never take OC3D's numbers at face value. I'd always compare them to other sites.

We are hoping to implement some changes in the future to address this. The only problem is that this change will require a CPU/motherboard change in our test system, so retesting and changing our graphs would be pointless before this happens.

Your criticisms have been noted. It is just going to take time to change things.
 
To be fair I don't think he needs to provide a source when the graph says Vega64 outperforms the RVII, but it's almost certainly down to the fact that trying to actually reliably benchmark BFV is almost impossible.


Yeah A Vega 64 should not be outperforming a Radeon 7 in anything considering the Radeon 7 is the successor to Vega 64 and faster in every way but then again as you said benchmarking a game that has no built in benchmark is quite tricky.
 
Last edited:
Good review as usual Tom, although I think the bench charts should have been sorted by 1440 as that is where the card really sits, it is an entry level at best for 4k.
 
To give more feedback, personally, I prefer the more YouTube style of doing things. Where a lot of reviewers will bench a set of cards in the same price or performance range and put them against each other. You can make an easy video off of that and get a pretty quick article up as well that goes into more depth of any issues or outliers, really anything, that isn't straight performance.

This let's you have more frequently updated results and therefore more accurate. You would have to make new graphs and sort by test date/driver release. But multiple smaller more readable graphs imo are far more user friendly for everybody including mobile users which this site is not entirely up to par on for 2019 and could slightly improve the SEO rating.

To add on before I forget, I feel like this more popular method of doing things currently is because they are straight to the point and quickly gets information across. Since less and less people are watching longer videos in general (not saying OC3D videos just people as a whole) it is something to consider.

Although I think you're trademark case reviews should definitely stay long. Those are literally gold and phenomenal. Way better than the 5-10 minutes everyone else gives a case review.
 
Last edited:
To give more feedback, personally, I prefer the more YouTube style of doing things. Where a lot of reviewers will bench a set of cards in the same price or performance range and put them against each other. You can make an easy video off of that and get a pretty quick article up as well that goes into more depth of any issues or outliers, really anything, that isn't straight performance.

This let's you have more frequently updated results and therefore more accurate. You would have to make new graphs and sort by test date/driver release. But multiple smaller more readable graphs imo are far more user friendly for everybody including mobile users which this site is not entirely up to par on for 2019 and could slightly improve the SEO rating.

To add on before I forget, I feel like this more popular method of doing things currently is because they are straight to the point and quickly gets information across. Since less and less people are watching longer videos in general (not saying OC3D videos just people as a whole) it is something to consider.

Although I think you're trademark case reviews should definitely stay long. Those are literally gold and phenomenal. Way better than the 5-10 minutes everyone else gives a case review.

I was hoping for a you tube one for these cards and agree the case ones are 2nd to none. I love detailed videos more than quickies though
 
too expensive for what it offers.


waiting for intel to step into the game so we can have at least some competition.


having just amd and nvidia fixing prices is not good.
 
50% more performance for a given price in less than 12 months is a pretty good step up, not sure what more you're expecting nowadays, unless you mean this Strix model specifically.

Intel should be competitive sub £300 but expecting them to come out the gate with competetive high end gaming cards is unrealistic, these GPUs are their first 7nm products so won't be too big, and any high end focus will be going to winning back all the datacentres, supercomputer & AI wins that NVidia & AMD have taken from them over the past 5 years, that's the real reason they're now making GPUs, the biggest market in computing, AI (Hence why they hired Raja Koduri). They'll make multi-purpose high end cards I'm sure but given they'll definitely be HBM by their own comments I wouldn't expect them to be too competetive on price that high up the stack unless they resort to their old trick of selling things at a loss to capture marketshare.
 
Last edited:
To be fair I don't think he needs to provide a source when the graph says Vega64 outperforms the RVII, but it's almost certainly down to the fact that trying to actually reliably benchmark BFV is almost impossible.

Agreed, there is no built in benchmark, the game gets updated regularly, drivers change, windows updates, bios updates, what mood an AV is in ect ect there are far too many factors to take into account here.

I am first to call out when something is blatantly wrong or biased but without benching everyone of these cards everytime a new review is published you can't call it wrong. It was correct at the time of testing dude, theres more to life than benching systems.
 
Back
Top