Gadgetory

the AMD rx 560 launches today with two gigabyte versions priced at around $100 and four gigabyte versions around 115 we're reviewing the Sapphire rx 560 pulse Oh see today a 4 gigabyte variant at $115 the gtx 1050 is the most immediate competition and previously out ranked the RX 460 and the majority of our tests with similar pricing the RX 560 returns in an attempt to change that hierarchy continuing and these at line of polaris refresh cards flanked by the 1050 and 1050 Ti with the rx 570 at the higher end before that this content is brought to you by custom backplate makers at v1 techcom do you want AK builds GP back place to order with their online customization tool making it easy to theme upcoming PC builds back plates cost $25 end up and are installed via magnets and can be seen in some shots of the cards reviewing that lately used code gamers Nexus 5 for five dollars off your order or click the link below this review will be pretty quick and easy by our standards anyway the 560 isn't anything crazy new it's so Polaris architecture just like the 580 was it's a refresh of an existing line of and the architecture from last year and with the 560 the main change here rather than just going with frequency and voltage changes like we saw the 580 over the 480 or 572 the 470 what we have here is a fully unlocked 560 so the 460 previously had a some locked cores that had to see use that were unavailable to the user so it was 14 rather than the full 16 on the chip this one goes for the full 16 s he used which means you're not only getting a frequency difference depending on which card you buy or look at but also getting a core count difference so the differences here in performance are more than just frequency changes basically we do have some chord changes as well as for this particular card the post OC is a $115 model which makes it one of the sort of borderline limit cards where you start bumping into territory where other cards might make more sense and the lower end to gigabyte models around $100 directly competing with the gtx 950 where previously the RS 460 was competing this model is pretty straightforward about half of the PCB doesn't even really long about a third of the PCB doesn't even really need to be there it's basically blank and what you end up with is a very tiny GPU die it's one of these smallest set AMD is shipping right now with a couple of VRAM modules and a three-phase VRM for the vrm and we as you can see by the thermocouples did test thermals on the B around the vram the vrm is using magnet chip semiconductor fats so they're using MDU 1514 and mvu 1517 fats they're not amazing it's a three-phase it's pretty basic but this isn't really an overclocking card anyway so a pretty basic card very cheap and it's just targeting an ultra budget build and we're going to test it today versus the GTX 1050 primarily 1050 T I look at the previous rx 460 a couple of our tests have a bit of everything in there so you can see the full stack up as this compares to things like a 475 70 and up for full testing methodology as always check the link in the description below that's the complete review written review we've got 3d mark tests in there that we won't have here but otherwise most of stuff will be here in in addition to the normal testing methodology one quick note some of these cards somewhat annoyingly have OC and their name which means that in the charts when we list OC it means either it's actually overclocked by us or the card is named OC an example would be the MSI GTX 1050 OC card which is not really overclocked it's one of the cheapest ten fifties that is out there we're testing it because it's cheap it's a hundred dollars normally $80.00 after rebates right now MSI just tags the letters OC on it so we're going to put quotes around OC if it's part of the card name as opposed to actually overclocked just a quick note starting off with total power consumption as measured at the wall our RX 5 60 pulse configurations are 6 to 10 watts on idle with the overclocked Varian's mostly equal to this this is a lower power consumption than the rx 460 for two reasons one and the tune to the 500 series of Polaris cards to draw lower power one Idol this is really just a BIOS update and to the RX 460 we test it anyway has LEDs and extra hardware on the board like another fan which increases power draw for comparison our MSI GTX 1050 model which MSI calls the OC unit but we're running stock as the equal power consumption in our test to the rx 550 pulse moving to 3dmark firestrike the RX 5 60 pulse faces a total system power job 148 watts lower than the RX 460 nitro only because the 460 hosts two fans LEDs and different volt frequency curves otherwise as we saw in our RS 580 versus our X 480 results which we can highlight here the 500 series has generally been drawing more power given the same PCB in cooler this one is a bit different though regardless 148 watts for the pulse 4 gigabyte card jumping to 172 an overclock the GTX 1050 configuration draws 139 watch with the MSI card and the 1060 SSE for reference sits at around 219 watts total system power draw in real world gaming workloads starting with 4 honors the RX 5 60 pulse runs a power consumption of around 170 to watch total system power draw the RX 460 nitro is at around 180 watts and both fans are cranking in this case but 560 pulls overclocked card the hovers around 193 watch total draw below the gtx 960 SSC at 218 watts total system power draw the msi 1050 card hovers at 153 watts total system power job marking it about 11 percent reduction in power consumption from the pulse 560 playing Ghost Recon wildlands we observe a total system power consumption of around 170 to watch for the RX 560 pulse that's a reduction about 5.7 percent from the particular rx 460 that we tested with the overclocked pulse build to draw in 194 watts this is about an 8% reduction from the GTX 960 SOC at the low end the 1050 MSI card runs at 152 I've settle system power draw moving now it's a thermal test and we start with a simple frequency versus time chart to plot GPU core stability under a power bi scenario clocks are enumerated differently in these torture workloads than in gaming workloads meaning the clock is lower than in gaming of power consumption and thermals are higher given the low-power cooler on the Sapphire 560 pulse we expected wears cooling performance than what was output in our torture workload it seems that thanks to the relatively low power consumption of the GPU compared to its larger counterparts things are kept under control relatively even with the heatsink that is similar to an Intel stock cooler we're looking at an average GPIO to reading of approximately 75 Celsius put the target seeming to be about 73 to 75 C GPU clock maintains just below 1200 megahertz and we see some clock dithering at this point ideally to observe a perfectly straight line for the clock but our range is still about 20 megahertz which really isn't bad to maintain this clock and temperature the fan ramps up maximally to around 36% fan speed for 1600 rpm on this particular cooler the next question is how the vrm and vram components hold up to the small cooler fortunately because this card isn't driving much power the vrm is actually do okay with our thermocouple testing we measured one of the hottest MOSFETs at around 63 Celsius which is well within any reasonable operating temperature and we're so far below the 125 C threshold that that temperatures are completely a non-issue here at least for the stock card the top left the vram module is one of the hottest vram modules on the card and measures about 75 Celsius or still well with inspect here that would get a bit warm in a higher ambient environment with restricted airflow TV is io temperature remains steady at around 73 to 75 Celsius and we'll move on to gaming next but for 3dmark results check the article into the description below starting our game testing with for honor the Sapphire rx 560 pulse OC operates at 43 FPS average with this in-game test course with GM's coined 1% low frame rate and 0.1% lows we see thirty two point seven fps and twenty eight foot seven fps this position sorry 560 poles below the GTX entity from MSI which despite its name is not actually overclocked and above the rx for 64 gigabyte card we've got a lead of about 5.7 percent over the previous RS 464 gateway card running with the slower clock rate and fewer cores and the GT extensively produces and a lot point six percent lead over the RX five six people's OC to compare versus other nearby cards the $140 msi gtx 750ti runs out to the seven FPS average 49 fps 1% lism 47 FPS 0-100 in for some blows this place is the 10 50 TI approximately 19% ahead of the $100 GTI 1050 and 32% ahead of the rx 560 RX 570 at $180 around 35% faster than the GTX 1050 TI while those cards aren't the focus of today they help us understand if and when the extra money could be worth while frame time consistency isn't cane for the rx 560 in this particular title as we can show on the frame time plot on the screen now although the GTX 1050 does reasonably well when compared to the 560 moving on to Mass Effect Andromeda we start with higher settings past to get some comparative numbers for the higher end cards and we'll move on to tests with medium settings the $115 Sapphire rs5 60 pulse runs an average FPS of 37 with loads of 31 and 30 trailed by the rx 460 nitro 4 gigabyte card which runs 834 FPS average the 560 weeds by about 8 point 8 percent over the 460 next card up the msi 10 50 100 dollar card operates an average of 46 FPS with lows tightly times as on the previous cards the 1050 Ti runs a 51 FPS average out matched quickly by an Rx 570 at 63 FPS average let's look at a medium result although more limited as we've only just begun testing lower end GPUs we get a better idea as to relative performance on these cards with me setting the GTI 1050 is able to hit 61 FPS average with the RX 560 stock running 53 average 45 FPS 1 percent lower than 42 0.1% lows this landed about 6.6 percent better than the rx 460 nitro 4 Giga card with the overclocked 560 pulse running at 8.8 present faster than the stock rx 560 moving now to doom with Vulcan and async compute the RS 560 finally begins to pull ahead of the gtx 950 in fact once overclocked it also pulls ahead of the more expensive 1050 Ti this is in line with what we've seen out of this particular game in the past or doom tends to produce more favorable results for AMD with its Vulcan integration and the msi 10:50 OC struggles a bit maintaining an average FPS of 44 with lows dipping to 25 and overall produces less consistent fame times than the RX 564 q a card the 560 a vertically took up PS which is about 6.5 percent faster than the rx for 64 gigabyte nitro card overclock the 560 moves up to 74 FPS average which is a 13% increase over its stock performance irreverence the rx 574 gigabyte card averages at around 120 FPS or 60% faster than the rx 560 postcard a quick note here this is what doom configured to ultra graphics at 1080p not bad performance for these low-end devices even a 1050 and although we wouldn't really recommend it for this game we now move to Ghost Recon wildlands where we've got tests at very high for comparison and then medium for more realistic representations of the cards we've got a fairly complete view of performance from top to bottom with this chart with regard to scaling the RX 5 60 pulse operates an average FPS of 33.3 followed by the RX 4 64 gigabyte card at 32 FPS average the GTX 1050 is able to sustain a higher average FPS at 36 with the 1050 TI running about 42 average for perspective the RX 570 runs an average FPS of 51 leading the rx 560 by 52% it's not until the RS 5 75 80 and DK 1060 territory that we start reaching numbers where it be possible to play on a mix of very high and high settings the gtx 1060 is the real Victor of that stack but at more than two times the cost of the 1050 or the 560 it can't really be considered as an alternative in the price point instead of trying to make that work let's run the test with lower settings moving down to 1080p medium hears medium where you'd more reasonably be playing on these $100 class cards we've got fewer cards on this chart but enough to start building an understanding of performance the RX 560 pulse now runs an average of 43 FPS about 10 FPS higher than previously landing it around 7.5% faster than the rx for 64 gigabyte nitro card overclocking gets us to 46 FPS average which is about on par with the MSI 1050 card stock performance that was IG tier 10 2d stock card operating point 6 percent faster than the RX 5 64 gigabyte card for reference the GTX hunted the GI stock card runs a the three FPS average makes it 12.8% faster than the 1050 and 15.3% fastened of the RX 560 sorry 570 is about 50% faster than the RX 560 for reference and finally we've got Sniper Elite 4 which is another game where we've only just started running at low end card tests so it's not as populated as our 4k test charts looking at 1080p high the RX 5 64 gigabyte card runs an average of 52 FPS with the overclocking producing a 9% performance improvement over baseline the RX 460 runs a 49 FPS average making the gap between the 460 and 560 about 6.7 percent the GTR 1050 is led by all of these links to its 43 FPS average the overclocking nearly gets its averages on par with the 460 despite lower performance and frame time consistency overall as illustrated by GM's low values these performance numbers plot the 560 close to 1050 Ti with the RS 570 hitting 100 FPS averages at times just for reference so it looks better than the RS 460 did at launch with the 460 especially the 4 gigabyte pricing for the units that we received initially it was hard to justify the purchase first of the 1050 because the 1050 was really just either performed equally or better in really all cases that we tested at a time and it was about a $10 price difference at worst when the 1050 was honored $10 and that was cheaper $200 sometimes you get them with 80 if you consider a rebate as part of the price but these cards the 2 gigabyte models are at $100 we're planning to do a 2 V for test 4 gigabytes bit more this one is 115 from what we've been told so it's really not too bad in the testing here today the 560 makes up some ground where the 460 was losing finally the rx 560 is competitive with the GTX 1050 in about half the title be tested and by competitive it's either a head or tied and on the other half it's a bit behind the 1050 still but when the 560 is ahead it's ahead by a good deal so that's the real difference here with Dumon Sniper Elite 4 being the main titles here the seminal examples of low level API is being done correctly the RX 560 takes the lead with the DX 11 titles we tested the gtx 950 takes the lead or is pretty close to it anyway so what you need to do for these cards is look at the type of games you're playing figure out what API they're on apparently and from there you should be able to figure out which one's worth the buy more realistically what you do is you look at the price and you see if there's more than a fifteen dollar difference between them at which point it's time to just take based on price for the most part because they're not hugely different in performance once you get above $120 for any of these cards ten 55 60 or whatever they start quickly losing value the 1050 TI is still in that really weird position it was when we first reviewed it where it's an okay performer but it sits between really strong performers at the low end which would now be the five six pm at ten fifty and pretty good performance at the mid-range which would be VR x4 70 or now the five seventy with of course ten sixties and 40s after that so the ten fifty TI is still in a weird spot it's worth buying for those really specific use cases where you have one hundred forty dollars and you can't possibly budget an extra twenty to thirty for a better card like the five seventy at this point or you just you have ten more dollars then you can spend on something like a five sixty or ten fifty that's the only time the tentative DGI makes sense it's always been a weirdly positioned card as for the low end though ten visi still does really well it is getting older at this point by standards in the GPU market so that means it's harder to recommend the five sixteen does just fine it's not terrible it's not impressive it's just fine so these are not the most exciting cards to look for reviews for but if you are buying them hopefully this helps you make a decision we've got the numbers for you you can check the article below for more of those numbers if this was not enough as always at patreon.com slash gamers next to tell us that directly can get a store doc gamers access not net to grab one of our shirts like this one thank you for watching subscribe for more i'll see you all next time you