Best graphics cards: the GPUs I recommend for every budget in 2025

The most productive graphics card is objectively the RTX 4090 from Nvidia. Subjectively, however, you have to weigh the pros and cons of spending $1,600 or more on a single GPU. It is not for each and every one. That’s why we tested each and every GPU of this generation made by AMD, Intel, and Nvidia to find the best card in multiple price ranges.

For the high-end gamer, the most productive $600-$800 graphics card is the RTX 4070 Ti Super and just below that price, the most productive $500-$600 graphics card is the RTX 4070 Super. They’re fast and feature-rich, with better ray tracing, scaling, and symbol generation performance than the competition, but they’re quite expensive.

Although AMD’s highest-performance graphics card is the high-end RX 7900 XTX, its lower-spec models offer fair value for money. The most productive graphics card in the $350 to $500 range is the RX 7800. It’s all about saving as much as possible while still getting smart performance, the most affordable graphics card is AMD’s Radeon RX 7600.

It’s worth noting that the launch of the next generation of GPUs has already begun, with Intel launching its B-series of Arc graphics cards powered by Battlemage. In addition, Nvidia’s RTX 50 series GPUs and AMD’s RDNA 4 GPUs are also rapidly approaching; The first will be launched at the end of January and will last until February 2025. That said, whatever occupies the first in this consultant is still a selection for your next graphics card purchase, now and in the future. Below, I’ve indexed the maximum applicable GPUs you can buy today, all in order of gaming performance, so you can make the most informed selection.

Jacob has loads of experience with the latest and greatest graphics cards, reviewing many generations of Nvidia and AMD GPU over the years. He’s au fait with the latest architectures, and makes sure to rotate through the latest cards from all three major manufacturers to get first-hand experience of what they’re like to game with. Not just of their performance, but also which offer the most useful features and have the most reliable drivers.

The general GPU

The RTX 4090 is where it’s at. This is the graphics card that indeed represents a true next-generation experience, but it is also very expensive. However, if you are lately looking for the most productive graphics card in terms of performance, look no further.

Read more below

Our expert review:

✅ You need the best: the RTX 4090 is the toughest GPU you can buy today for your gaming PC. The internal silicon is incredibly resilient, and with DLSS 3 and Frame Generation, it delivers a true next-generation experience. ✅ You need to master 4K gaming: This is the card that makes 4K gaming exceptionally smooth. That 24GB frame buffer means you probably won’t run out of VRAM anytime soon. ✅ Or you’re an author and gamer: time is cash if you’re working with professional GPUs, and the RTX 4090 may start to pay for itself without delay given its incredible processing and computing power.

❌ You have to ask about the price: it’s fair to say it’s one of the cheaper Ada GPUs given its relative price, but it’s still $1,600 at best. That’s much less expensive than the RTX. 3090 Ti and RTX 3090 if you’re worried about inflation. ❌ You’ve been given a compact platform – this thing is BIG, comically cool. In some cases, you will find it difficult to adapt, so make sure you know that first.

? The RTX 4090 is the true next-gen delight that was never noticed on any of this new generation’s other AMD or Nvidia cards, and it’s almost worth that exorbitant price.

The most productive graphics card out there is Nvidia’s GeForce RTX 4090, and there’s nothing fancy about this monster of ultimate gaming functionality. It’s a huge pixel pusher, and while a few more curves are added to what might otherwise look like a Carenado RTX 3090 respin, it still has that novel graphics card aesthetic.

It looks like some semi-satirical plastic model made up to skewer GPU makers for the ever-increasing size of their cards. But it’s no model, and it’s no moon, this is the vanguard for the entire RTX 40-series GPU generation, a complete beast of a gaming component that leaves all others in the dust.

On the one hand, it was an advent of the kind of excessive functionality that Ada can offer when given a long leash, and on the other hand, it was a soft, tone-deaf release in light of a global economic crisis that has made launching a graphics card for a small minority of busy gamers feel a little off.

But we can’t forget about this for this consultant of the most productive GPUs just because, as it stands, no option for the RTX 4090 can come close to its performance. It’s unstoppable and will stay ahead of the curve given that AMD’s highest-performing graphics card, the RX 7900 XTX, is more of a competitor to the RTX 4080.

It’s worth noting that a replacement appears to be on its way, which we’re keeping track of here. That being said, the RTX 4090 is powerful enough that it’s still very likely to stand head and shoulders above most of the next generation cards, too.

This is a massive GPU that packs 170% more transistors than the incredibly giant GA102 chip that powers the RTX 3090 Ti. And, for the most part, it makes the last-gen Ampere flagship card look behind the scenes. And that’s before we even get into the equivalent combination of majesty and black magic that is DLSS 3 and all its scaling and frame-generating tricks, further elevating its functionality to the stratosphere.

Look, it’s fast, it’s okay. With everything enabled, and especially with DLSS 3 and Frame Generation joining the party, the RTX 4090 is monumentally faster than the RTX 3090 that preceded it. 3DMark Time Spy Extreme’s direct score is double that of the larger Ampere core, and before ray tracing or DLSS comes in, raw silicon also delivers twice the frame rate of 4K in Cyberpunk 2077.

There’s no denying that this is an ultra-niche, ultra-enthusiast card, and that at its best makes the RTX 4090 a little more than a benchmark for most of us PC gamers. Then we have to count the days until Ada descends to the worthy of us, mere mortals, which is unlikely until we see what the next generation of cards brings.

However, on its own, the RTX 4090 is a wonderful graphics card and will satisfy the functionality cravings of anyone who can one day justify spending $1,600 on a new GPU. And no wonder it will sell to those who can buy it, like no other. The GPU can get close to that right now.

Read our full review of the Nvidia GeForce RTX 4090.

Our opinion:

✅ You can find a card with an MSRP (or less): The price relief is the real highlight of the Super variant, so you really have to find one with its MSRP of $999 or less to get the best value/performance. You need smart performance in 4K ray tracing games: Ada is efficient and very smart when it comes to handling the rigors of ray tracing. Add DLSS 3 and Frame Generation to the mix and the RTX 4080 Super excels at delivering solid 4K frame rates.

❌ Ray tracing doesn’t mean anything to you: in terms of natural raster games, AMD’s RX 7900 XTX can outperform the RTX 4080 Super, making it a very tempting option, especially if you can find one that’s less expensive than the Nvidia card.

? The RTX 4080 Super is a much more tempting prospect given its $200 value relief over the popular RTX 4080. That brings it down to the same value as the AMD 7900 XTX and the balance of ray tracing features and functionality gets us back to Nvidia’s side. . Now that the RTX 4090 is ridiculously expensive, it has temporarily become the most productive GPU you can buy at a moderate price today.

Less is more, right? I mean, technically the RTX 4080 Super is really more, it’s less when you get a full AD103 GPU for less money. But what really stands out is the worthwhile replacement, as it’s the only tangible difference between this and the original RTX 4080.

While it’s not like you’ll ever have to compare the differences between the RTX 4080 and the RTX 4080 Super; the previous edition has been removed. The RTX 4080 Super sits atop the high-end (but not ultra-high-end) throne and remains as desirable a card today as it was when it was first released.

So, it’s all about weighing up how much of a difference that price cut has made to the positioning of the RTX 4080 Super. Let’s be honest, with a $999 MSRP, it’s still nobody’s idea of cheap, and arguably should still be cheaper than what we’re left with here.

But we can’t get too bent out of shape over what might have been; this is about the product I have in front of me, and the RTX 4080 Super is the same super-powerful graphics card it was when its predecessor launched. It’s a card which makes the once top GPU of the Ampere age look utterly laggardly—and incredibly inefficient—by comparison. At $999, it looks even better when put up against the $1,500 RTX 3090 of the previous gen.

The price cut and slight performance bump also now make it tougher for AMD’s best RDNA 3 card by comparison, though it is still very tight. Our two original reference RX 7900 XTX cards were beset by thermal issues and performed very badly, but third-party cards, and subsequent driver improvements, resulted in a Radeon GPU that was generally priced below the $1,200 RTX 4080 but in pure raster terms often outperformed it.

With a much more competitive MSRP of $999, the RTX 4080 Super rarely changes the game in terms of comparative functionality (on average, 2% slower at 4K settings), but it does make it less expensive than the speedy third-party OC RX 7900. Combine that with the weight of DLSS 3 and Frame Generation support, and that balances it out quite a bit for the RTX 4080 Super.

But it is still a close-run thing, and I certainly wouldn’t begrudge anyone picking the RX 7900 XTX over the latest GeForce GPU if they could find it for less.

However, after probably lackluster sales of the RTX 4080, and worse press surrounding its launch and price, the RTX 4080 Super turns out to be a fairly successful relaunch. It has a nicer fairing and the same functionality but at a lower price.

The RTX 4080 Super is a serious bit of gaming hardware, and if you’ve got serious cash for a majorly fast GPU and want to take advantage of Nvidia’s extra goodies, all without stretching to RTX 4090 prices, here’s where you probably want to be.

Read our full review of the Nvidia RTX 4080 Super.

Our opinion:

✅ You find a significant discount: At its $1,200 price or above it’s a poor deal, but if you can find the RTX 4080 below $999 it might be worth a punt for the serious gaming performance it offers, although the RTX 4080 Super variant should be hovering around that price level as well.

❌ You can find an RTX 4080 Super for less – the RTX 4080 Super is the cheaper card and given its incredibly similar performance, it’s a much better buy due to the price difference.

? It’s almost unlikely that the RTX 4080 will come with a value of more than $1,200. At $999 or less, it becomes much more tempting due to its impressive functionality and DLSS 3 support, although in reality, you deserve to be able to find the RTX 4080 Super for the same price. Oh, and it’s not even in production anymore, so you’ll probably only find it on pre-built PCs those days.

The Nvidia RTX 4080 is a fast graphics card, and when you think of DLSS 3, you get twice the functionality of the last-gen RTX 3080 Ti, for the price.

But reviewing the RTX 4080 is tougher than being Jen-Hsun’s spatula wrangler. For a start, it’s been pretty much entirely superseded by the RTX 4080 Super, a card that basically cuts the retail price of the original model while also delivering blink-and-you’ll-miss-it performance gains.

Seriously, we found the Super version to be just 1% faster than the OG at 1440p, and around 2% faster at 4K. That’s practically margin-of-error differences, but the fact that the RTX 4080 Super has a $200 cheaper MSRP means you shouldn’t really pick up the RTX 4080 standard unless you can find one cheaper than the Super variant.

However, if you do manage to find a smart deal, also note that the RTX 4080 handily beats similarly priced cards from the previous generation, adding the $1,200 RTX 3080 Ti and thus really beats this generation over generation. . the functionality improvement we were looking for when the card was launched. But honestly, none of those GeForce cards deserve to have had a $1,200 GPU.

Nvidia greatly reduced the silicon to create the AD103 GPU compared to the AD102 chip in the RTX 4090. Overall, it’s 60% the size, it has 60% of the transistors and 60% of the CUDA cores, and yet it had 75%. of the RTX value 4090. Si wished to do some undeniable calculations, the RTX 4080 deserves to have charge around $960, and if you can find one at this price compared to the Super, you’ll have seen a lot.

So how does the standard card compare to the Radeon RX 7900 XTX? We are looking for a very close case. The AMD card runs roughly the same as the RTX 3090, and for a $999 card, that would have made it tempting compared to the faster $1,200 RTX 4080, especially with its advanced ray tracing capabilities.

Now the RTX 4080 Super delivers roughly the same performance as the old card for a $999 price tag, I’d plump for that particular model over the AMD instead, mostly thanks to DLSS 3 and Frame Generation support.

The original is still in this guide as a reference, but once again, you should only be looking here if you can find a great deal—which will only be in pre-builts, these days, now the RTX 4080 is out of production. Otherwise, the RTX 4080 Super makes much more sense at its cheaper MSRP.

Read our full Nvidia GeForce RTX 4080 review.

Our opinion:

✅ You want the most productive AMD has to offer: This is ultimately the pinnacle of AMD’s RDNA 3 technology, and the fact that the red team has a GPU chiplet that performs so well in its first generation is truly amazing. I want many videos. Memory: With 24GB of GDDR6 at your disposal, for much less than the RTX 4090, RX 7900, the raster functionality of the Navi 31 GPU is excellent. It has improved its RT capabilities compared to AMD’s last generation, but it is still behind Nvidia at this point.

❌ The AMD reference card is the only option: Normally we’re big fans of both AMD and Nvidia’s reference cards, but the RX 7900 XTX had a heat issue with the reference cooler not found on third-party versions. We experienced the problem on both the review cards AMD provided for testing.❌ You’re looking for consistent RTX 4080 Super performance: We’d hoped for a more regularly competitive gaming experience from the RX 7900 XTX, but sometimes it’s a long way behind Nvidia’s second-tier card.

? As the most productive Radeon to date, the AMD RX 7900 XTX has a lot to offer. If it was consistently closer to the RTX 4080 in terms of gaming, we wouldn’t hesitate to feature this high-end GPU from the red team.

The most productive AMD graphics card of the moment is the Radeon RX 7900 XTX. We’re used to seeing generations of GPUs arrive at smaller compute nodes, redesigned architectures, larger caches, redesigned shaders, more memory, the list goes on. But all this, all at once? That’s what the RX 7900 XTX with RDNA 3 technology offers: all in one go.

At its original price of $999, the RX 7900 XTX was a very good 4K graphics card, but can now be found for around $850. AMD’s top-tier card from the last generation, the RX 6950 XT, is nowhere near its current full value ($1,099) and regularly goes on sale for $800 or less. The new RDNA 3 card has enough speed to justify its premium value in comparison, typically outperforming the previous RDNA 2 card by 20% and, in some games, much more.

That’s because the RX 7900 XTX has 20% more shaders, 8% higher boost clocks, twice as many FP32 units per shader, and 67% more memory bandwidth than the RX 6950 XT. The increase in memory capacity from 16 to 24 GB with the RX 7900 XTX is also a nice bonus, and the ray tracing performance on RDNA 3 is much more convincing to make me part with my money.

Still, as a competitor to the RTX 4080 or RTX 4080 Super, the RX 7900 XTX is less compelling. It can rarely fit any of the cards. The RTX 4080 is up to 28% faster in my tests, it’s more like 15% on average. For an RTX card that demands at least 20% more cash than the RX 7900

What helps the XTX’s price proposition is that it has more memory and on rare occasions outperforms the RTX 4080, Super or not. If you only play Far Cry 6, laugh with an XTX, but let’s be honest, that’s not the case.

Our review RX 7900 XTX sample also suffered from an issue with GPU hotspot temperatures exceeding the normal expected range under load. We reached out to AMD and received a replacement, but unfortunately had the same issue strike again. Fun, eh?

You can check out our reviews for the Asus TUF Gaming Radeon RX 7900 XTX OC Edition and Sapphire Nitro+ RX 7900 XTX Vapor-X, as these cards are entirely unaffected by the issue and better show what sort of performance you can expect from this card’s spec. The benchmark numbers above for the top AMD card are from the excellent Sapphire Nitro+ version.

All of this makes Nvidia the leader in the ultra-high-end segment. Make no mistake, the RX 7900 XTX is a wonderful 4K graphics card for an ultra-high-end PC built in 2024. However, the RTX 4080 Super may be next. There’s no doubt that this makes it a hard sell, but it’s still the most productive AMD GPU you can buy.

Combined with a high-end CPU and a 4K (or ultrawide) monitor, you get very good frame rates with the RX 7900 XTX in your build, and if you’re looking for the most productive AMD GPU performance, this is the card you need. .

Read our full AMD Radeon RX 7900 XTX.

Our opinion:

✅ Price helps keep dropping: At launch, its $899 sticker was too close to the incredible RX 7900 XTX, but now well below $700, it becomes a more tempting 20GB GPU option.

❌ Appreciate Nvidia’s DLSS 3 and Frame Generation technology: At around the same value as the RTX 4070 Ti and swapping shots at popular raster frame rates, the truly tangible benefits of the full DLSS 3 package can make a real difference.

? The AMD RX 7900 XT makes things a little more awkward for the RTX 4070 Ti and RX 7900

AMD’s Radeon RX 7900 XT is a slightly simplified edition of the Navi 31 GPU and the company’s most productive graphics card, the RX 7900 XTX. Starting at $899, and now available for around $700, so it offers a particularly less expensive way to get into the RDNA 3 generation, and you can also be forgiven for thinking it’s not much less expensive than the most productive.

So why would you pick up the cheaper RX 7900 XT? That’s a good question, and I’m not sure I have a good answer, apart from the price.

Overall, I’d say the RX 7900 XT does a few things well. For starters, it’s a smart upgrade, even on the RX 6950 XT, and considering the price difference between the two at launch, it’s a smart sign. of AMD’s progress with the RDNA 3 architecture. The reference cooler also performs well for its price, with relatively cool temperatures considering its performance.

Sometimes the difference between the XT and XTX is minimal and the functionality delta is almost non-existent. The XT is also the more effective and cooler style of the two. Overall, though, you get what you pay for with the high-end XTX card, or even a little more.

Is it bigger than an RX 6950 XT? Yes. Cheaper than an RX 6950 XT at launch? Yes. An RTX 4080 Super competitor? No. Does this style save you money instead of the XTX? Probably not.

It’s a smart 4K graphics card if you look at frame rates in isolation, but with a sometimes much larger card on hand, you’re better off going to need to find the extra $100-200 somewhere in my build and get the XTX or RTX 4080 Super instead.

Read our full AMD Radeon RX 7900 XT review.

Our opinion:

✅ You’re upgrading from an RTX-20 series card or an older card – Upgrading from one of the wonderful xx70 cards of yesteryear will give you spectacular Array upgrade functionality, not to mention the “free” upgrade offered by DLSS 3 and Frame Generation. will bring. deliver. ✅You need the security of having a card with 16 GB of VRAM: Forget the 16 GB RTX 4060 Ti. This is the Nvidia card to get now if you need a little more life. With the arrival of Grand Theft Auto 6, it would arguably be better to have too much VRAM in the coming years than not enough.

❌You already have a non-Super RTX 40-series card: The Super refreshes, while welcome, aren’t enough of an upgrade over non-Super counterparts to justify going out and dropping another wad of cash on one.❌ The RTX 4070 Super drops further in price: The RTX 4070 Super is a full $200 less than the RTX 4070 Ti Super. Should it drop to $549 or lower, it’d become hard to ignore.

? I would have hoped for more of a performance difference with the RTX 4070 Ti Super considering it’s using the bigger GPU. But even if it’s only 10-15% quicker than the RTX 4070 Ti, it’s still the card I’d want for the memory increase and bigger GPU on the whole.

The mid-life update to the RTX 4070 Ti Super addresses the considerations we had when the initial lineup went on sale and right now it’s the most productive graphics card at $600 to $800. With the RTX 4070 Ti Super, any worries about low-memory specs disappear. This is the RTX 4070 Ti we wanted, and it probably would have ended up as our initial generation pick, assuming it was still presented at $799.

At just over half the price of the formerly mighty RTX 3090, while handily beating it, the RTX 4070 Ti Super is the perfect advertisement for an intergenerational performance improvement. If it beats out the RTX 3090, just imagine the kind of upgrade it will deliver for owners of popular cards like the GTX 1070 and RTX 2070. Add to that the benefits of DLSS 3 and Frame Generation and gamers looking for an upgrade that’s a big step above the likes of the RTX 4060 Ti will be very happy.

The AMD RX 7900 XT is still a smart competitor, especially after its recent price drop. However, AMD’s rival is still lagging behind in ray tracing functionality and has now lost its VRAM advantage.

The day before our review was published, Nvidia informed us of a BIOS factor with the MSI Ventus card we tested. This factor can result in a loss of functionality of up to 5%. An updated BIOS was provided, but even that BIOS still suffered from a reported 3% loss of functionality. MSI has now released a vBIOS patch that would improve “the overall functionality of the graphics card to be in line with our expectations. “

But with stock sitting in warehouses, it’s a struggle to recommend the MSI RTX 4070 TI Super Ventus right now when users are likely going to have to be very wary of what vBIOS version is on their board and deal with the sometimes scary (though honestly super straightforward) trial of updating the BIOS of their new GPU.

However, there is a wide variety of RTX 4070 Ti Super cards, including factory overclocked models from other manufacturers, so this small factor is not worth putting you off purchasing which makes this a very capable card. BIOS problems are rare for many GPUs, it’s not the norm, and enough time has passed that the early quirks of third-party cards deserve to have been well and actually fixed.

Graphics card brands take great delight when it comes to taming cards with much higher thermal and power requirements, so in fact, any of the entry-level RTX 4070 Ti Supers will offer almost identical functionality. Dual-fan models may require a bit more fan speed and therefore noise levels to bring the watches to life, but ultimately, the aesthetics and loyalty of the logo (if any) will be the main differentiators.

All that aside, gaming at maximum 1080p and 1440p refresh rates is a breeze with the compatible RTX 4070 Ti Su. Add to that the benefits of seamless functionality consistent with watts, 16GB of cutting-edge VRAM, Nvidia’s AI and artistic tools, and the ability to game with a gorgeous view with DLSS 3 and Frame Generation in 4K and $799 suddenly turns out to be a moderate price. Get it for $749 or a little less and we’ll be really happy.

Read our full Nvidia RTX 4070 Ti Super review.

Our expert review:

✅ You can find one significantly cheaper than the RTX 4070 Ti Super: Yep, much like the RTX 4080 above, this card only really makes sense if you can find it significantly cheaper than the new variant, especially as the Super version has a little bit more performance. It’s still a brilliant GPU though, so worth keeping an eye out for a great price.

❌ $800 is too rich for your blood: If you’re considering selling a kidney, hold that thought. The RTX 4070 is considerably less expensive and is only around 25% slower. And remember 25% slower still means really high 1440p and even 4K frame rates.

? It’s a smart card, but like almost all Ada GPUs, the RTX 4070 Ti is still too high a value to feel like a true generational step. The RX 7900 XT is also almost the same value now and is rarely faster, and the RTX 4070 is rarely much slower for less. It is also very likely that it will only be found for sale in pre-built PCs, as it is no longer in production.

The unlaunching and subsequent rebadging and repricing of the RTX 4080 12 GB was the best thing to happen to this third-tier Ada GPU. Now and forever to be known as the RTX 4070 Ti, this is the card that made it impossible to recommend AMD’s RX 7900 XT.

However, just like the previous debate about RTX 4080 vs RTX 4080, the RTX 4070 Ti Super is there for the same MSRP. Not only that, but it also provides a moderate increase in functionality, meaning that you really only deserve the non-Super edition. if it can locate many, which will be increasingly unlikely given that the RTX 4070 Ti is no longer in production and will become increasingly rare.

Perhaps the most impressive thing that can be said about the RTX 4070 Ti is that it is consistently on par or faster than an RTX 3090. When you think that this is the $1,500 GPU of the last generation, it seems like a great generation evolution. to generation. performance, especially when using the higher 4K resolution.

What’s less exciting is that, when we talk in terms of directly raster games, it’s not much faster than the older, less expensive 10GB 4K RTX 3080. It’s faster, especially when you include the third-gen RT cores in the equation, but it’s clear that the higher clocks and heavier L2 cache have to work hard to give it the edge in frame rate. Raw images compared to the old Ampere card.

Where it is much more positive is in AMD’s RDNA 3 cards, RX 7900 XTX and RX 7900 XT. It is sometimes slower than the high-end Radeon GPU, but compared to the even more expensive RX 7900. XT, the RTX 4070 Ti shows superior 4K performance.

Without any of the DLSS 3/Frame Gen stuff in attendance, the RTX 4070 Ti is a very capable performer, but once again the power of Nvidia’s upscaling tech is preposterously good. I keep trying to see where the Frame Generation technology fails but I can’t do it. Every time I’m like ‘Aha, there it is, the tell-tale artefact of fake AI frames!’ I then check out the native rendering and it looks exactly the same. If not worse.

With the extra genuine performance of the upscaled frames and the interpolated smoothness of the AI-generated frames, the performance improvement is spectacular where DLSS 3 is available. This should be more and more often, with Nvidia’s Streamline SDK offering devs a one-stop option for enabling it and other vendors’ upscaling tech too.

In terms of gaming, the 4K functionality of the RTX 4070 Ti is impressive, even without upscaling, and it’s amazing.

Still, given that the RTX 4070 Ti Super easily beats it and can regularly be found for the same price, this is all pretty basic. It’s a wonderful card, but only if you can find it particularly less expensive than its newer, flashier sister. And this will most likely only apply to pre-built PCs, since the non-Super RTX 4070 Ti is no longer in production.

Read our full Nvidia GeForce RTX 4070 Ti review.

Our expert review:

✅ You need great mid-range functionality: The RTX 4070 Super has brilliant 1440p performance and, thanks to DLSS 3, is a wonderful card for building a mid-range gaming PC. ✅ You can’t have everything justified with an RTX 4070 Ti. Great: While it’s a big jump in functionality to move up the ladder, the value also increases significantly. The RTX 4070 Super, however, is still very attractively priced given the current state of the GPU market.

❌ RTX 4070 falls below $499: With a functionality delta of 18% at most, the RTX 4070 becomes a very attractive option if its value continues to fall beyond the $549 MSRP that Nvidia has tentatively set.

? The RTX 4070 Super is the Ada update that has seen the biggest improvement in terms of core specs, although it is the same GPU. But the fact that the original RTX 4070 still exists and the RX 7800 XT is so competitive makes the war more difficult.

The RTX 4070 Super is a reputable graphics card and we wouldn’t hesitate to present it as the most productive $500-$600 card available today.

This card is closer to the RTX 4070 Ti in terms of specs and GPU functionality than the RTX 4070, which is honestly the way we’d prefer. So while the original was an RTX 3080 with advantages, it’s more like a replacement for the RTX 3080 Ti and, in real terms, it particularly beats it, although it only employs almost a portion of the forceArray.

This makes it an excellent card for 1440p gaming and can even be located in 4K territory with the help of DLSS 3. We liked the RTX 4070 for its perfect mid-range performance, but the Super variant has around 10% in brain. benchmarks, and that’s a pretty healthy margin for a little more money, with an MSRP of $599.

Still, while the RTX 4070 Ti and RTX 4080 have necessarily been pushed out of the game via their Super replacements, the original RTX 4070 remains and, with a discounted MSRP of $549, it’s still a wonderful deal.

So why is the old map preserved? This is most likely because the RX 7800 XT offers similar functionality for $499 or less, which means Nvidia has to remain a convenient card in terms of value and the new RTX 4070 Super puts a bit of transparent air between them in terms of gaming frame rate. .

That said, the RTX 4070 Super is very well positioned for the functionality it offers and is well ahead of the festival in this segment. That cumulative 10% in functionality is big enough to justify its position here, and while the RTX 4070 is still a wonderful buy, the Super variant does enough for us to make a point of picking up the newer style.

Read our full Nvidia RTX 4070 Super review.

Our expert review:

✅ You don’t care as much about ray tracing performance – RT is here to stay, but enabling it still has a big effect on performance. Rasterization is important and the RX 7900 GRE does it very well. ✅ You need a silent card: The Sapphire Nitro edition is an exceptionally silent card. You can’t actually hear it in most typical operating scenarios.

❌ You care a lot about ray tracing and DLSS: There are more and more ray tracing games in fashion. While the RDNA 3 generation is a big step forward in this regard, the DLSS and ray tracing functionality of the RTX 4070 Super are still a step up.

? The RX 7900 GRE took a while to hit the global market, but now that it’s here, AMD has a formidable competitor on its hands. For $550, it’s not a reasonable card, but it’s arguably one of the most productive RDNA 3 cards for what it offers.

AMD’s Radeon RX 7900 GRE started as a China-exclusive release. GRE stands for Golden Rabbit Edition and refers to the Year of the Rabbit in Chinese culture. Initially released in mid-2023, it will be available on OEM systems in the general market for a few months before launching on the DIY market as a standalone graphics card in late February 2024.

RX 7900 GRE is essentially a trimmed-down RX 7900 XT, which itself is a trimmed-down RX 7900 XTX. It uses the same Navi 31 chiplet GPU as those two cards, but with fewer memory controller dies activated, fewer shaders, and a lower TGP. At 260 W, top clock speeds are a little lower too.

The result is a total of 16 GB of GDDR6 memory at 18 GT/s connected to a 256-bit bus, which puts the RX 7900 GRE in a smart position compared to the RTX 4070 Super with its 12 GB of memory and 192 – bit bus matrix

1440p is where the RX 7900 GRE shines. It’s capable of easily hitting 90fps and up on non-easy titles, but fails when easy ray tracing effects aren’t allowed. This can be offset by enabling AMD’s mature technologies, adding FSR 3 and Fluid Motion Frames. If you do, it becomes a perfectly moderate 4K option, at which point its 16GB frame buffer starts to become relevant.

Although it arrived on the market more than a year after its older brothers RX 7900, the RX 7900 GRE was above all one of the most attractive cards in the entire RX 7000 line. But as the value of the RX 7900

The RTX 4070 Super might be the first to be the more attractive overall option of the two, but the RX 7900 GRE with its 16GB of VRAM and seamless rasterization functionality means it holds up very well. The $500 to $600 graphics card market is still a very competitive place, and the RX 7900 GRE holds its own, if you can find it at the right price compared to its siblings.

Read our full AMD Radeon RX 7900 GRE review.

Our expert review:

✅ You’d like to save a little over the Super variant: While the RTX 4070 Super is 10% faster than the original card, if your budget’s tight, then the OG version here is usually cheaper.✅ You’re building a tiny gaming rig: The RTX 4070 is cool, quiet, and supremely efficient. It’s also a lot smaller than any other card of similar performance.

❌ You already own a decent RTX 30- or RX 6800-series GPU: At ~20% higher performance than the RTX 3070 Ti, and similar performance to the RX 6800 XT, spending another $550 so soon isn’t worth it just for Frame Generation.

? The RTX 4070 may be more expensive than the previous RTX 3070, but it can keep up with and even surpass the RTX 3080 when using DLSS 3.

The Nvidia RTX 4070 is a less expensive $100 RTX 3080. This is the simplest, but probably also the simplest, way to describe the Green Team graphics card. This is the fourth entry in the Ada GPU generation and, in terms of popular metrics, it performs as well as the fourth-tier card in the Ampere line. At first glance, it’s simply a less expensive chip.

But it’s not just that. The RTX 4070 is like an RTX 3080 with benefits.

This was the first Ada graphics card to use the same GPU as the previous version, with just a few smart elements trimmed to create a more affordable offering. And it also means that Nvidia was able to go all out with any chip that failed to match the point of the RTX 4070 Ti, which otherwise uses the full AD104 chip, or a mobile RTX 4080.

The RTX 4070 is rarely a monstrous mass of GPU with a PCIe connector; It’s a modest card the length of its RTX 3070 predecessor. This makes it a nice sight. Well, at least in terms of scale; This Nvidia Founders Edition brushed aluminum frame still looks pleasantly serious.

And that’s more than aesthetics, too. The size of the card hints at the efficiency of the 4nm Ada GPU quietly thrumming away inside of it. If you want a powerful, but low-power card, the RTX 4070 fits the bill.

In fact, this is one of the benefits I mentioned earlier in the RTX 4070 vs RTX 3080 debate, but more importantly the fact that the Ada card has access to DLSS 3 and frame generation. And when that comes into play, it’s a game-changer, especially for titles that would otherwise struggle with 4K ray-traced settings.

It’s a huge deal, though, and it comes in the form of the RTX 4070 Super, which now takes its position as our most sensible tip for cards in the $500 to $600 range. It costs $50 more but also offers 10% more in appearance, which puts it ahead of the original RTX 4070 in our tips. The RTX 4070 is still a brilliant card, make no mistake, but the Super puts it ahead in our estimates.

Read our full Nvidia GeForce RTX 4070 review.

Our opinion:

✅ You have a $500 budget: This is the best card in this price range right now, offering the standard gaming performance of two of the best GPUs from the last generation.✅ You want the security blanket of 16 GB VRAM: With a grand total memory capacity, and a 256-bit memory bus, the RX 7800 XT has a memory subsystem that will be ample for 1440p gaming for the foreseeable future.

❌ You need excellent ray tracing performance: AMD’s RT acceleration is still a generation behind Nvidia, and with more games supporting it, this can be a valid fear if you crave next-gen lighting effects. ❌ You were hoping to upgrade your RX 6800 series card: Despite the chiplet technology, the game has not evolved technologically from the same GPU point as the last generation Radeon.

? While AMD still has some catching up to do in terms of ray tracing performance, in many other ways the RX 7800 XT is a better value for your dollar card than anything Nvidia has to offer in the mid-range today . Although many will opt for an RTX 4070 or RTX 4070 Super for a little more money.

There would have been a time when we’d have baulked at calling a $499 graphics card “mid-range”, but that’s the world of PC gaming nowadays. If you’re prepared to drop that sort of cash on a graphical upgrade for your PC, at least you can buy the RX 7800 XT knowing it’s going to net you great performance, and that makes it our top recommendation for the best $350-$500 graphics card.

How much value do you put on ray tracing as a PC gamer? That’s got to be the question at the forefront of your mind if you’re considering dropping $500 on an AMD Radeon RX 7800 XT. Because if you are 100% sold on ray tracing you probably need to consider whether you’re actually willing to pay a bit more for the privilege of extra RT performance and drop the cash on the RTX 4070 Super instead.

If, however, you’re convinced rasterised performance is the only GPU metric worth a damn in this topsy-turvy world of PC gaming, then the RX 7800 XT is probably the best upper mid-range graphics card you can buy today.

As you’d expect, and it’s to be expected with a $500 graphics card, the RX 7800 XT is an excellent 1080p GPU. Only close to Cyberpunk 2077 does it not exceed 60 frames per hour with the most complex graphics settings. For the rest of our benchmark test, the card is capable of exceeding 100fps, and in fact, the maximum commonly approaches the 144fps you’d need for a maximum refresh rate. monitor.

What AMD needs you to know, however, is that this is a 1440p graphics card, designed to run more responsively at 60fps on newer titles at settings. And honestly, it goes way beyond that outside of, again, that brutal Cyberpunk 2077 outlier.

The RX 7800 XT has 4K gaming capabilities, as you’d expect from a card that offers functionality on par with AMD’s previous-generation high-end GPUs. At this level, you want to scale to take care of ray-traced lighting effects. However, once you move away from natural raster rendering, you can see that the RX 7800 XT starts painting with so many pixels on the screen.

It’s arguably an upper-mid-range card, but it arguably hits the point of the last generation’s two most productive GPUs, the RX 6800, roughly equivalent to the RTX 4070 in terms of popular games. – faster – with only games specializing in the latest ray-tracing effects, which gave Nvidia a fake win.

It all depends on the value of this card, and that’s where the RX 7800 XT wins. It’s by no means a reasonable card, as $500 is a lot to spend on any component, but it does reduce the value of that point of raster functionality and it’s an affordable way to run games at maximum at 1440p, giving you the ability to keep ray tracing to a minimum.

Read our full AMD Radeon RX 7800 XT review.

Our expert review:

✅You’re actually worried about the limitations of 8GB of VRAM. For now, 8GB of VRAM is enough, however, some games are already hitting those limits, and 8GB will only be more limiting, however, with 12GB on the 7700 XT, you literally save. You can’t find a little more budget ✅ for an RX 7800 XT. The RX 7800 XT costs only $50 more than the RX 7700 XT. If you don’t find those extra or RTX 4060 Ti-inspired greenbacks, then the RX 7700 XT is rarely a bad choice.

❌ Ray Tracing and DLSS 3 are a must. AMD would have possibly released FSR 3 and FSR 3. 1 in some games, but the technical credentials of Nvidia and its ever-growing developer can’t be overlooked. ❌ You can wait before upgrading. The RX 7700 XT is a smart graphics card, but it’s too expensive. Over time, its value will most likely drop, similar to what happened with the RX 6700 XT, which eventually became a fantastic card. for their money.

? We’d rather have the RX 7800 XT, but the RX 7700

The RX 7700 XT looks wonderful compared to its predecessor, the RX 6700 XT. It has more shaders, excellent ray tracing performance, faster memory, and includes AI accelerators that were notably missing on the RX 6000 series cards. AMD positions it as a competitor to the RTX 4060 Ti 8 GB, and it does the job well. task.

The RX 7700 XT has some features and specifications that the competing RTX 4060 Ti 8GB lacks. Lately, DisplayPort 2. 1 is more of a feature to check out, as the ultra-superior gaming solution with maximum refresh rates requires a high-end graphics card. However, the RX 7700 XT’s fantastic memory subsystem deserves praise.

One of our criticisms of the RTX 4060 Ti is its low-memory 8GB, 128-bit configuration. The RX 7700’s 12GB of VRAM and 192-bit bus

The RX 7700 XT’s challenge with Nvidia’s competition is that the RX 7800 XT costs only $50 more.

The RX 7700 XT is built around the Navi 32 chiplet GPU, just like the RX 7800 XT, with some features disabled. Both come with a single 5nm Graphics Computing Array (GCD) and four surrounding 6nm Memory Cache Arrays (MCDs). RX 7700 XT, on the other hand, has one of the MCDs disabled.

The RX 7800 XT is a better graphics card in every way, and for $50 more than the RX 7700

If you can’t stretch your budget, and I understand that not everyone can, the RX 7700 XT is still a tough card in max games at 1440p. It handles 60 FPS in all our proven titles, with the exception of Cyberpunk 2077. 4K. it’s still the domain of the most expensive cards, but if you’re betting on older titles, competitive shooters, or well-optimized MOBA games, the RX 7700 XT will run them in 4K without issue.

Read our full AMD Radeon RX 7700 XT review.

Our expert review:

✅ You can’t quite stretch to the RTX 4070: While the RTX 4070 is the better mid-range card, the RTX 4060 Ti puts on a decent showing for itself, especially if you want to take advantage of DLSS 3 without breaking the bank.✅ You’re rocking an RTX 20-series or RX 5000-series GPU: With a sub-20% boost over the RTX 3060 Ti, it’s not necessarily a particularly effective upgrade over most RTX 30-series GPUs, but it will be a significant upgrade to an older Nvidia or AMD card.

❌ You have a mid-range GPU from the latest generation: There is a huge improvement over the RTX 30 or RX 6000 series graphics cards when it comes to the mid-range of those previous GPU generations.

? The RTX 4060 Ti is the card you deserve to buy if you’re in the market for a new midrange Nvidia GPU and can’t expand it to the RTX 4070. You have to be a bit converted with ray tracing to need it, so badlyArray because compared to AMD’s mid-range it falls short. But it’s faster than the RTX 3060 Ti and RX 6700 XT, and it supports DLSS. 3.

The RTX 4060 Ti was, in many ways, the card that many gamers were hoping for in the Ada generation. With prices through the roof, the anticipation was that there might be a reasonably priced 40-series card with great performance that filled in the gaps, something with the sort of punchy performance that the RTX 3060 Ti provided in the previous generation.

The successor to the Ada-based RTX 3060 Ti will be the 40-series graphics card for the GPU-hungry masses. Surely it would be time to bring up the cause of a new RTX 40 series card.

Oh, how I wish I could tell you it was a resounding “yes”!But, while this is surely the card you buy if you’re ever considering an RTX 3060 Ti or an RX 6700 XT, all due to its complex gaming performance, I don’t feel right about making this recommendation.

The RTX 4060 Ti is a tightly regimented graphics card, rigidly taped out to hit a certain performance level as cheaply as possible and to provide Nvidia with an avenue for any AD106 GPUs that don’t make the grade as higher-spec mobile RTX 4070 chips.

It was designed to beat an RTX 3060 Ti with as narrow a margin of functionality as imaginable (and honestly, you can’t argue with the business reasons for doing so), but without DLSS 3 and Frame Generation, it doesn’t actually feel like a device with specific capacity. model. Exciting generational update to the old Ampere card.

And, to me, it doesn’t really feel like a slice of silicon that’s really worth the $400 sticker price, even if the performance on offer arguably is. That all makes it a very uneasy recommendation even though it is absolutely the fastest graphics card at this exact price point.

It’s also worth saying that we’ve seen decent price drops over time, particularly on the 8 GB variant, and for a good discount, the RTX 4060 Ti certainly becomes the GPU darling of the mid-range PC gamer. If you can pick one up around the $350 mark, you’ve got yourself a bit of a steal, but it’s still tough to truly recommend it much beyond that.

Read our full Nvidia GeForce RTX 4060 Ti 8GB.

Our opinion:

✅ Looking for a 1440p card at a great price: The RX 6700 XT was terribly priced at launch, but with cards available for $330 or less, it’s a wonderful gaming GPU today. It also has 12GB of VRAM, more than the more beloved RTX 4060 Ti, sometimes faster.

❌ The 7800 XT and 7700 ❌ Stock is almost out: As this is a three-year-old card, inventory has decreased and while it hasn’t affected the value much, trying to locate one is a challenge.

? When it comes to next-gen GPUs, the RX 6700 XT still has to hold its head high. It has a 12GB frame buffer and decent 1440p functionality for its value of less than $350. If you can find a. . .

If you can find one in stores, the best graphics card between $250 and $350 that we suggest you buy is the Radeon RX 6700 XT. However, as stock is almost out of stock, we are presenting it with caution as it is no longer available.

The AMD Radeon RX 6700 We are not talking about the cheapest chips here either. The Radeon RX 6700

There’s more to the Radeon RX 6700 XT than a simple halving of silicon from AMD’s top RDNA 2 chip, the Radeon RX 6900 XT. In some ways, sure, it’s a straight slice down the middle. The RX 6700 XT features 40 compute units (CUs) for a total of 2,560 RDNA 2 cores and is equipped with 64 ROPs—exactly half of the maximum configuration of the Navi 21 GPU—but the card comes with more than its fair share of memory and Infinity Cache.

A headline feature of AMD’s RDNA 2 lineup has been bigger memory capacities compared with rival RTX 30-series GeForce GPUs, and the RX 6700 XT doesn’t buck that trend. There’s 12 GB of GDDR6 packed onto this card: an attempt at what we optimistically call ‘future-proofing’. That’s greater VRAM capacity than the RTX 4060 Ti 8GB and RTX 3070 and is a match for the RTX 3060 12 GB.

With a price tag closer to the GeForce RTX 3070, yet performance between it and the GeForce RTX 3060 Ti, most often closer to the latter, the Radeon was a solid alternative but hadn’t been my first port of call for this sort of cash for a good portion of its life. That’s all changed now that it’s often quite a reasonable amount cheaper than anything the newer generation can manage.

If ray tracing doesn’t bother you and you’d prefer the extra memory over some of the other budget to mid-range offerings in this list, however, the RX 6700 XT is still a decent all-round GPU to buy right now. Should you be lucky enough to find one, of course.

Read our full AMD Radeon RX 6700 XT review.

Our opinion:

✅ Requires DLSS 3 and Nvidia Software: If you’re tied to Green Team’s software and scaling solutions, this is your only option at this type of value point.

✅ YOU WANT A COOL AND QUIET CARD: This card offers exceptional functionality consistent with the watts. Nvidia has solved this challenge with the Ada architecture.

❌ You have an RTX 30 series GPU: There is not enough progress with the RTX 4060 to justify a single generation upgrade. It’s only worth buying if you have to shoot on an older GPU.

? The RTX 4060 isn’t the kind of card that makes us scream about its performance, but it’s decent in its price range. We wish it were less expensive and more competitive, but it’s not. It’s just a worthwhile upgrade in moderation for anyone with a much older card that supports the Nvidia ecosystem.

Bringing up the rear of the RTX 40-series is the RTX 4060. Not quite the budget-friendly champion we’d perhaps want it to be, when there’s not a huge amount of current-gen competition at the $300 mark, it’s a bit of a shoe-in for this price bracket. That said, the RX 6700 XT still provides plenty of competition, despite being hard to find in stores, and the RX 7600 is the newer, cheaper, but still pretty competent alternative.

Nvidia has positioned the RTX 4060 as a 1080p card with just enough power and power for key Nvidia technologies, adding DLSS 3 with Frame Generation. In fact, it’s actually DLSS 3 that remains the flagship feature of the RTX 4060.

Perhaps the biggest question, controversy, or triviality, depending on how you look at it, is the inclusion of 8 GB of VRAM. As is the case with the RTX 4060 Ti 8 GB or the RX 7600 8 GB, there’s no doubt that 8 GB will eventually become a bottleneck, as it is increasingly becoming for the RTX 2060 6 GB and even more so for 4 GB cards. Given its predecessor, the RTX 3060, came with 12 GB and a 192-bit bus, the RTX 4060 certainly feels like a step backwards.

If you’re playing at 1080p, that’s fine for now, but will that be the case if millions of players download Grand Theft Auto 6 and find they can’t fully enjoy it? Future-proofing is a big word in the tech world, but 12GB of VRAM would have been welcome.

For now, the RTX 4060 has moderate performance. At $299, the RTX 4060 is rarely expected to work with a 4K 144Hz gaming monitor, but it surely deserves to be able to play all games in 1080p with smart frame rates on the top settings. Some games like Cyberpunk 2077 with full ray tracing are a challenge for much more expensive cards, even in 1080p, but with the magic of DLSS 3 and Frame Generation, even the modest RTX 4060 can put on a smart show.

The RTX 4060 is also a very energy efficient GPU. It is very suitable for players with smaller instances or instances with poor airflow as it emits a lot of heat.

While we’re not surprised by the RTX 4060’s lackluster improvement over previous cards in purely raster games with no DLSS to back it up, we also don’t like its small GPU – at least it’s a little less expensive than the 12GB RTX 3060 it replaces. . It is important to note that he is even still very competitive in certain matches. If you need to stick with Nvidia on a budget, there’s rarely a better option today.

Read our full Nvidia GeForce RTX 4060 review.

Our opinion:

✅ You know it runs the games you like: If you’ve seen others playing the games you want to play using the B580 without issues, then you can rest assured you’re not rolling the dice so much as making a great value purchase. And thankfully, that list of working games should be quite expansive and will only grow as drivers improve.✅ You want great raw performance for the price: Intel’s GPU ecosystem might not be as reliable and fleshed out as Nvidia’s and AMD’s—not yet, anyway—but when the B580 works, boy, does it work. In some cases it outshines the RTX 4060 by quite a big margin.

❌ You need absolutely reliable performance: While the Arc B580 offers a lot and its drivers are improving, it can’t be said to offer the point of reliability that the RTX 4060 or RX 7600 do. These cards will work in all cases, but unfortunately the same cannot be said for the B580.

? The Intel Arc B580 offers stellar performance for the price in the $250–$350 price bracket, but unfortunately, while its drivers are certainly improving, they’re not as reliable as those from Nvidia and AMD. In most cases you’ll get RTX 4060-beating value, but there’s a chance that in some small portion of cases you’ll get far less than this, or even find a game doesn’t work at all. It’s up to you whether you think that dice roll’s worth it.

The Intel Arc B580 brought the second generation of Intel Arc graphics cards in December 2024, and the GPU is aimed squarely at the RTX 4060 and its semi-budget competition in the $250 to $350 range. In some tactics it was successful, but without some pretty serious warnings at launch.

The BMG-G21 chip at the heart of the B580 is incredibly capable, and most of our tests obviously attest to this. It trades blows with the RTX 4060 and beats it hands down, even using frame generation (in Intel’s case, it’s using the new XeSS-FG or the GPU-independent AMD FSR equivalent). But “” is rarely always, and we found that it was far from being the best in all cases.

So far from perfect, in fact, that some games refused to work at all. In his testing, Dave found Cyberpunk 2077 to crash with the B580 when upscaling was disabled, thus the lack of frame rate data for that game in the charts above. And using the review drivers, Homeworld 3 wouldn’t run in DX12 mode, and when running in DX11 the performance isn’t impressive.

The problems weren’t limited to gaming, either, as some art apps like DaVinci Resolve also failed testing, and Blender’s functionality was nothing to write home about, despite its smart VRAM mapping.

However, these come from engine problems, like those that happened with the first generation of Arc GPUs. And that means there is room for improvement, perhaps even significant improvement. In fact, we are already seeing that Homeworld has resolved its problems and is now working well. For the maximum component things worked well, even with the evaluation drivers.

And “working well,” for the B580, means impressive functionality for the price. It outperforms the RTX 4060 by a wide margin, even at 1440p, and with the upscaling and symbol generation at 1080p, it’s on track to hit the max upgrade. Rate the games.

You also get the full package, this package is still in its infancy: decent ray tracing, upscaling, ray tracing, low latency mode (XeSS-LL), 190W TBP compared to the RTX 4060’s 115W and 12 GB of VRAM in comparison. to the RTX 4060’s measly 8GB. In fact, it’s a great price proposition at the lower end of the spectrum. In the $250 to $350 value range, you just have to be willing to roll the bucket a little on upcoming controllers and games.

Read our full Intel Arc B580 review.

Our opinion:

✅ You’re upgrading from a years-old card – This is a smart upgrade if your old GPU suffers from playable frame rates at 1080p in popular titles. ✅ You build a compact gaming rig: Thanks to its small size, the RX 7600 SKU is a great choice for a small gaming rig or living room PC. Its AV1 encoding and DisplayPort 2. 1 are also great advantages.

❌ You already own a mid-range RX 6000 series or RTX-30 GPU. The RX 7600 is very much a card aimed at gamers on a budget. It can’t compete with the latest generation mid/high tier cards and its 8GB buffer will be increasingly a bottleneck in the coming years.

? It wasn’t a smart addition to the Navi 33 GPU, however, the Radeon RX 7600 is still a faster GPU than the expensive RX 6600 XT it apparently replaces. The challenge is that Intel’s Arc A750 is rarely $200 and rarely faster.

After the release of AMD’s Radeon RX 7900 XT and RX 7900 Surprisingly, it was the RX 7600 that appeared on shelves next, and despite some drawbacks, it remains the most affordable GPU on the market today.

The RX 7600 is easily the cheapest of the current generation of GPUs from either AMD or Nvidia. Though it’s important not to overlook Intel’s Arc cards either. For now, the RX 7600 is a solid option if you’re after a card on a tight budget.

I was surprised at how small the Radeon RX 7600 reference card was. In fact, it’s small compared to some of the triple-slot giants we’ve seen. This alone will impress some buyers with compact cabinets looking to take advantage of its impressive demo features and media engine.

Then there’s the elephant in the room. AMD did miss a trick by sticking with 8 GB of VRAM. It’s enough for 1080p in almost every current game, but can we say that two years from now?

But that’s in the future. For now, the RX 7600 is a competent, though not stellar 1080p performer. It’s easily capable of high framerates in eSport titles and maintaining that important 60 fps level at 1080p in almost every game. It is still the dominant gaming resolution and the RX 7600 does offer a decent all-round upgrade over the aforementioned cards.

It’s got a good media engine that includes AV1 encode and decode support up to 8K, DisplayPort 2.1 and support for up to four simultaneous 4K displays at 144 Hz each. That makes it a very capable HTPC or productivity GPU. There’s nothing available (yet) at its $269 or cheaper price point that ticks all of those boxes.

Here’s a small caveat, though: if you approve with the AMD card, you may not be able to take advantage of DLSS 3 and frame generation as you would if you spent a little more on an RTX 4060. It’s a small problem. Demerit to the AMD card, but since it’s so cheap, it’s still the best budget recommendation.

While it’s not what you’d necessarily call impressive in terms of overall performance, once you think about that MSRP, it makes a lot of sense. It’s possibly small and relatively cheap, but if you’re looking to upgrade from an older GPU on a very tight budget, then this affordable little number is a pretty smart choice.

Read our full AMD Radeon RX 7600 review.

Our opinion:

✅ You should back the little guy: Well, billion-dollar Intel is rarely the consistent “little guy,” but in the GPU industry, it is. And its A750 has become a strangely effective first entry for the company, providing a true third option in the budget segment of the market. ✅ You only have $200 to spend: Let’s face it, we all have strict limits on our spending, and if you only have that much money, the Arc A750 will give you the most productive value for your money.

❌ You’re risk averse: The driver stack wasn’t great at launch, with rather inconsistent performance. That has changed in the intervening months, and it’s far more solid, potentially with more room to grow. But there are still questions over gaming performance and stability concerns, and those are likely to be around for a while yet.

? It’s not what we expected, but we actually introduced the Intel Arc A750 as a smart graphics card in 2024. This is mainly due to a massive price drop to $200, which comes and goes, but also some British pound sums. paints to the functionality of your drivers.

The Intel Arc A750 got off to a bit of a rough start, and infrequently, it still has a weird driver-induced moment to this day. But it offers just the right functionality for the price, as long as you don’t spend more than $200 on it. This card can be found for this type of cash those days, probably because it doesn’t seem to have sold very well.

The Arc A750 offers a moderately scaled-down edition of the Arc A770 GPU, the G10, which you can learn more about in my Intel Arc A770 review. It comes with 28 Xe cores, the building blocks of the Xe-HPG architecture, just 4 fewer than the Arc A770’s 32 Xe cores. So it’s also not far behind the speed of this card in terms of performance, but it runs slower and has some of the overall memory capacity.

But the Arc A750’s memory specs, for a card of its price, are immense. This is basically because Intel’s Arc A750 is more of a graphics card than its price. Intel invested in physically powerful specs for its first-generation GPUs, adding a large memory spec, and in the end, the drivers couldn’t get to where Intel wanted them to be. This means that some games don’t run well on the Arc A750, but it’s not too bad. This also means that the Arc A750 has abundant untapped potential.

A potential that Intel is gradually exploiting with every new driver update. Still, while progress has been made, it’s still fair to say that Intel is well off the pace of its rivals when it comes to keeping its drivers up to date, stable, and optimised.

Meanwhile, Intel and its partners (like ASRock) have reduced the price of the Arc A750 to compete with AMD’s RX 6600. The RX 6600 was my previous pick for this spot, but at the end of the day, I think Intel has more graphics cards to offer. Offer for cash than the competition. And there are still some elements to sweeten the matter.

The Arc A750 comes with impressive AV1 acceleration for encoding in the new, bandwidth-savvy codec. That’s a big deal if you’re a streamer or content creator looking to improve the quality of your videos. The A750’s ray tracing ability is similarly beefed up compared to the competition’s.

There will be cases where the A750 is way off the mark, and that pretty much rules this card out for anyone on an older system without Resize BAR support, but generally I think it’s a savvy buy for its new lower price and definitely a bit of a budget underdog right now. At a time when graphics cards are often underwhelming for the money and dreadfully prescriptive at the lower end, the Intel Arc positions itself as something of a wild card—but if you’re looking for an easy ride on a very tight budget, it might be better to go with something more tried and tested on the driver front.

Read our full Intel Arc A750 review.

Nvidia GeForce RTX 3090 Ti | March 2022″The RTX 3090 Ti is undoubtedly the fastest consumer GPU known to man, but it comes with a steep price and how long it remains on top is up in the air. This is as far as the Ampere GPUs can be pushed and, with the next-gen architectures from AMD and Nvidia on the way this year, its dominance could well be short lived.”PC Gamer score: 71%

Nvidia GeForce RTX 3090 | September 2020 “This frankly huge graphics card is incredibly powerful, but it’s more worthy of its Titan credentials than the GeForce brand. For the average gamer, it doesn’t perform well enough for the RTX 3080 to make sense, but for the professional. Creator is a card that crushes the workload.

PC Player Rating: 84%

Nvidia GeForce RTX 3080 Ti | June 2021 The GeForce RTX 3080 Ti is an excellent 4K gaming graphics card, very similar to Nvidia’s RTX 3090. However, with prices closer to those of Ampere, and well above the RTX 3080, this card will be discovered only on the most expensive PC versions on the market.

PC Gamer score: 84%

AMD RadeonRX 6950XT | May 2022 “It runs wonderfully, is quiet and incredibly powerful. Although it is much thirstier than previous RDNA 2 cards. However, the real problems are all external and arise from the expectation that the next-generation GPUs are missing some months for AMD and Nvidia. ”

PC Gamer Rating: 73%

AMD RadeonRX 6900XT | December 2020 “The AMD RX 6900

PC Gamer Rating: 66%

Nvidia GeForce RTX 3080 | September 2020″The new Nvidia card houses a monster of a GPU, tearing up the Turing generation and making ray traced gaming worthwhile. And this Founders Edition is the ultimate expression of the GeForce RTX 3080.”

PC Gamer score: 92%

AMD RadeonRX 6800XT | November 2020 “With the launch of Radeon RX 6800

PC Gamer Rating: 90%

AMD RadeonRX 6800 | November 2020 “The RX 6800 has no true Nvidia counterpart, however I can say that it more than matches an RTX 2080 Ti at 4K and offers incredible functionality at 1440p for gaming at the maximum refresh rate without compromise. “

PC Gamer Rating: 80%

Nvidia GeForce RTX 3070 Ti | June 2021 “The GeForce RTX 3070 Ti offers frustratingly variable performance, too close to an RTX 3070 for a price between it and the RTX 3080. It doesn’t have the same effect as the more productive Ampere gaming cards, but at least “At least it gives PC gamers another chance to get a graphics card at an MSRP this year. “

PC Player Rating: 78%

NVIDIA GeForce RTX 3070 | October 2020 “This small, third-tier Ampere GPU delivers gaming functionality that would have surpassed the Turing generation, and it does so while running at a cooler temperature, with less power, and in a much smaller space. “

PC Gamer Rating: 90%

Nvidia GeForce RTX 3060 Ti | December 2020 “The RTX 3060 Ti is precisely what we expected from a fourth-tier RTX 30 graphics card, but that’s not a mark against it. Nvidia’s Ampere cards offer an almost unprecedented leap in gaming functionality compared to previous generations, and the RTX 3060 Ti manages to offer the same on a smaller budget.

PC Gamer score: 90%

AMD RadeonRX 6750XT | May 2022 “Asus edition of the RX 6750 at a time when the craft market is reborn and the perception of an MSRP product is ever relevant. “

PC Gamer score: 66%

NVIDIA GeForce RTX 3060 12GB | February 2021 “A healthy improvement over the RTX 2060, but some of the surprising is gone for the less expensive Ampere. At least you can be sure that you won’t run out of memory, and that turns out to be the case. It’s going to be vital for gaming through 2021 and beyond, even if we’re not there yet. »

PC Player Rating: 84%

Intel Arc A770 | October 2022 “The limited edition Arc A770 is far from the best, but it offers very playable frame rates at 1440p and with ray tracing enabled. It’s not a bad start for Intel Arc. But while it focuses on Nvidia and the popular RTX 3060, It’s AMD’s RX 6600 XT that offers the most options at the moment. “

PC Player Rating: 69%

AMD Radeon RX 6600 XT | August 2021″In an alternate universe, without the current silicon drought, AMD’s RX 6600 XT is less than $300 and a great mid-range option. In this, the darkest timeline, it’s almost $400, negligibly quicker than an RX 5700 XT and only slightly cheaper. Though at least it’s better than the RTX 3060.”

PC Gamer Rating: 67%

AMD RadeonRX 6600 | October 2021 “The Radeon RX 6600 offers enough for modern 1080p gaming, which is a smart thing considering AMD has upgraded this package to a much more power-efficient design than the RX 6600 Nvidia’s 12GB is still the card we’d recommend at this price if you can find one in stock. Otherwise, the RX 6600 is it. viable, otherwise.

PC Gamer Rating: 65%

NVIDIA GeForce RTX 3050 | January 2022 “Basically, we’re looking at Nvidia cutting off the ‘GTX’ prefix and giving us an RTX 1660 Ti. This makes it a smart 1080p GPU, the addition of DLSS help is a much more tempting proposition, especially at this level. . de GPU, which promises 1080p ray-traced gaming. We’re keeping our hands crossed that it remains in stock at a price. moderate.

PC Gamer score: 77%

AMD RadeonRX 6500XT | January 2022 “A graphics card designed to offer the bare minimum of hardware fails to advance the game and looks like the most cynical GPU failing graphics card ever. »

PC Player Rating: 47%

Each new generation of GPUs offers new features and possibilities. But raster rendering is still the most important metric for overall gaming functionality in the PC gaming world. Sure, Nvidia’s GPUs could do better in ray tracing tests that have more or less incentive. , however, when it comes to popular gaming features, AMD’s newest lineup can still hold up, at least until it hits the higher end.

It’s also worth noting that the past generation of graphics cards still has something to offer, with something like the GTX 1650 Super to outperform a more modern RTX 3050 in the peak benchmarks.

We’re not saying you deserve to buy an older card (cheap features from Intel or AMD are much more important these days), but it’s worth knowing where your current GPU sits, or just knowing the lay of the land. But there’s also the fact that there will be gaming rigs on sale with older graphics cards, and if they’re reasonable enough, they’ll possibly still be worth it as a reasonable entry into PC gaming. That’s a pretty big caveat though, so be sure to compare everything you have. you are looking with an older card for economical features with newer silicon.

We’ve benchmarked all the latest GPUs of this generation, and have tracked their performance against the previous generation in terms of 3DMark Time Spy Extreme scores. Where we don’t have the referential numbers for an older card we have used the average index score from the UL database. These figures track alongside an aggregated 1440p frame rate score from across our suite of benchmarks.

Except for Intel’s Arc GPUs. We haven’t included them in this graph as the A770 and A750 3DMark results skew higher than their actual gaming performance and would be misleading. The two Intel GPUs generally post frame rate numbers that put them around either the RTX 3060 or RTX 3060 Ti at a pinch.

Here’s a list of the manufacturer set retail prices (MSRP), or recommended retail price (RRP), for most of the latest graphics cards. Those in bold font are for the current generation of GPUs. For the most part, these are the set prices for the stock or reference versions of these cards, if applicable, and not representative of overclocked or third-party graphics cards, which may well be priced higher.

NVIDIA

amd

1. Everything within your budget

2. Buy the graphics card not the brand

3. Uh. . . yeah, it also depends on your instructor.

4. You can wait for a new generation, but don’t automatically skip the latest generation cards.

5. It’s a simple update

In the United States:

Amazon – savings on Nvidia & AMD graphics cards

Best Buy – the only place to buy Founders Edition cards in the US

Walmart – Rare discounts on older or budget GPUs

Newegg – save on the list price of graphics cards

In the UK:

Amazon UK: Deals on the newest GPUs

Scan – on Nvidia graphics cards

Box: save on GPU

Ebuyer – often AMD cards with discounts

Overclockers – on newest generation AMD and Nvidia GPUs

Currys: Some on GeForce GPUs

Laptops Direct: Some GPU deals, but you have to look for them. . .

The older GTX prefix is now used to denote older Nvidia graphics cards which don’t have the extra AI and ray tracing silicon that the RTX-level cards do.

This RTX prefix was brought with the Turing-based RTX 20 series and highlights cards with GPUs that feature the Tensor Cores and RT Cores necessary for real-time ray tracing and deep learning supersampling (DLSS).

These days you only find older 16-series GPUs with the GTX prefix attached, so it’s pretty much RTX in every way.

The RTX prefix is only used to indicate cards that house Nvidia GPUs with compromised ray tracing hardware, but they are the only ones.

AMD’s RDNA 2 GPUs and RDNA 3 GPUs all support real-time ray tracing acceleration, as do Intel’s Alchemist graphics cards.

Keep in mind that with cheaper offers, you’ll expect maximum frame rates when enabled. Otherwise, AMD and Intel’s ray tracing acceleration is pretty clever but slower than Nvidia’s.

If you’re looking for maximum performance, you used to run two cards in SLI or CrossFire. However, it’s not uncommon for large games to completely forget about multi-GPU users. This includes all DXR games. There is also the fact that there are fewer and fewer fashion cards that link two cards together.

Then no. It’s one thing.

The obvious answer is: Only if you have a 4K gaming monitor. But there are other things to consider here, such as what kinds of games do you play? If frame rates are absolutely king for you, and you’re into ultra-competitive shooters, then you want to be aiming for super high fps figures. And, right now, you’re better placed to do that at either 1440p or 1080p resolutions.

That said, the more games incorporate upscaling technologies like DLSS, FSR, and XeSS, the closer the cards will get to reaching 4K visuals on your 4K monitor, but at higher frame rates.

Founders Edition cards are just Nvidia’s internal designs for their graphics cards, as opposed to those designed through their partners. These are reference cards, meaning they work based on inventory clocks.

Briefly, for the RTX 20 series, Nvidia will offer Founders Edition with factory overclocks. This had made it a bit tricky to compare cards, as the Founders Edition cards give us a baseline for performance, but Nvidia has since rebuilt them for reference.

Intel also offers something similar with its Limited Edition Arc Alchemist cards featuring its own in-house cooler design, as does AMD with its reference cards.

For basic 1080p games, the minimum amount of VRAM (video RAM) that any graphics card has is 4GB. However, you’ll notice that all of our recommended cards are 8GB or larger, and that’s because the vast majority of today’s cards are available. Budget 3D games have graphics that can consume more than 6GB.

Once you’re gaming at 1440p or higher, and using high graphics settings, 6GB just won’t cut the mustard and even 8 GB might be a bit limiting in some games. This is why the latest mid-range graphics cards, between $350 and $600 in price, sport 12 or 16 GB of VRAM.

The latter is more than enough for almost every game. You’ll want it in a small number of games and even then, in very expressive circumstances, like 4K with max settings and frame generation enabled.

However, if you use your GPU to create content or AI workloads, the more VRAM you have, the greater the graphics card’s ability to handle those tasks.

Jacob received his first byline for his own technology blog. From there, he turned pro as a Hardware Editor at PCGamesN and later led the team as a Hardware Editor. He joined PC Gamer’s controller team as a senior hardware editor before becoming chief editor of the hardware team. Now you’ll find him reporting on the latest developments in the generation and gaming industries and testing the newest PC components.

If AMD isn’t spilling the beans on the RX 9070 and 9070 XT at least online retailers are, but a release isn’t imminent

Turns out there’s ‘a big supercomputer at Nvidia… running 24/7, 365 days a year improving DLSS. And it’s been doing that for six years’

Climbing a lighthouse where they were determined to kill me in Peaks of Yore was truly eye-opening; Now I understand climbers.

PC Gamer is from Future plc, a major foreign media organization and virtual publisher. Visit our corporate site.

Leave a Comment Cancel Reply