Intel heard your screams of anguish, PC players. Finances graphics playing cards which can be truly price your cash have all however disappeared this pandemic/crypto/AI-crazed decade, with trendy “budget” GPUs going for $300 or extra, whereas concurrently being nerfed by substandard reminiscence configurations that restrict your gaming to 1080p decision until you make some critical visible sacrifices.
No extra.
In the present day, Intel introduced the $249 Arc B580 graphics card (launching December 13) and $219 Arc B570 (January 16), constructed utilizing the corporate’s next-gen “Battlemage” GPU structure. The Arc B580 not solely comes with sufficient firepower to finest Nvidia’s GeForce RTX 4060 in uncooked body charges, it has a 12GB reminiscence system target-built for 1440p gaming – one thing the 8GB RTX 4060 sorely lacks regardless of costing extra.
Intel
As if that wasn’t an interesting sufficient mixture (did I point out this factor is $249?!), Intel is upping the ante with XeSS 2, a more moderen model of its AI super-resolution know-how that provides Nvidia DLSS 3-like body technology for even extra efficiency, in addition to Xe Low Latency (XeLL), a characteristic that may drastically scale back latency in supported video games.
Add all of it up and Intel’s Arc B580 appears poised to essentially, really shake issues up for PC players on a finances – one thing we haven’t seen in years and years. When you’re nonetheless rocking an OG GTX 1060, take a critical have a look at this improve. Let’s dig in.
Meet Battlemage and the Arc B580
Intel’s debut “Alchemist” Arc GPUs launched in late 2022, rife with all of the bugs and points you’d anticipate from the primary technology of a product as advanced as trendy graphics playing cards. Intel diligently ironed these out over the following months, delivering driver updates that supercharged efficiency and squashed bugs at a torrid tempo.
In a briefing with press, Intel Fellow Tom Petersen mentioned a significant drive throughout Battlemage’s growth was enhancing software program effectivity, to be higher in a position to unleash the total energy of Intel’s {hardware}. However keep in mind, it ran on first-gen {hardware}, too. Battlemage improves effectivity on that entrance, utilizing methods like reworking the vector engines from two slices right into a single construction, supporting native SIMD16 directions, and beefing up the capabilities of the Xe core’s ray tracing and XMX AI directions to, sure, make all the pieces run smoother and higher than earlier than.
Battlemage’s {hardware} effectivity enchancment are illustrated right here, exhibiting how a Fortnite body runs on the brand new Arc B580 versus last-gen’s (costlier) A750 — the time to render a body dropped from 19ms to a silky clean 13ms
intel
I’ve included a bunch of technical slides above, so nerds can decide by the small print. However right here’s the upshot: The Arc B580 delivers 70 % extra efficiency per Xe core than final gen’s Arc A750, and 50 % extra efficiency per watt, per Intel.
Cue Keanu Reeves: Whoa. That’s completely bonkers. You virtually by no means see efficiency leaps that substantial from a single-generation advance anymore!
Intel
That’s at an architectural degree; the slide above reveals the particular {hardware} configurations discovered within the Intel B580 and B570. A few issues stand out right here, initially the reminiscence configuration.
Nvidia and AMD’s present $300 gaming choices include simply 8GB of VRAM, tied to a paltry 128-bit bus that each one however forces you to play at 1080p decision. The Arc B580 comes with an ample 12GB of quick GDDR6 reminiscence over a wider 192-bit bus – so sure, this GPU is actually constructed for 1440p gaming, in contrast to its rivals. The Arc B570 cuts issues down a bit to hit its $219 price ticket however the identical broad strokes apply.
Additionally price noting: Intel’s new GPUs characteristic a lavatory customary 8-pin energy connector (although third-party fashions might add a second one to assist Battlemage’s ample overclocking chops). No fumbling with fugly 12VHPWR connectors right here.
Intel’s homebrew Restricted Version reference GPUs will return for the B580 in a more moderen, smaller design with blow-through cooling. You’ll additionally have the ability to decide up third-party customized playing cards from the companions proven above, and the B570’s launch in January will likely be unique to customized boards, with no Restricted Version reference deliberate.
Intel
As a part of the launch, Intel can be introducing a redesigned gaming app with superior overclocking capabilities, together with the flexibility to tweak voltage and frequency offsets.
Intel Arc B580 efficiency particulars
Now let’s dig into precise efficiency, utilizing Intel’s equipped numbers.
Intel says the $249 Arc B580 performs video games a mean of 25 % sooner than final technology’s higher-tier $279 Arc A770 throughout a check suite of 40 video games. In comparison with the competitors, Intel says the Arc B580 runs a mean of 10 % sooner than Nvidia’s RTX 4060 – although crucially, these numbers have been taken at 1440p decision moderately than the 1080p decision the overly nerfed RTX 4060 works finest at.
Intel
Intel additionally made a degree of stressing how the RTX 4060’s restricted 8GB of RAM over a 128-bit bus can straight affect efficiency at the moment. The slide above reveals Forza Motorsport working at 1440p decision. At customary Excessive settings, the RTX 4060 truly holds a efficiency benefit. As you scale up the stressors, flipping on ray tracing and transferring to Extremely settings, the benefit immediately flips, with the B580 taking a transparent, substantial lead whereas the RTX 4060 hits the bounds of what’s attainable with its reminiscence setup.
Talking of, Intel says a lot of the key applied sciences underlying ray tracing have been improved by 1.5x to 2x in Battlemage in comparison with the first-gen Arc “Alchemist” choices. Contemplating that Intel’s debut Arc playing cards already went toe-to-toe with Nvidia’s vaunted RTX 40-series ray tracing, there could possibly be a fierce battle brewing in reasonable real-time lighting subsequent 12 months – which isn’t one thing I’d thought I’d say within the $250 section earlier than even flipping the calendar to 2025. When you’re nonetheless rocking a GTX 1060 or 1650 from again within the day, the Arc B580 could be a huge improve in each velocity and superior options like ray tracing.
Uncooked {hardware} firepower alone is simply a part of the graphics equation nowadays, nonetheless. Nvidia’s RTX know-how pressured the facility of AI upscaling and body technology into consideration this decade – and Intel’s new software program options are designed to supercharge body charges and decrease latency even additional.
Meet XeSS 2 and Xe Low Latency
Intel’s XeSS know-how debuted alongside the first-gen Arc playing cards, serving as an AI upscaling rival to Nvidia’s core DLSS know-how. (These render frames at a decrease decision internally, then use AI to supersample the ultimate consequence, resulting in increased efficiency with little to no loss in visible high quality.) However then Nvidia launched DLSS 3, a know-how that injects AI-generated “interpolated” frames between each GPU-rendered body, completely turbocharging efficiency in lots of video games and eventualities.
XeSS 2 is Intel’s response to that. Whereas DLSS 3 requires using a {hardware} Optical Movement Accelerator solely current in RTX 40-series GPUs, Intel’s XeSS 2 makes use of AI and Arc’s XMX engines to do the work as an alternative – that means it’ll additionally work on previous-gen Arc playing cards, and the Xe-based built-in graphics present in Intel’s Lunar Lake laptops.
Intel
And as we see with DLSS 3, the efficiency enhancements could be excellent. Intel says that in its in-house F1 24 checks with the B580, activating XeSS 2 with supersampling and body technology can enhance efficiency by a whopping 2.8x to three.9x, relying on the High quality setting used. Whereas the sport runs at 48fps on the chosen settings with out XeSS 2 enabled, turning on XeSS 2’s Extremely high quality lifts that each one the way in which as much as 186fps – a literal sport changer.
Intel
Assist for XeSS 2 is coming to the video games proven above, with extra to reach within the coming months. First-gen XeSS hit 150 video games up to now, so the hope is that XeSS 2 (which makes use of completely different APIs for builders to hook into) ramps rapidly as effectively.
Injecting AI frames between custom frames has a facet impact although – it will increase latency, or the response between your mouse click on and the motion occurring onscreen, as a result of the interpolated AI frames can’t reply to your instructions. Enter Intel’s Xe Low Latency characteristic.
XeLL primarily cuts out a bunch of the ‘middleman’ rendering and logic queues that occur behind the scenes in a body, letting your GPU render a body a lot, a lot sooner than typical. (Nvidia’s superior Reflex know-how works equally.) Activating it drastically lowers latency. You’ll be able to tangibly really feel the advance in video games that don’t have body technology lively, however by enabling it alongside XeSS 2, it claws again the latency created by body technology.
You’ll be able to witness the enhancements attainable within the slide under, which reveals the efficiency of an F1 24 body with a wide range of XeSS options (supersampling, body gen, XeLL) lively. It actually illustrates the necessity for a latency-reduction characteristic alongside body technology.
Intel
Latency discount is so essential to border gen “feeling right” that Intel requires builders to incorporate XeLL as a part of the broader XeSS 2 bundle, following in Nvidia’s footsteps. As with DLSS 3 and Reflex, you might even see the choices offered individually in some video games, whereas others will silently allow them collectively – it’s as much as the developer.
Battlemage brings the warmth?
At all times take vendor numbers with a giant punch of salt. We’ve seen vendor benchmark controversies through the years, together with this 12 months. Company advertising exists to promote stuff to you initially. Hashtag: Look forward to benchmarks et cetera et cetera.
All that mentioned, whereas Battlemage doesn’t push for the bleeding fringe of efficiency, I’m wildly excited by what I see on paper right here. Finances GPUs have been an absolute quagmire ever for the reason that pandemic, with none of the present Nvidia or AMD choices being very compelling. They really feel like rip-offs.
Intel
Intel’s Arc B580 and B570 really feel like real worth choices, lastly giving players with out deep pockets an attractive 1440p choice that’s truly reasonably priced – one thing we haven’t seen this decade regardless of 1440p gaming turning into the brand new norm. Delivering better-than-4060 efficiency and 12GB of VRAM for $250 is downright killer if Intel hits all its guarantees, particularly paired with what seems to be a considerable enhance to Arc’s already-good ray tracing efficiency. And with XeSS 2 and XeLL, Intel is conserving tempo with Nvidia’s superior options – assuming builders embrace it as wholeheartedly as they’ve finished with first-gen XeSS.
Add all of it up and I’m excited for a really mainstream GPU for the primary time in a lengthy time. The proof is within the pudding (once more, await impartial benchmarks!) however Intel appears to be brewing up one thing spicy certainly with Battlemage and the Arc A580.
