Intel needs to release a new generation of CPUs every so often, so that PC manufacturers have higher numbers to promote and buyers feel as though they’re getting something better and faster. Of course, as buyers, it’s nice to get all the latest standards and the best possible performance at every price level, but the simple fact is that older hardware is very often good enough, and very few people actually need to upgrade every year. Another challenge for Intel is that its manufacturing efforts have suffered a few setbacks over the past several years – we were supposed to be well beyond the 10nm node by now, but the company is still unable to serve all its market segments with 10nm parts.
It’s been nearly six years since Intel first announced that we would not see 10nm desktop CPUs on schedule in 2016, and that instead, the previous Skylake architecture on the 14nm manufacturing process would be improved and kept in use. Until now, Intel has repeated the process each year, delivering solid improvements by tweaking the same old architecture.
Limited 10nm manufacturing capacity has wisely been allocated to the laptop segment where power efficiency and chip density are far more important, but Intel unfortunately tied its chip designs to a particular manufacturing node. That has meant a split strategy, where laptops have benefited from new features with more modern architectures on the 10nm node, while desktops have lagged somewhat, stuck with derivatives of the same old 14nm design, yet both have been marketed as the same generation of chips. Intel is hoping to bridge that gap with the 11th Gen ‘Rocket Lake‘ desktop family, with one very interesting trick.
The many derivatives of Skylake have finally been laid to rest, but we’re giving 14nm one last hurrah with a backported version of the 10nm ‘Ice Lake’ architecture which served the previous generation of laptops. What Intel has done is essentially transpose a more modern architecture onto an older manufacturing process – a huge undertaking, but one that it hopes will pay off.
This is a short term goal, since native 10nm 12th Gen desktop ‘Alder Lake’ CPUs have already been confirmed for a late 2021 launch. At the end of the day, though, it’s all about performance, as indicated by model numbers – most users don’t care about the architecture inside their CPUs, much less the minutiae of how they are manufactured. If you do want to know all the details though, we’ve got you covered. In this review, we’ll see whether Intel’s unusual manoeuvre has paid off, and whether there are any surprises that you need to be aware of.
Intel ‘Rocket Lake’ 11th Gen Core architecture and platform
Intel now uses different codenames for its CPUs and the architecture of the cores within them, so we have ‘Rocket Lake’, the 11th Gen desktop family, based on ‘Sunny Cove’ cores (now named ‘Cypress Cove’). If you’ve been following product roadmaps, you’ll recognise Sunny Cove from the 10th Gen ‘Ice Lake’ laptop processor family as well as the hybrid ‘Lakefield’ CPU shown off last year. That means there’s a significant difference between the 11th Gen ‘Rocket Lake’ desktop and ‘Tiger Lake’ laptop CPU families.
This CPU core architecture marks a break from the past several years of refreshes, so 11th Gen desktop Core CPUs gain a few features that have already debuted on laptops, most notably AI and neural network acceleration. (To add to any potential confusion, there are now also 11th Gen desktop ‘Tiger Lake’ CPUs, although these are likely to be found only in OEM systems since they are designed to be soldered to motherboards and not sold as modular socketed CPUs).
Along with these CPU cores, most Rocket Lake CPUs also have integrated Intel Xe-LP graphics, which you might recognise from the 11th Gen ‘Tiger Lake’ laptop family. Only models with an -F suffix do not have integrated graphics capabilities. If you choose one of those models you’ll also miss out on updated video encoding and decoding hardware.
Intel has just gone with its old UHD Graphics tag rather than playing up the Xe name, which tells you not to get too excited. That’s probably because Intel has cut back on the number of graphics execution units compared to Tiger Lake CPUs. This won’t be a problem for most people, since high-end PCs with CPUs like this usually have discrete graphics cards, but a big leap in integrated graphics power would have been nice, especially with graphics cards so scarce these days and prices shooting through the roof.
Despite the significant difference in architectures between 10th and 11th Gen desktop Core CPUs, Rocket Lake is compatible with most previous-gen 400-series motherboards, and older CPUs will work on new boards with 500-series chipsets as well. Intel surprisingly managed to implement its backported architecture to use the same socket as last year’s CPUs, but you don’t get all the benefits of the platform unless both CPU and motherboard are new. Of course you’ll want to check exact details of required BIOS updates and support with your motherboard vendor if you’re planning a drop-in upgrade.
Platform-level features include 20 PCIe 4.0 lanes, extra bandwidth between the CPU and platform controller, support for up to DDR4-3200 RAM, and native USB 3.2×2 (20Gbps). Thunderbolt 4, Wi-Fi 6E, and 2.5Gbit Ethernet can be added by motherboard OEMs. Memory overclocking is now supported on mid-range 500-series chipsets, not just the top-end Z590.
Having more, faster PCIe lanes means that you can run a graphics card at x16 as well as an SSD at x4 simultaneously. There are still 24 additional PCIe 3.0 lanes originating from the platform controller. It will be up to motherboard makers to allocate lanes between slots and other onboard hardware, and to decide what to let users configure.
Intel ‘Rocket Lake’ 11th Gen specifications and features
As of launch time, there are only Core i5, Core i7, and Core i9 models in the Rocket Lake family. The market for lower-end Core i3 and Pentium CPUs will be served by refreshed 10th Gen models, at least for now. The lineup includes low-TDP models with the -T suffix, overclockable unlocked -K models, and -F models without integrated graphics. There are also a few new differences in terms of differentiation between these product tiers that you should know about.
Core i5 models all have six cores and 12 threads, and Core i7 models all have 8 cores and 16 threads. Most interestingly, Core i9 models do not have more cores – this generation tops out at eight. This might seem like a huge regression, considering that the 10th Gen ‘Comet Lake’ Core i9-10900K had 10 cores. It’s also a bold move since prime competitor AMD has offered consumer CPUs with 12 and 16 cores for nearly two years now, and consumers definitely do look at this as a measure of performance.
Intel acknowledges that this is a potential weakness in terms of marketing, and that in fact the previous-gen flagship might outperform the new one in workloads that are heavily multi-threaded. However, the company says this is balanced by Rocket Lake’s overall performance uplift, which is claimed to be in the region of 19 percent in terms of instructions per clock, plus the new integrated graphics and the platform-level improvements. It isn’t clear exactly why Intel decided to risk being perceived as regressing – it could have to do with die space, cost, heat and power budgets, boost frequency targets, or manufacturing efficiencies.
So with identical core counts, how are the Core i9 and Core i7 tiers differentiated? One factor is a fancy retail box for the Core i9. More importantly, there’s memory speed – all models are rated to support DDR4-3200 RAM, but only the Core i9 can do so at 1:1 memory timing, while others top out at DDR4-2933 officially, or DDR4-3200 at 2:1. Interestingly the Asus ROG Maximus XIII Hero motherboard used for this review claims to support overclocking up to DDR4-5333, so this limitation is clearly more of a guideline.
The other big differentiator is Intel’s Thermal Velocity Boost feature, which is exclusive to the Core i9 tier. As its name suggests, this is essentially an opportunistic way to boost clock speeds beyond even the Turbo Boost rating for a short period of time if power and thermal conditions allow for it.
This means there are now multiply frequency levels in play – the Core i9-11900K has a base frequency of 3.5GHz, and then Intel’s longstanding Turbo Boost feature will allow all eight cores to go up to 4.7GHz or trade off so that two cores go up to 5.1GHz when needed. On top of that, Turbo Boost Max 3.0 allows two “favoured cores”, the ones that are found to be most efficient, to go up to 5.2GHz. On top of all that, Thermal Velocity Boost (TVB) allows one or two cores to go up to 5.3GHz on demand as long as your cooler can keep the CPU below a certain temperature threshold. And as if all that wasn’t enough, Intel’s new Adaptive Boost Technology (ABT) works counter to TVB, bumping up the other six cores to as much as 5.1GHz (the two-core Turbo Boost limit) when it makes more sense to have all cores running slightly slower as opposed to two cores running a lot faster.
It can get quite confusing, and it really just shows that clock speed isn’t a single number and can’t easily be used to compare one processor against another from a different tier, generation, or manufacturer. In fact, multiple tests of the same CPU under different conditions might also produce different results – your motherboard, power supply, cooler, case, and even ambient conditions will affect when these boost targets are reached and how long they can be sustained. To simplify it all, “classic” Turbo Boost is what your CPU should be capable of all the time, while TVB and ABT will allow power users to make the most of high-end cooling hardware. Intel has also stated that ABT does not violate your warranty since it an advertised feature and not considered overclocking.
Our Core i5-11600K is a lot simpler – it has a 3.9GHz base speed and 4.9GHz Turbo Boost speed. It does not get Turbo Boost Max 3.0, TVB, or ABT. The -K suffix means it is overclockable just like its sibling. Both CPUs have 125W TDP ratings, and neither comes with a stock cooler in the retail box.
Asus ROG Maximus XIII Hero specifications and features
If you’re going all out on a Core i9-11900K, you’ll want an ultra-premium motherboard to go with it. The Asus ROG Maximus XIII Hero is one of the top-end ATX gaming motherboards from Asus this generation. It has plenty of bells and whistles, as you might expect given its street price of approximately Rs. 45,500. It is of course based on the Intel Z590 platform which was released specifically to let buyers take advantage of the incremental updates that Rocket Lake offers over Comet Lake.
Key features include high-end power circuitry with 14+2 stages and Japanese capacitors; massive heatsinks covering the power components, M.2 slots and chipset itself; dual Thunderbolt 4 ports plus a USB 3.2×2 header; and of course aggressive styling with RGB LEDs. Asus claims that this board supports memory overclocking up to DDR4-5333.
This motherboard looks suitably expensive, with nearly all-black components and chunky matte metal heatsinks that feel solid. The RGB LED lighting is relatively subtle and is contained to two zones: the heatsinks around the CPU socket, and the heatsink over the platform controller. Thankfully Asus has ditched its cringey “cybertext” graffiti-like pattern.
There’s adequate clearance around the CPU socket for all but the largest of air coolers, plus headers for all kinds of liquid cooling apparatus. This motherboard has three full-sized PCIe slots (only the first one can harness a full 16 lanes) and one more x1 slot, plus there’s somehow room for four M.2 slots, all laying flat with thermal pads and heatsinks. Unfortunately, all these slots plus the six SATA ports share bandwidth, so you can’t use them all at the same time. This is a limit that Intel imposes on its consumer CPU tier. You’ll also need a Rocket Lake CPU to benefit from PCIe 4.0 speed, and even then it’ll only benefit the first PCIe and M.2 slots. Some slots will be disabled and others will fall back to PCIe 3.0 if using a previous-gen CPU.
The rear port cluster is absolutely packed, and there’s an integrated IO shield which makes life so much easier when assembling a PC. To the left, there are buttons to reset and update the BIOS right off a USB drive. Then, you get two 2.5Gb Ethernet ports and two Wi-Fi antenna terminals, HDMI 2.0 video out, five gold-plated 3.5mm audio jacks plus optical S/PDIF output, two USB 2.0 ports, six USB 3.2 Gen2 Type-A ports, and two Thunderbolt 4 ports which also support DisplayPort 1.4 video output.
Audio is handled by a Realtek ALC4082 codec and there’s also ESS Sabre DAC for the front panel headphone connector. You also get tri-band Wi-Fi 6E (802.11 ax/az) and Bluetooth 5.2, which require the latest versions of Windows 10 to work. There are internal headers for one USB 3.2×2 Type-C port plus two USB 3.2 Gen 1 ports and four more USB 2.0 ports.
Enthusiast-friendly features include a surface-mount power button and another programmable ‘Flexkey’ button that will reset your PC by default but can be changed to an RGB lighting toggle or safe boot trigger. There’s a two-digit alphanumeric code readout for diagnostics. Of course, you also get plenty of RGB and ARGB LED headers to make your whole case look snazzy. The box includes the Wi-Fi antenna, all the cables and screws you’ll need, a bracket to take the weight off a large graphics card, a detailed manual and even a driver DVD, plus some Asus ROG branded merchandise.
The graphical BIOS is easy to use and while there are loads of tweaking options, things are labelled well enough and there are some text descriptions. Asus’ Armory Crate software lets you synchronise RGB lighting across supported Asus products and manage BIOS and driver updates. AI Suite 3 is for Windows-based overclocking, fan control, and monitoring.
The Asus ROG Maximus XIII Hero is a great-looking high-end board and has nearly every feature you might want if you’re an enthusiast. The physical layout is very good and the BIOS is easy to work with. You’ll have to really consider whether you need all these features though, since there’s no shortage of other Z590 motherboards that cost quite a lot less.
Intel Core i9-11900K and Core i5-11600K setup and performance
In addition to the ROG Maximus XIII Hero provided by Intel along with the CPUs for this review, our open-air test bench consists of a 2x16GB Corsair Dominator Platinum RGB DDR4-3600 RAM kit, Sapphire Nitro+ Radeon RX 590 graphics card (when not testing the CPUs’ integrated graphics), WD Black (2018) 1TB PCIe SSD and 1TB Samsung SSD 860 EVO SSD, Corsair H150i Elite Capellix 360mm AIO liquid cooler, Corsair RM850 power supply, and Asus PB287Q monitor. All tests were run under Windows 10 20H2 with the latest patches and updates applied. Intel provided a BIOS update and graphics drivers, and all other drivers used were the latest available public releases at the time.
Setup and installation were very easy. The motherboard socket has enough clearance, and the layout is convenient, but there are some rather sharp edges and corners on the heat sinks. Setting up the Corsair H150i AIO cooler took the most time, but it comes with clear instructions. Every possible screw and accessory for Intel as well as AMD CPUs is neatly labelled and sorted into bags within bags so you can never go wrong. The pump head, fan power and fan RGB leads all plug into a separate “iCue Commander Core” module which is about the size of a SATA SSD. It gets its power from a SATA connector and needs one internal USB 2.0 header which will block off two ports. You might need to be a bit creative with cable management, but adhesive strips are provided.
The Dominator Platinum RAM has individually addressable RGB LEDs. There are more on the H150i Elite Capellix cooler’s pump head and the three 120mm fans. The Commander Core module has additional connectors for your case fans or other products that you might want to synchronise. Corsair’s iCue software can now also work with Asus’ Aura Sync RGB framework so the motherboard, components, and peripherals can all be managed as one.
Thanks to lockdown-related restrictions throughout most of 2020, we don’t have last year’s 10th Gen Core i9-10900K or a Ryzen 9 5950X to compare performance numbers with. However, most users will be upgrading after a gap of a few years, so the comparisons we’ve presented below should still be meaningful. We can see how the Core i9-11900K and Core i5-11600K match up against the slightly older Core i9-9900K and Ryzen 3xxx series. The Core i9-10980XE based on the Cascade Lake-X (Skylake-X derivative) architecture is also an interesting reference point, being priced only a little higher. We’ve even thrown in some scores from our Ryzen Threadripper 3970X to illustrate the differences (and in some cases, the similarities) between these vastly different classes of CPUs.
In some cases where benchmarks have been updated and scores for older CPUs might not be comparable, they are not shown here, but you can click through to their individual reviews to read more about them. Test bench configurations, OS updates and driver versions also do affect scores, so there are some variances to account for as well.
Intel Core i9-11900K | Intel Core i5-11600K | AMD Ryzen 9 3900X |
AMD Ryzen 7 3700X | Intel Core i9-9900K | Intel Core i9-10980XE | AMD Ryzen Threadripper 3970X | |
---|---|---|---|---|---|---|---|
CPU tests | |||||||
Cinebench R20 CPU single-threaded | 628 | 595 | 495 | 489 | NA | 452 | 515 |
Cinebench R20 CPU multi-threaded | 5,927 | 4,292 | 6,785 | 4,521 | NA | 8,729 | 17,069 |
Cinebench R23 CPU single-threaded (10 mins) | 1,676 | 1,542 | NA | NA | NA | NA | NA |
Cinebench R23 CPU multi-threaded (10 mins) | 15,373 | 11,094 | NA | NA | NA | NA | NA |
PCMark 10 standard | 7,474 | 5,036 | 6,597 | 6,861 | 4,957 | 6,914 | 6,637 |
PCMark 10 extended | 8,466 | 8,137 | 6,807 | 7,937 | 3,435 | 7,967 | 7,681 |
3DMark Fire Strike Ultra (physics) | 26,716 | 22,328 | 27,471 | 23,933 | 21,550 | 28,111 | 22,010 |
3DMark Time Spy (CPU) | 11,000 | 8,446 | NA | NA | NA | NA | NA |
Geekbench 5 single-threaded | 1,777 | 1,654 | NA | NA | NA | NA | NA |
Geekbench 5 multi-threaded | 9,536 | 7,582 | NA | NA | NA | NA | NA |
POVRay* | 55 seconds | 1 minute, 16 seconds | 41 seconds | 1 minute, 1 second | 57 seconds | 35 seconds | 18 seconds |
VRAY CPU* | 54 seconds | 1 minute, 14 seconds | 48 seconds | 1 minute, 9 seconds | 1 minute, 2 seconds | 37 seconds | 20 seconds |
Corona Renderer Benchmark* | 1 minute, 35 seconds | 2 minutes, 11 seconds | 1 minute, 19 seconds | 1 minute, 57 seconds | 1 minute, 42 seconds | 57 seconds | 29 seconds |
WebXprt | 308 | 310 | 260 | 270 | 272 | 237 | 256 |
Jetstream 2 | 209.32 | 196.946 | NA | NA | NA | NA | NA |
Speedometer | 216 | 201.8 | NA | NA | NA | NA | NA |
SiSoft SANDRA CPU arithmetic | 313.18GOPS | 224.44GOPS | 366GOPS | 244.52GOPS | 282.45GOPS | 496GOPS | 940.69GOPS |
SiSoft SANDRA CPU multimedia | 1.3GPix/s | 948.64MPix/s | 1.26GPix/s | 876.32MPix/s | 918.22MPix/s | 2.13GPix/s | 3.31GPix/s |
SiSoft SANDRA CPU cryptography | 14GBps | 16.88GBps | 18.09GBps | 18.36GBps | 12.12GBps | 25.65GBps | 41.42GBps |
SiSoft SANDRA cache bandwidth | 429GBps | 351GBps | 589.9GBps | 378GBps | 307.32GBps | 701.53GBps | 1.73TBps |
7Zip file compression* | 1 minute, 37 seconds | 1 minute, 40 seconds | 1 minute, 33 seconds | 1 minute, 37 seconds | 2 minutes, 12 seconds | 1 minute, 8 seconds | 56 seconds |
Handbrake video encoding* | 32 seconds | 41 seconds | 35 seconds | 43 seconds | 39 seconds | 37 seconds | 30 seconds |
Civilization VI AI benchmark (average) | 8.83 seconds | 8.87s | NA | NA | NA | NA | NA |
Integrated GPU tests | |||||||
Unigine Superposition 1080p Medium | 1,188 | 1,207 | |||||
3DMark Time Spy | 698 | 702 | |||||
3DMark Night Raid | 9,423 | 9,448 | |||||
Discrete GPU tests | |||||||
3DMark Fire Strike Ultra | 3,638 | 3,746 | |||||
3DMark Time Spy | 5,290 | 5,200 | |||||
Far Cry 5 1080p Ultra | 82fps | 81fps | |||||
Far Cry 5 1440p Ultra | 58fps | 57fps | |||||
Assassin’s Creed Odyssey 1080p Very High | 61fps | 62fps | |||||
Assassin’s Creed Odyssey 1440p Very High | 45fps | 44fps | |||||
*lower is better |
There’s usually a clear distinction between the Core i9-11900K and Core i5-11600K except in rare cases that don’t benefit from having more cores, and where perhaps the latter’s higher clock speed gives it an advantage. The higher core count of the Ryzen CPUs pull ahead in multi-threaded content creation tests but the Rocket Lake CPUs are no slouches, and you can see that there are gen-on-gen advantages. If you’re upgrading from even a two-year-old PC, you’ll see improvements in many productivity as well as gaming workloads.
We also tested the onboard GPU with a few synthetic benchmarks. Intel has the advantage over AMD here, since most Ryzen CPUs do not have integrated graphics at all. Sure, if you’re buying a CPU at this level, you’ll very likely want a powerful GPU, but in times like these when graphics cards are simply unavailable, and for troubleshooting or surviving a GPU failure, at least you can boot up your PC and use it. We clearly aren’t expecting gaming-grade performance, and that’s what we get. Interestingly, the Core i5 is just marginally better than the Core i9 in these tests. Gaming performance of course depends mainly on the GPU you use, and in our tests, we see that the CPUs are not causing bottlenecks here.
As with many “gaming” motherboards, some of Intel’s stock recommendations are overridden so that the CPU can draw more power and run hotter for longer than officially intended. A BIOS option called “Multi Core Enhancement” is set to Auto by default, and that was left in place for all tests. This is a relatively safe automatic overclocking feature that allows all cores to run at the single-core Turbo speed, which has somewhat the same effect as Intel’s new Adaptive Boost Technology.
In fact, ABT is turned off by default and this was tested separately on the Core i9-11900K. In Cinebench R20, the single-core and multi-core scores went from 628 and 5,927 to 635 and 6,116 respectively. Geekbench 5 also presents separate single-core and multi-core scores, when went from 1,777 and 9,536 to 1,770 and 9,847 respectively.
One of the main reasons to spend money on Intel’s -K CPUs and a Zxxx-series motherboard is of course overclocking, and you’ll need to spend even more on a good AIO liquid cooler as well as high-spec RAM. You can tinker manually in the BIOS or use Asus’ freely downloadable software. If you choose the latter approach, there’s a one-click overclocking option which stress-tests your PC to determine how far your cooler and other hardware will let the CPU go. In our case, we didn’t get spectacular results, only managing 5.3GHz sustained across multiple cores. Cinebench R20’s single-core and multi-core scores went up a bit to 641 and 6,184, but Geekbench posted lukewarm figures of 1,747 and 9,841 respectively.
The Core i9-11900K will require heavy duty cooling if you’re planning to overclock, and even ABT will strain your cooler more than usual. The Corsair H150i Elite Capellix that we used is certainly expensive at around Rs. 15,500, but its enormous radiator surface area and three 120mm 2400rpm maglev fans handled our review workloads well. The fans of our cooler, power supply and graphics card could all spin down completely when not stressed, which means even a very high-end PC build around today’s flagship CPUs can be completely silent.
Verdict
If things had gone according to plan for Intel many years ago, the 11th Gen Core family would have been a very different product, possibly based on a 7nm or even smaller process. Intel certainly wouldn’t have wanted to regress in terms of core count for its halo product of this generation, and performance has not grown by huge leaps across the board. Intel decided that it would be best to have eight cores in the thermal and power envelopes that could be managed with this architecture, but ten or more would have shown significant improvement in multi-threaded tests. The integrated graphics could also have used a major boost over what we’re currently getting.
With that said, the Core i9-11900K is a formidable processor and should be able to keep all your AAA games running nicely for the foreseeable future. Intel has managed to keep pace with AMD’s latest and greatest desktop Ryzens when it comes to gaming, and that’s what most people will be looking at buying this CPU for.
Intel has strenuously denied that the Rocket Lake family is a stopgap measure to keep the market fresh, but we already know that the 10nm 12th Gen Alder Lake architecture is on track to ship in late 2021, which means Rocket Lake will not have a long shelf life. Alder Lake will benefit from 10nm transistor sizes, Intel’s evolving SuperFin production process, the hybrid design with heterogenous cores, a new socket, DDR5 RAM, and (reportedly) PCIe 5.0 – all in all, it’s shaping up to be a massive shift and potentially one worth waiting for.
Buying a new CPU to replace a 10th Gen model won’t get you very much return on investment – the Core i9-10900K with 10 cores sells for around Rs. 48,000 these days which is not bad at all, compared to the Rs. 61,000 average street price for the Core i9-11900K. In fact, people might still want to go for the older model given these prices.
A lot of potential Core i9-11900K buyers will also have to ask themselves whether the Core i7-11700K, priced at Rs. 38,500 with the same number of cores, makes more sense. If Adaptive Boost Technology isn’t going to make a huge difference to you, you can save quite a bit of money. If you aren’t going to overclock, the non-K SKUs make more sense too. Similarly, with the Core i5-11600K, you might find better value for money a little down the stack, in the form of the Core i5-11500.
AMD’s Ryzen 9 5900X with 12 cores will do better in multi-threaded applications and might be a better bet in terms of future-readiness. Street prices are a bit lower right now, at around Rs. 55,000, but you do have to look at the costs and availability of motherboards as well. The Ryzen 9 5950X with 16 cores costs around Rs. 76,000, putting it a tier above the Core i9-11900K, but again, it’s the better option if cores and threads matter and if your time is money. For example if you play games and do a lot of content creation work on the same PC, you need a balance, not just excellent single-threaded performance.
It’s generally a bad time to be building or upgrading a gaming PC anyway, thanks to the complete lack of graphics cards in the market (at reasonable prices at least) and slowly rising prices of RAM and SSDs. What it all comes down to is that for most people, the 11th Gen Core desktop family, and particularly the Core i9-11900K, will be skippable. If you can wait till the end of this year for the 12th Gen to launch (and hopefully for graphics card prices to stabilise as well), you’ll be much happier. However, if you do need a new, powerful gaming PC right now or are upgrading only your CPU and motherboard, the 11th Gen Rocket Lake family does make sense.
Intel Core i9-11900K
Price (MOP): Rs. 61,000
Pros
- Great performance in lightly threaded tests
- Platform features including Thunderbolt 4, Wi-Fi 6E
- Will work with previous-gen motherboards
Cons
- Expensive
- Basic integrated graphics
- Fewer cores than the Core i9-10900K
Ratings (out of 5)
- Performance: 4.5
- Value for Money: 4
- Overall: 4.5
Intel Core i5-11600K
Price (MOP): Rs. 24,500
Pros
- Good performance overall
- Platform features including Thunderbolt 4, Wi-Fi 6E
- Will work with previous-gen motherboards
Cons
- Basic integrated graphics
Ratings (out of 5)
- Performance: 3.5
- Value for Money: 4
- Overall: 4
Asus ROG Maximus XIII Hero
Price (MOP): Rs. 42,000
Pros
- Good looks, clean layout
- Four M.2 slots, two Thunderbolt 4 ports
- Stable performance
- Well designed UEFI BIOS interface
Cons
- Very expensive
Ratings (out of 5)
- Features: 4.5
- Performance: 4.5
- Value for Money: 4
- Overall: 4.5