Get quick — the list of thebest graphics cardsis go to look a lot different in the next couple of months . For the first prison term , Nvidia , AMD , and Intel are ready to found new generations within weeks of each other . Whatever you know about the three major players is about to modify . Not only are we cause raw generations but there are also shifts in strategy between Nvidia and AMD , tariffs to grapple with , and next - gen AI boast like FSR 4 in the pipeline .

Over the next few calendar month , everything we currently jazz about the current slate of GPUs will vary — that much I can say for sure . I ’m here to not only get you up to speed on the past 12 months of leaks , rumour , report , and confirmations , but also distill all of that information to get a better grasp on the GPU market of 2025 . A circle is changing — some undecomposed and some tough — but one matter is undeniable : We ’re standing on the boundary of an exciting new epoch for PC play .

The easy one: Nvidia

I have continue using weasel Holy Scripture like “ could ” and “ may ” to describe Nvidia ’s plans for itsRTX 50 - seriesGPUs because the companionship has n’t even confirm they exist . But if there are any GPUs we love the most about , it ’s the next generation from the marketplace leader . There is a mountain of passing water , even more rumors , and brands likeCorsair support the next generationis arrive before long , and Nvidia ’s CEO Jensen Huang isset to take the stage at CES 2025 .

Nvidia usually gives some special address at CES each twelvemonth . However , it has never launch an totally Modern propagation of artwork cards at CES . Nvidia usually plunge fresh generations at its annual — and sometimes semiannual — GPU Technology Conference , or GTC . The spring GTC result , which is held annually , ordinarily focuses on Nvidia ’s robotics , AI , and enterprise projects , while the periodic fall GTC is earmark for young gaming graphics cards . That state , earlier this year , we saw Nvidia set in motion new gaming GPUs like the RTX 4080 Super at CES , not a GTC event . GTC has or else force more toward AI , perhaps encouraging Nvidia to move its gambling launching to a broader technical school upshot like CES .

That ’s conjecture on my part , but the fact that Nvidia is host the chief CES keynote , and that it usually has a much smaller mien at the show , is not meditation . After all , Nvidia has alreadydeployed its Blackwell architecturein the data centre , which is the computer architecture it will use in its next - gen GPUs , andNvidia has reportedly lay off productionon several RTX 40 - serial models , presumptively to make way for next - generation graphics cards . That ’s something even Nvidiahinted at during its latest earnings call .

Jacob Roach in a promotional image for ReSpec

So , no , Nvidia has n’t confirm RTX 50 - serial GPUs are launching in January . But I ’d be shock if we did n’t see them amount ahead of time next yr .

That ’s just what I ’ve seen about when these lineup are come out , and there ’s even more to dive into on the spectacles front . It look like Nvidia is tilt into the moral force we saw with theRTX 4090and RTX 4080 this generation . The RTX 5080 will apparently have less than half of the sum of the RTX 5090 , placing an even spacious interruption between the two GPUs than what we see this coevals . It makes sense . For as potent as the RTX 4090 is in game , the reality is that an RTX 4080 can do just about everything on the gaming front . The RTX 4090 is that Titan - stratum alternate , branching out to professional workloads rather of just centre on gaming alone , and it looks like the RTX 5090 will behave that forward .

Something that should show up across the lineup is newGDDR7 retention . We ’ve cognize about GDDR7 for over year , and in recent calendar month , we’ve seen broader productionof the mental faculty get afoot . There are some leak pointing to Nvidia using next - gen memory , but those really are n’t what ’s authoritative . Nvidia has been on the cutting edge of graphics remembering for the last several generations . That fact , combine with the rollout of mass production — I ca n’t imagine Nvidiawon’tuse GDDR7 .

I ’m not going to theorize on performance here because , candidly , it ’s up in the air . Nvidia is uncontested at the top - last — something that will be even more straight in the next coevals , at least fromwhat AMD has say — so it ’s really a thing of far Nvidia want to bear on things ( and how much money client are willing to spend ) .

GPU pricingis a unlike story . It ’s no secret that Nvidia has sharply pushed the damage of GPUs up in the last two generation , and I do n’t see that changing with RTX 50 - series option . Not only is there conjecture that Nvidia will push prices up but we ’re alsodealing with likely tariffson China and Mexico — two major importation countries of personal computer computer hardware — that could push prices up across the board .

The wildcard: AMD

AMD is in an interesting topographic point right now . The company has already corroborate that itsRDNA 4 graphics visiting card — the RX 8000 serial — will go far in “ early 2025 . ” That ’s a inverted comma from the company ’s CEO Dr. Lisa Su , and it ’s a time anatomy that AMD has reiterated for several calendar month . That does n’t corroborate that the card will show up at CES 2025 alongside Nvidia , but that ’s the time frame I ’m keeping an eye on .

The release date is the least interesting part , though . AMD ’s position is well tot up by a quote shared by the company ’s own Jack Hugh earlier this twelvemonth . The executive aver thatAMD does n’t planto play “ king of the J. J. Hill ” with Nvidia . Rumors have been swirling for more than a year thatAMD wo n’t point the uttermost high endwith RDNA 4 graphics calling card , and Hugh ’s quote all but substantiate that those rumors had virtue .

It piddle mother wit for AMD to make this shimmy . The company has been chasing Nvidia ’s pate for more than a decade . Sure , there ’s been plenty of conversation between if AMD or Nvidia is better lower down the flock , especially with card like theRX 7800 XT , but at the flagship horizontal surface ? Nvidia has dominated for a long time . Team Green has essentially squashed AMD ’s flagship dream the past two generations . My guess — and this is just a guess — is that AMD does n’t see a bring home the bacon conflict there , and it believably lose some money chasing the crown .

Some context is important here to see why . With RDNA 2 nontextual matter cards , AMD finally had a flagship on - par with Nvidia in the form of the RX 6900 XT . brusque of ray trace power , it could stand up toe - to - toe with Nvidia ’s RTX 3090 for a much lower price . Then , Nvidia released the RTX 3090 Ti . This generation , Nvidia did n’t even have to free a Ti model of the RTX 4090 . It just made the flagship super muscular , super expensive , and relegatedAMD ’s RX 7900 XTXto a struggle with the RTX 4080 .

As I ’ve write aboutpreviously here on ReSpec , the flagship conversation sets the whole step for a new generation of graphics board . These halo products , generally , do n’t sell in gamy numbers , at least equate to more pretty priced options like theRTX 4060orRX 7600 . What doughnut products do is establish some wider narrative about a particular propagation and where the players stand . Absent a flagship contender , AMD can shift that conversation lower down the mountain as it has done in the past times with popular GPUs like the RX 5700 XT and RX 580 .

AMD can likely make up some ground for those essay higher performance with FSR . Like in GPUs , AMD has been tag Nvidia ’s DLSS with its own upscaling and frame multiplication tech . AMD says the next variation of FSR will be AI - free-base , and I suspect that next - gen RDNA 4 GPUs will be jam some sort of computer hardware to dulcify the AI deal . There ’s some discussion that we ’ll demand to have around AMD ’s change in approach shot to FSR and limiting package feature to special hardware , but that ’s a chromatography column entry for a different daytime .

The shoe-in: Intel

I ’ve care in a lot of rumors and speculation up this point , but now let ’s move to something concrete : Intel . The company has dropped a bunch of official details about its upcomingBattlemage GPUsin the many years that it ’s been in the works . No , the launch date is n’t confirmed , but a handful of passing water have hit over the preceding week , suggesting we ’ll learn about these graphics wit in a subject of days , not weeks . Somereports even indicatewe’ll get an announcement as soon as next workweek , but Intel has n’t confirm that timeline .

Intel released its Battlemage tech originally this twelvemonth — the Xe2 architecture that Battlemage GPUs will use launch in Lunar Lake laptop CPUs . Just this past week , the companionship posted a serial of videos detailing the architectures in spite of appearance of Lunar Lake , where Intel ’s Tom Petersen confirmed that “ Battlemage is next , ” suggest a reveal is veracious around the corner . This all joins months of supposition that Intel will launch its GPUs before of AMD and Nvidia , likely to carve out a culture medium hertz off from the challenger . For as telling as GPUs like theArc A770 and A750were , Intel has such a low-pitched market place shareright now that it does n’t even file against AMD and Nvidia . It needs some space by from the handsome musician to stand out .

As if that was n’t enough , ASRock spoiled Intel ’s party a bit and listed two computer graphic card prematurely on Amazon this past hebdomad . Two role model for the Arc B580 went up . The listing are no longer live , and there were n’t any pricing item , but a few other key pieces of information were revealed . First , the low - remnant example — from ASRock ’s Challenger serial — sported a single 8 - pin force connector , intimate the GPU wo n’t describe more than 150W. They also revealed the card will make out with 12 GB of VRAM and a PCIe 5.0 x8 user interface .

These details suggest the B580 is just one of the models Intel plans on launching , and in all probability the lower - end model . The A770 had duple 8 - pin tumbler power connection and 16 GB of VRAM . Meanwhile , theArc A580came with only 8 GB of VRAM . The PCIe 5.0 x8 port is interesting . It ’s not an issue if you have a PCIe 5.0 motherboard , but it could represent bandwidth upshot if used in a PCIe 4.0 motherboard one-armed bandit . That ’s something I ’ll need to turn over into once the card is here , however .

As for higher - close models , Intel has establish that it ’s point AMD , not Nvidia , to swipe away some marketplace share . sharply focussed on value , it looks like the flagship card from the range will direct the operation of the RTX 4070 . It ’s hard to say that for sure , though . I ’ve seen slews of leak over the past six months occupy Battlemage GPUs , all with wildly different public presentation . That suggests we ’ve meet a circumstances of early engineering run , which are n’t representative of final carrying out .

Still , there ’s a good chance Intel is n’t holding out for flagship performance , or even performance that ’s remotely closemouthed to what Nvidia could accomplish with the RTX 5090 . The strategy for Intel , it seems , is to undersell last - gen options while AMD and Nvidia are focused higher up the tidy sum . That ’s a scheme Intel put to good use with its first generation of desktop GPUs .

A trio of contenders

Between leaks , rumors , and official confirmation , one matter is clear — PC gamers are about to geta tonof new computer hardware to opt from . AMD and Nvidia normally set up within a similar sentence frame , but now Intel is in the premix . meld with AMD ’s shifting scheme , tariffs that could push Price upwards , and Intel ’s ambition in the screen background GPU market , it ’s an exciting clock time in the world of PC gaming — for better and tough .

Looking out toward the next generation , I suspect we ’ll see a rejection of crazily expensive flagship like the RTX 4090 . There was already peck of pushback against Nvidia ’s pricing scheme in the former multiplication , so much so that we get aggressive cost cuts with theRTX 4080 Superas well as a rebranding with the RTX 4070 Ti at a scurvy cost . If we see aggressive tariffs , that ’ll force the Mary Leontyne Price of these in high spirits - ending GPUs even further , and push them out of budget for most microcomputer gamers .

The pricing billet is only compounded by Intel and AMD ’s midrange scheme for the upcoming generation . It ’s alone at the top , and it looks like AMD , especially , is abandoning Nvidia at the top in the upcoming generation . Most of the conversation concerning value already happens lower down the stack , but I suspect there will be an even not bad focusing on midrange GPUs this generation chip in the contest from Intel and AMD .

On the other helping hand , there ’s some dependable opportunity for last - gen GPUs in the upcoming generation . This generation was vitiate by remnants of the GPU deficit , so last - gen alternative like the RTX 3070 did n’t see large drops in price — prices just dropped back down to number price . With GPUs like the RTX 4060 already seeing sale below list price , hopefully we ’ll see some secure great deal on last - gen option . They could really be the winner of the next multiplication look on where pricing lands .