The AMD Ryzen 9 7950X3D Review: AMD's Fastest Gaming Processor

https://ift.tt/NhPlakI

This week is the long-awaited launch of AMD's second generation of V-Cache equipped consumer chips, the Ryzen 7000X3D family. Aimed primarily at gamers, tomorrow morning AMD will be releasing a pair of their latest-generation Ryzen 7000 chips with the extra cache stacked on, including the Ryzen 9 7950X3D (16C/32T) and the Ryzen 7 7900X3D (12C/24T). Both chips build upon their Ryzen 7000X-series predecessors by adding a further 64MB of L3 cache, bringing them to an impressive total of 128 MB of L3 cache.

Meanwhile, a third SKU, the AMD Ryzen 7 7800X3D, is in the works for April 6th. That part will offer 8 CPU cores and 96 MB of L3 cache, making it the most direct successor to the Ryzen 7 5800X3D.

Ultimately, all three chips will serve to update AMD's product stack by combining the strengths of the Zen 4 CPU architecture with the performance benefits of the extra L3 cache, which during the overlapping period of the last several months, has been split between the Ryzen 5000 and Ryzen 7000 families. In short, PC gamers will finally be able to have their cake and eat it too, gaining access to AMD's Zen 4 microarchitecture and its myriad of benefits (higher IPC, higher clockspeeds, DDR5, PCIe 5) with a nice helping of additional L3 cache slathered on top.

From that stack, today we're reviewing the new flagship Ryzen 9 7950X3D. The 7950X3D offers 16 Zen 4 cores spread over two CCDs (8C/16T per CCD). AMD had to elect one of the CCDs to stack the additional L3 cache onto, resulting in a new-to-AMD heterogeneous CPU design, but they do have some special sauce as a garnish to make it work. We aim to determine if the Ryzen 9 7950X3D is the chip gamers have been yearning for and how it stacks up against other Ryzen 7000 chips (and Intel's 13th Gen) in our test suite.

from AnandTech https://ift.tt/p3OVWUZ
via IFTTT

AMD plans to harness the power of AI to transform gaming with its next-gen CPUs

https://ift.tt/7KiWEcZ

AMD executives David Wang and Rick Bergman have confirmed that we’ll be seeing a lot more AI in the next generation of graphics cards from the tech giant, which will be built on AMD’s RDNA 4 architecture.

In a recent interview with the Japanese gaming website 4gamer, the AMD execs detailed some of what we can expect from RDNA 4. Naturally, front and center was confirmation that we’ll be seeing the second iteration of Team Red’s AI Accelerator cores (similar to Nvidia’s Tensor cores), which were first introduced in the current-gen RDNA 3 GPUs - such as the excellent Radeon RX 7900 XTX, currently the best AMD graphics card on the market.

Nvidia’s tech is still lightyears ahead of AMD when it comes to AI processes - just look at the RTX 4090 - but these second-gen AI cores should offer a serious step up. Beyond the Accelerator cores, the pair also discussed some other nifty new features, most importantly a new self-contained GPU pipeline that allows for rendering and texture processes to be generated exclusively on the GPU without needing to communicate with the CPU.

This has massive potential to boost the processing speed of RDNA 4 GPUs, since it won’t need to rely on the CPU and system RAM to carry out some of its workloads, effectively cutting out two potential system bottlenecks. According to Wang and Bergman, we can expect a massive 2.2x performance boost over the current RDNA 3 cards.

Analysis: AMD is now fully aboard the AI bandwagon

Perhaps even more interestingly, Wang was keen to discuss the implementation of AI within the gaming space. In the 4gamer interview, he espoused that AI hardware shouldn’t just be used purely for improving the graphical aspect of games, positing that inferences cores on the GPU could actually be used to improve the gameplay experience itself.

This has some potentially frightening implications - imagine playing a shooter where the enemies are actually powered by a deep-learning AI housed in your PC, which genuinely reacts, learns, and adapts to your actions. It does sound just the teensiest bit dystopian to imagine an in-game boss battle against ChatGPT.

Still, on a slightly less terrifying level, it’s interesting to consider the uses AI cores could have for things like NPC pathfinding in games. Every gamer has, at some point, complained about game characters walking too slowly or getting stuck on a doorframe due to wonky pathfinding - Wang believes AI cores have the potential to solve a lot of immersion-breaking issues such as this.

It’s somewhat amusing to see AMD commit so strongly to AI now, since Team Red basically mocked Nvidia last year for relying too heavily on AI tech in its RTX GPUs and DLSS. In its defense, AMD did prove that upscaling tech such as DLSS can be achieved without the use of AI, as seen in the Radeon FidelityFX Super Resolution (FSR) software. It sounds like AMD’s higher-ups have become AI converts, but they have different - and intriguing - ideas about how it should be employed compared to Nvidia’s plans. We can’t blame them for wanting to go in a different direction; Nvidia has been in a lot of hot water recently.

While we don’t know exactly when RDNA 4 will land, AMD’s 2-year cadence when it comes to GPU launches means that we can most likely expect them sometime in late 2024. Bergman assured 4gamer that the new cards will be released in the near future, though ‘near’ might be a stretch here.

from TechRadar: computing components news https://ift.tt/TUtHgD5
via IFTTT

Silicon Motion SM2268XT DRAM-less NVMe SSD Controller: PCIe 4.0 Speeds on a Budget

https://ift.tt/ezs82hB

The excitement in the client SSD space recently has understandably been on the Gen 5 front. However, cooling requirements have made it difficult for vendors to bring effective M.2 NVMe Gen 5 SSDs to the market. In that context, it appears that Gen 4 M.2 SSDs will continue to have a much longer runway than was previously estimated. In order to better serve that market segment, Silicon Motion is introducing a new product in their PCIe 4.0-capable NVMe SSD controller lineup. The company's roadmap is not much of a secret, as we do expect their Gen 5 client controllers to land in Q4 2023 - in fact, an end product based on it from ADATA was already demonstrated by ADATA at the 2023 CES. However, today marks the official launch date for the SM2268XT - the third-generation DRAM-less Gen 4 SSD controller meant to play in the entry-level segment in terms of pricing, but move to the high end in terms of performance .

The key updates in the SM2268XT over the SM2267XT and SM2269XT are the increase in the per-channel data rate from 1600 MT/s to 3200 MT/s and support for the latest 2xx layer 3D TLC (as well as QLC) from different flash vendors. The new controller also brings in support for some NVMe 2.0 features (compared to NVMe 1.4 in the SM2267XT and SM2269XT). Like the SM2269XT, the LDPC engine codeword size in the SM2268XT is also 4KB (compared to 2KB in the SM2267XT). The move to a 12nm process also brings in better power efficiency.

THE SM2268XT will be competing against in-house controllers from Western Digital (such as the one used in the WD_BLACK SN770), and the upcoming Phison E21T as well as InnoGrit's RainierQX IG5220. The claimed performance numbers across all four corners for the SM2268XT are leading in its class.

Silicon Motion Client/Consumer Gen 4 NVMe SSD Controllers
	SM2264	SM2267	SM2267XT	SM2269XT	SM2268XT
Market Segment	High-End Consumer	Mainstream Consumer
Manufacturing Process	12nm	28nm		12nm
Arm CPU Cores	4x ARM Cortex R8	2x ARM Cortex R5		2x ARM Cortex R8
Error Correction	4KB LDPC	2KB LDPC		4KB LDPC
DRAM Support	DDR4, LPDDR4(X)	DDR3, DDR3L, LPDDR3, DDR4, LPDDR4	None / HMB
Host Interface	PCIe 4.0 x4
NAND Channels, Interface Speed	8ch 1600 MT/s	8ch 1200 MT/s	4ch 1200 MT/s	4ch 1600 MT/s	4ch 3200 MT/s
CEs per Channel	8	8	4
Sequential Read	7500 MB/s	3900 MB/s	3500 MB/s	5100 MB/s	7400 MB/s
Sequential Write	7000 MB/s	3500 MB/s	3000 MB/s	4800 MB/s	6500 MB/s
4KB Random Read IOPS	1.3M	500K	500K (HMB) / 200K (No HMB)	900K	1.2M
4KB Random Write IOPS	1.2M	500K	500K	900K	1.2M

The company indicates that the SM2268XT is currently sampling to its key customers and the launch of SSDs based on it should be imminent. We expect the usual suspects such as ADATA to announce SSDs based on the new controller soon.

from AnandTech https://ift.tt/0Jb6pqs
via IFTTT

Intel’s secret overclocking app is supremely user-friendly – so why hasn’t it been released yet?

https://ift.tt/Uk4fnXT

Intel has overclocking software which could really take the sting out of juicing up a CPU to perform better, and make the process more accessible to less tech-savvy types – but you can’t get hold of this app yet.

Why not? Well, as Tom’s Hardware reports, the ROC app (which stands for Real-time OverClocking) was shown off in a YouTube video by expert overclocker Der8auer, who got to play with the software during a visit to an Intel lab in Portland, Oregon. And for the moment, this application remains something that’s only used by Intel staff to internally test and play with CPUs – but we’re hoping that could change down the line.

Der8auer used ROC in conjunction with an Intel Core i9-13900HK processor, which is a (flagship) laptop chip – but the app can be used with any CPU, of course, desktop or mobile – and the results were extremely impressive.

The 13900HK could be driven to 5.8GHz and remained stable, on air cooling, with temperatures not exceeding 75C, and only when 6GHz was reached (Der8auer was working in 200MHz increments) did the laptop crash.

Remember, this is a laptop processor, not a desktop model, so these are eye-opening speeds indeed (achieved without exotic cooling).

---

Analysis: Intel could – and should – ROC the overclocking world

This again underlines what Intel’s 13th-generation silicon (Raptor Lake) is capable of in terms of overclocking (as well as performance), taking some considerable strides forward compared to the previous-gen (Alder Lake).

The trouble is, overclocking isn’t for everyone, but it certainly sounds like ROC takes some healthy steps towards making the process more widely accessible to desktop PC and laptop owners alike.

As Der8auer explains, ROC is like a slimmed-down version of Intel’s current XTU app (Extreme Tuning Utility), with a far more intuitive interface that is much easier to work with than XTU.

It sounds like something that really needs to be out there for everyone who has an Intel CPU to make use of, but thus far, the company hasn’t indicated any plans to release the app. That doesn’t mean it won’t happen, of course; and even if ROC itself isn’t debuted for public consumption, maybe the interface or some of the functionality will make its way across to XTU. Or we’ll see a different app entirely with some of ROC’s features.

At any rate, this is a promising glimpse that the future of overclocking could well be a much less intimidating place (keep those fingers crossed).

from TechRadar: computing components news https://ift.tt/ecXu79t
via IFTTT

ASRock Industrial's 4X4 BOX 7000/D5 Series Brings Zen 3+ and USB4 40Gbps to UCFF Systems

https://ift.tt/cZ29On0

ASRock Industrial's lineup of ultra-compact form-factor machines in the NUC BOX (Intel-based) and 4X4 BOX (AMD-based) series has gained popularity over the last couple of years. Being the first to market with the latest platforms has been one of the key reasons behind this. In 2022, the company had launched the Intel Alder Lake and AMD Cezanne UCFF systems together, with the NUC BOX-1200 series and the 4X4 BOX-5000 series becoming available for purchase within a few weeks of each other. Earlier this year, the Intel Raptor Lake-based NUC BOX-1300 series was launched (our review) and is already available for purchase. The company recently took the wraps off the 4X4 BOX 7000/D5 series based on AMD's low-power Rembrandt-R APUs. These APUs sport Zen 3+ cores along with a RDNA2 iGPU fabricated on a TSMC 6nm process.

One of the key updates in the new 4X4 BOX systems is the move to DDR5 SODIMMs. The other updates in the platform such as support for a PCIe 4.0 x4 M.2 NVMe SSD and USB4 40 Gbps bring it almost on par with the premium UCFF systems based on Intel processors. Full-fledged USB4 support inclusive of PCIe tunneling has been somewhat of a hit or miss on AMD platforms, as many OEMs have refrained from integrating the necessary board components to enable it. AMD itself had some work to do on the firmware side before the feature baked into the hardware of Rembrandt and later APUs could be enabled. The good news here is that the ASRock Industrial 4X4 BOX 7000/D5 series has two USB4 ports capable of supporting DisplayPort 1.4 signals as well as PCIe tunneling - this means that Thunderbolt 3 peripherals should work when connected to those Type-C ports.

The 4X4 BOX 7000/D5 series comes in two flavors. The specifications of both models are summarized in the table below.

ASRock Industrial 4X4 BOX 7000/D5 (Rembrandt-R) Lineup
Model		4X4 BOX-7735U/D5	4X4 BOX-7535U/D5
CPU		AMD Ryzen 7 7735U 8C / 16T 2.7 GHz (Up to 4.75 GHz) 28W	AMD Ryzen 5 7535U 6C / 12T 2.9 GHz (Up to 4.55 GHz) 28W
CPU		AMD Radeon 680M (12 CU / 768 Shaders) @ 2.2 GHz	AMD Radeon 660M (6 CU / 384 Shaders) @ 1.9 GHz
DRAM		Two DDR5 SO-DIMM slots Up to 64 GB of DDR5-4800 in dual-channel mode
Motherboard		4.02" x 4.09" UCFF
Storage	SSD	1x M.2-22(42/60/80) (PCIe 4.0 x4 (CPU-direct))
	DFF	1 × SATA III Port (for 2.5" drive)
Wireless		Mediatek MT7922 (RZ616)? Wi-Fi 6E 2x2 802.11ax Wi-Fi (2.4Gbps) + Bluetooth 5.2 module
Ethernet		1x 2.5 GbE RJ-45 (Realtek RTL8125) 1x GbE RJ-45 (Realtek RTL8111EPV with DASH Support)
USB	Front	1 × USB 3.2 Gen 2 Type-A 2x USB4 (with DisplayPort 1.4a Alt Mode)
	Rear	2 × USB 2.0 Type-A
Display Outputs		1 × HDMI 2.1 (Rear, up to 8Kp60) 1 × DisplayPort 1.4a (Rear, up to 4Kp60) 2 × DisplayPort 1.4a (using Front Panel Type-C ports, up to 4Kp60)
Audio		1 × 3.5mm audio jack (Realtek ALC233)
PSU		External (19V/90W)
Dimensions		Length: 117.5 mm Width: 110 mm Height: 47.85 mm

Note that the M.2 2280 support is enabled by a separate bracket, similar to the previous NUC BOX and 4X4 BOX systems with 2.5" drive support and dual LAN capabilities.

Overall, these systems bring the AMD UCFF scene on par with the high-end Intel NUCs and its clones - except for the newer NUC BOX-1300/D5 series which has Thunderbolt 4 ports that also have USB 3.2 Gen 2x2 support for 20Gbps PSSDs. The USB4 ports in the 4X4 BOX 7000/D5 support only up to USB 3.2 Gen 2 speeds in legacy USB (non-PCIe tunneling) mode. The RDNA2 iGPU in the new machines should also help these systems perform as well as the the Alder Lake and Raptor Lake mini-PCs in graphics-heavy workloads.

We have reached out to ASRock Industrial for clarity on market availability dates and pricing, and will update the article with the details after receiving them.

Gallery: ASRock Industrial's 4X4 BOX 7000/D5 Series Brings Zen 3+ and USB4 40Gbps to UCFF Systems

from AnandTech https://ift.tt/ybpLkNr
via IFTTT

Is your CPU struggling when gaming? Microsoft DirectStorage could come to the rescue

https://ift.tt/ioPpfvq

Microsoft’s DirectStorage tech has had its first outing on PC courtesy of Forspoken, and according to some new testing, the feature boosts not just loading times considerably (for those with the right hardware), but also in-game frame rates in certain scenarios.

Neowin flagged up a YouTube video in which Compusemble tested Forspoken, observing that: “DirectStorage has a pretty significant effect on frame rate, frame time consistency, and GPU utilization at 1080p, whereas it has virtually no effect at higher resolutions.”

What seems to be happening here, Compusemble elaborates, is that DirectStorage only kicks in like this to bolster frame rates when the game is CPU-limited, meaning the processor is struggling with its workload, whereas the graphics card is ticking along just fine.

That’s the case when playing Forspoken at 1080p (Full HD resolution) for Compusemble, but when the resolution is cranked up to 1440p, the YouTuber becomes GPU-limited instead – and the frame rate boost vanishes into thin air.

The testing was performed with a Sabrent Rocket 4 Plus-G SSD (2TB) in a gaming PC with an RTX 3080 Ti graphics card plus Ryzen 7 7700X CPU. Remember, to get the full benefit of DirectStorage you need an NVMe SSD, and furthermore running Windows 11 gives better results than Windows 10 (the test PC used Windows 11).

---

Analysis: New GPU decompression tech at work? Not the case, in fact…

We knew that DirectStorage would do more than just speed up loading times, and that it’d make a difference in big open world environments, supercharging the in-game loading of assets on-the-fly. But this is the first tantalizing evidence that the tech really can boost frame rates on PC – at least in certain scenarios, namely where the GPU isn’t breaking a sweat, but the CPU is getting hammered.

So, this could make you think that this is clear evidence that the GPU decompression tech in DirectStorage 1.1 – the latest version that Forspoken uses – is coming into play here. This allows the GPU to handle decompression of game assets (which are compressed for size reasons), and the graphics card can do this far more efficiently than the CPU – helping to take a load off the latter when it’s struggling, and thus improving frame rates in CPU-limited scenarios.

That’s the obvious theory, but as Digital Foundry separately points out, even though Forspoken does utilize DirectStorage 1.1, apparently GPU decompression isn’t actually active here, as there is “no GPU compute usage spike when the game is doing a dedicated load” (when obviously there should be if the card was busy with decompression work).

All in all, this is a bit of an odd one, and as Digital Foundry (and others) have observed, there are quite a few issues and glitches with Forspoken on PC. (Not exactly uncommon with games which are subject to multiple delays, and indeed console ports, it has to be said).

The conclusions we can draw from a single game are, of course, limited, but we can say that if these are the results without GPU decompression being made use of, it’s exciting to imagine what other games that actually employ this tech might do for frame rates (at least away from high resolutions). Although that excitement is somewhat tempered when we think about what the next game to use DirectStorage on PC might be – which, well, isn’t yet clear.

We simply don’t know anything else about what games will be supported on the Windows platform after Forspoken, which is somewhat ominous in terms of suggesting the wait for a second title – with any luck featuring an even better implementation of DS 1.1 – could be a long one. But next year, maybe, DirectStorage could become a compelling reason for gamers to upgrade to Windows 11 if they haven’t already made the leap from Windows 10.

from TechRadar: computing components news https://ift.tt/CchGzbJ
via IFTTT

Intel’s next-gen Meteor Lake CPUs could be amazing for laptops

https://ift.tt/CA3G50s

Meteor Lake CPUs, Intel’s next-gen chips to follow Raptor Lake, are going to provide some huge boosts on the efficiency front if a fresh rumor proves to be correct.

As Wccftech reports, this is another from that fountain of speculation which is Twitter, and one of the better-known hardware leakers on that platform, namely Raichu.

The target of Meteor is to realize 1.5x+ efficiency compare to the Raptor when it has the same perf. (same core processor, P+E)🧐February 6, 2023

See more

Raichu believes that Intel is targeting 50% power efficiency gains – at least – with its 14th-gen Meteor Lake processors, compared to current Raptor Lake silicon. In other words, when running at the same performance level as a 13th-gen chip, a next-gen model will use a third less power.

Raichu does mention performance briefly in that Twitter thread, but only to confirm that Meteor Lake will increase it – as would be fully expected. However, the leaker doesn’t give us any idea of the kind of uptick to expect on that front from these next-gen processors.

We’re also told that integrated graphics for Meteor Lake will nearly double performance levels compared to Raptor Lake, which would be seriously impressive, too.

---

Analysis: Yet more evidence of Intel’s fresh focus on efficiency

All this is great news for laptops, if it pans out of course – we must always be cautious of how much stock we put in rumors. That said, there has already been quite a bit of buzz about how Meteor Lake will focus on efficiency, driving forward with ever-increasing amounts of efficiency cores that could benefit from a new architecture. (Indeed, past rumors have suggested that the 14th-gen may not even have Core i9 models for the desktop, such might be the focus on mobile).

A 50% increase in efficiency, or maybe more, will mean that laptops will be able to pack more powerful chips that’ll be thermally okay within the confines of a small chassis. And if integrated graphics are nearly twice as performant as Raptor Lake – with Intel already making great strides on that front – we can expect thin-and-light notebooks that offer impressive performance not just for apps, but also for light gaming duties too.

The word from the grapevine is that Lunar Lake, in theory the 16th-generation for Intel, will also drive hard with efficiency to the point that it’s built with laptops in mind. All of this might leave desktop users worrying that Intel could neglect performance, with only Arrow Lake, which will be the 15th-gen, likely to pack a Core i9 heavyweight desktop CPU to take the baton from the current 13900K.

Mind you, that’s really treading deeper into the territory of speculation, so desktop PC enthusiasts shouldn’t lose too much sleep – not yet, anyway. But it is very much starting to sound more and more like Intel will focus big on efficiency over the next few generations, making a distinct change from recent iterations of the Core family which have really pushed hard to get beefy performance levels, putting the power usage pedal to the proverbial metal (at least on the desktop at the high-end, anyway).

from TechRadar: computing components news https://ift.tt/J1PBS3w
via IFTTT